A systematic review of the literature - Health Services Assessment

A systematic review of the literature

July 2009 The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care

settings - the ABCs

David Brinson Wasan Ali

This report should be referenced as follows: Brinson, D and Ali, W. The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings - the ABCs. HSAC Report 2009; 2(9) Health Services Assessment Collaboration (HSAC) ISBN 978-0-9864551-1-7 (online) ISBN 978-0-9864551-2-4 (print) ISSN 1178-5748 (online) ISSN 1178-573X (print)

i

The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings - the ABCs

Review Team This review was undertaken by the Health Services Assessment Collaboration (HSAC). HSAC is a collaboration of the Health Sciences Centre of the University of Canterbury, New Zealand and Health Technology Analysts, Sydney, Australia. This report was authored by David Brinson, Researcher, who developed the scoping protocol, and undertook the initial scoping literature search. Frances L. Bluhdorn, Medical Librarian, undertook the systematic literature search and Franziska Gallrach, Research Assistant, assisted with the management of the database and retrieval of the articles. David Brinson and Dr Wasan Ali, Researcher, extracted the data, conducted the critical appraisals and wrote the results narratives. David Brinson collated the data, wrote the discussion, conclusions and recommendations, and prepared the final report. Acknowledgements Dr Ray Kirk, an HSAC Director, peer reviewed the final draft. Cecilia Tolan, Administrator provided document formatting. Staff at the University of Canterbury Libraries assisted with the retrieval of articles via interloan. Sub-editing was performed by Dr Christian Long. The current review was conducted under the auspices of a contract funded by the New Zealand Ministry of Health. This report was requested by Karen Evison, National Programme Manager, Tobacco Policy and Implementation, Sector Capability and Innovation Directorate, of the New Zealand Ministry of Health. We thank Karen and Dr Ashley Bloomfield, Acting Deputy Director-General Sector Capability and Innovation Directorate, Ministry of Health, and Dr Hayden McRobbie, Consultant, Inspiring Limited, for assisting in developing the scope of the review and providing background material for the review. The systematic review of the evidence will ultimately be used to inform policy decisions made in conjunction with other information. The content of the review alone does not constitute clinical advice or policy recommendations. Copyright Statement & Disclaimer This report is copyright. Apart from any use as permitted under the Copyright Act 1994, no part may be reproduced by any process without written permission from HSAC. Requests and inquiries concerning reproduction and rights should be directed to the Director, Health Services Assessment Collaboration, Health Sciences Centre, University of Canterbury, Private Bag 4800, Christchurch, New Zealand HSAC takes great care to ensure the accuracy of the information in this report, but neither HSAC, the University of Canterbury, Health Technology Analysts Pty Ltd nor the Ministry of Health make any representations or warranties in respect of the accuracy or quality of the information, or accept responsibility for the accuracy, correctness, completeness or use of this report. The reader should always consult the original database from which each abstract is derived along with the original articles before making decisions based on a document or abstract. All responsibility for action based on any information in this report rests with the reader. This report is not intended to be used as personal health advice. People seeking individual medical advice should contact their physician or health professional.

ii


The views expressed in this report are those of HSAC and do not necessarily represent those of the University of Canterbury New Zealand, Health Technology Analysts Pty Ltd, Australia or the Ministry of Health. Contact Details Health Services Assessment Collaboration (HSAC) Health Sciences Centre University of Canterbury Private Bag 4800 Christchurch 8140 New Zealand Tel: +64 3 345 8147 Fax: +64 3 345 8191 Email: [email protected] Web Site: www.healthsac.net

iii


Executive Summary Introduction The purpose of this systematic review is to provide a summary of the evidence pertaining to the effectiveness of interventions aimed at increasing the likelihood of Health Care Professionals (HCPs) initiating elements of the ABC approach in a primary care setting. That is, any intervention that aims to increase the documentation of smoking status (Ask), provision of brief advice to stop smoking (Brief advice), the offer and/or provision of, or referral to, smoking cessation treatments or services (Cessation support), compared to usual care. A current challenge is to get more smokers to see the importance of stopping as soon as possible, and to use the proven effective treatments available. Primary care has a key role to play in helping to meet this challenge, but lacks clear guidance on how to best undertake this task; therefore this systematic review was requested by the New Zealand Ministry of Health’s Tobacco Policy and Implementation Group. The systematic review was requested to inform national policy, assist District Health Boards (DHBs) and Primary Health Organisations (PHOs) at a regional and local level, and ultimately help people who smoke to stop. Methods The primary research question to be addressed by this review was: What interventions have been successfully implemented in primary care and increase (a) assessment of smoking status (b) the provision of brief advice and (c) the offer of effective smoking cessation treatments (including counselling, pharmacotherapy, or referral for smoking cessation treatment), compared to standard care?

A systematic method of literature searching, study selection, data extraction and appraisal was employed in the preparation of this report. The literature was searched using the following bibliographic databases: Medline and Embase. The Cochrane Database of Systematic Reviews (CDSR), the Database of Abstracts of Reviews of Effects (DARE), and other sources where applicable were also searched to help identify existing systematic reviews. In addition, the bibliographies of included papers were examined for relevant studies. Searches were limited to English-language material published from 1990 to 2009 inclusive. The Australian National Health and Medical Research Council (NHMRC) dimensions of evidence, levels of evidence and quality assessment criteria were used to evaluate each of the included studies. Data was extracted onto standardised data extraction forms by two reviewers. An iterative process was used to select articles for review. In the first step, studies were screened by their titles and abstracts or summaries. A publication was included for further review if it fulfilled all of the following criteria in its title and abstract or summary – 1. The study was conducted in a population comparable to that of New Zealand, 2. the interventions were appropriate and were administered in a primary care setting, 3. a valid comparison group was present, and 4. an outcome of the study included one or more of the following: assessment and/or

documentation of patients’ smoking status; the provision and/or documentation of brief advice; the offer and/or delivery of appropriate smoking cessation treatments, including counselling, pharmacotherapy, or referral to specialist smoking cessation services.

Studies that did not satisfy these four criteria were excluded. In addition, a study was excluded if it had a sample size less than 20 in each group, if it was published in a language other than English, if the study was pre-1990, or if the full text was not available from any source.

iv


Full texts of all included studies were critically appraised by two independent researchers to further establish the eligibility of each study on the basis of the rigorousness of the research methods, quality of evidence, quality of their reporting and relevance of the research with respect to the research question. In order to base conclusions on the best available evidence, only studies that were peer reviewed were selected. The information content of the publications was overviewed and the interventions were then categorised into six predominant intervention types based on the number of eligible studies reporting such interventions. These six categories were – (1) Training (2) Multi-component interventions (3) Reminders (4) Financial incentives (5) Audit & Feedback (6) Other interventions There was notable heterogeneity in the way in which authors described similar interventions, and in the terminology and claimed theoretical underpinnings for some of the interventions. However, some commonality and overlap was also evident within categories and between categories with respect to the essence of the intervention. Overall, the groupings reported here were formed in an attempt to best represent the fundamental component of each intervention within the total body of evidence in a way that is simplified and potentially generalisable to New Zealand’s primary care context. Components of the ABC approach for smoking cessation were the primary outcomes of interest for this review. During the initial phase of the search, selection, and retrieval of publications, it was apparent that these outcomes were commonly reported in differing combinations. For example, some studies only reported the assessment and/or documentation of patients’ smoking status, others reported only referral to specialist smoking cessation services, and others reported many outcomes. Studies were included if they reported any of these smoking outcomes. Studies within each of the six types of interventions were reviewed with respect to the ABC outcomes; summaries and conclusions are provided for each of the six groupings. Subsequently, the results of all of the six individual summaries were synthesised to arrive at an overview of the general patterns of effectiveness. Results The literature search identified a total of 464 citations. After consideration of titles and abstracts using the study selection criteria, 130 full papers were retrieved and scrutinised in detail for possible inclusion in the review. As a result, 42 publications were ultimately included in the review. Key findings Of the 42 papers identified as eligible, five were systematic reviews/meta-analyses and 37 were original primary research studies. The review articles differed considerably with respect to their inclusion criteria, in particular, the different types of study designs included for review, the publication year range, and the process and/or patient outcomes reported. One review included only randomised controlled trials or cluster randomised controlled trials, whereas the other four reviews also included non-randomised controlled clinical trials, controlled before

v


and after trials and interrupted time-series designs. The reviews also differed with respect to the interventions; one review evaluated the effectiveness of a single intervention (training), whereas the other four evaluated the effectiveness of up to nine different types of interventions, alone and in combination. The 37 original studies represented a range of study designs. These included 14 cluster randomised controlled trials (where the unit of randomisation was the clinic or practice not the patient) (C-RCTs) and 11 randomised controlled trials (RCTs); all were evidence level II. The remaining 12 studies were of other designs: two were non-randomised controlled clinical trials (evidence level III-2); six were controlled before and after trials (evidence level III-3); three were time-series designs (evidence level III-2); and one study analysed comparative data from two temporally different cross-sectional surveys (evidence level IV). The 37 original studies also differed widely with respect to study duration. The mean study duration of the randomised trials was 14 months; however, study durations ranged widely from three or four months to four years. For non-randomised trials, the mean duration was 30 months, with the longest running trial being a 15 year interrupted time series evaluation. Overall, the studies involved more than 3000 health care professionals, the majority were general practitioners (GPs) working in primary care/community settings. Other health care professionals included nurse practitioners, physician assistants, and other rooming and administrative staff. Health care professionals were studied as individuals, in groups, or within an entire clinic or practice. The number of patients involved in the studies ranged between a few hundred and 1.6M (electronic patient records), with the majority involving adult patients. Two studies involved adolescents, one study involved pregnant women, another study involved patients with diabetes, and another study involved patients with established coronary heart disease. More often than not, the patients were ‘passive’ and not formally enrolled in the study, although consent to complete interviews and questionnaires was usual. Finally, while the population (health care professionals) and setting (primary care) were similar in all of the studies, two variations of context were apparent: smoking specific studies (n = 32) and general prevention studies (n = 10). Smoking specific studies were those studies in which the intervention and outcomes related only to health care professionals’ smoking cessation activities. General prevention studies were those studies in which two or more preventive manoeuvres were addressed within one trial (e.g. smoking cessation and diet and exercise counselling). General ‘prevention’ in this context refers to services or activities directed primarily toward early detection and treatment or prevention of disease, not prevention of smoking initiation per se. When relevant smoking outcomes were reported separately within general prevention trials and the authors’ discussion and conclusions differentiated between the different outcomes, then the study was included, and only the smoking outcomes are reported in this review. Overall, the methodological quality of the original research studies was fair to good. In general, studies provided information on their analyses but many did not report on the intention to treat analysis or on confounding factors, although most provided a good level of detail when describing the analyses. Many of the studies did not clearly describe the comparison group. In the main, the interventions were compared with ‘no-intervention’ or ‘usual care’ but this was often not explicit. Due to the nature of the interventions and the challenges of conducting (pragmatic) randomised trials in clinic settings, the majority of the studies were not blinded or it was not clear who was blinded to the study or question. However, most used blinded research assistants for chart abstraction. Many of the studies used subjective measures for assessing outcomes, most commonly patient exit interviews. These are likely to be subject to some degree of recall bias. These tools for measuring the primary outcomes were usually adequately described, but often the reliability and validity were not cited or demonstrated. Studies in general were subject to selection biases (e.g. volunteerism), the percentages of

vi


eligible subjects actually participating were low in some studies and participants were not always wholly representative of the target population. Effectiveness As previously noted, ‘effectiveness’ in this context relates to the documentation of smoking status, provision of brief advice to stop smoking, and the offer and/or provision of, or referral to smoking cessation treatments or services– not quit rates. The inherent assumption here is that if people receive brief advice and support to quit, some will go on to do so (using proven effective treatments e.g. NRT and counselling) and quit rates will be further increased above the unaided quit rate. These observations suggested the following emergent patterns of effectiveness, as identified in the published literature – Training: training health care professionals is more effective than no training in increasing the numbers of smokers who have their smoking status recorded in the medical records, are offered brief advice about stopping, and subsequently receive at least some cessation support. It appears that the most effective form of training is training health care professionals in the skills of actually delivering smoking cessation treatments to their patients (for example, training in effective counselling techniques or methods). Training health care professionals to design and implement systems and to initiate system-wide changes does not appear to be a good use of resources. Thus, training will need to be focused on skills, be practice-based, utilise existing opportunities, and be supplemented by other interventions or systematic processes (for example, reminder systems). It appears that training-based interventions perform best when they are implemented alongside other interventions. Multi-component: multi-component interventions, based on training, have been shown to be effective in increasing the frequency with which primary care clinicians intervene with their patients who smoke. However, no clear patterns emerged regarding which combinations of intervention components were more or less effective or what relative contributions the individual components might make. Simply increasing the number of intervention components delivered within a programme does not necessarily increase its overall performance. Reminders: all the trials of reminders demonstrated an increase in the frequency of health professional intervention. Programmes were found to be generally effective in changing health care professionals’ advice-giving rates and counselling activities. Sharing the responsibility for office-based smoking cessation activities among all staff appears to be a promising strategy and addressing the barrier of ‘clinician forgetfulness’ is one probable core constituent. Financial Incentives: the studies of financial incentives suggest that performance targets coupled with financial incentives do appear to change health care professionals’ clinical behaviours; however, there is insufficient evidence to determine what the optimal performance threshold levels and/or the optimum level of incentives might be. In addition, the possibility of unintended consequences has not been fully explored and it is also unclear whether or not the apparent positive influence of financial incentives on clinical practice could be sustained over time. Audit & Feedback: audit & feedback appears to be an effective means of changing health care professionals’ behaviour. Audit & Feedback requires the widespread availability of electronic medical record systems capable of quickly and easily generating comparative feedback reports. However, the optimal methods of delivering such feedback are not fully understood.

vii


Other: there is insufficient evidence to draw robust conclusions for other intervention types. However, the results from two studies suggest that other low-intensity single component interventions are likely to be of limited value. Conclusions The review conclusions are based on the critical appraisal of the current literature published on the effectiveness of interventions aimed at increasing the likelihood of Health Care Professionals delivering elements of the ABC approach in a primary care setting. Being a systematic review, the overall quality of the review and its recommendations are based on the quality of information presented in the individual studies. Studies were selected on the basis that they reflected the New Zealand context/setting as much as possible, at the level of the patient-provider interaction. An exact match is not possible and no studies were identified that were conducted in New Zealand. A large proportion of the research was conducted in the USA, and differences in system and organisational aspects of care may have shaped the findings. Generalisability of findings to New Zealand populations should therefore be considered on a case-by-case basis with regard being given to the possible influences of contextual, environmental and high-level policy factors. Several authors question the effectiveness of interventions aimed at increasing the delivery of smoking cessation treatments outside of controlled clinical trial settings. That is, their feasibility and sustainability when established in typical primary care settings – removed from the external influences of researcher expertise, funding and other resources. While questions of generalisability are applicable to all clinical trials, this may be more problematic when the intervention is ‘operator dependent’, that is, dependent (at least in part) on the attitudes, enthusiasm, competence and motivation of health care professionals, funders, and other stakeholders. It is recommended that any issues of generalisability be carefully considered but that generalisability should not be seen as a barrier to implementation. Programmes that implement these interventions should carefully monitor and evaluate their impact. Policy-makers need to consider the wider implications associated with changes in clinical practice and assess their likely impact on resourse utilisation and/or any possible unintended consequences. For example, the ultimate success of any system that increases the rates of referral of smokers to speciality counselling services is contingent on these services being available and adequately resourced. Preventive activities in general (i.e. services directed primarily toward early detection and treatment or prevention of disease, not prevention of smoking initiation per se) are likely to be in competition with the other pressing demands on general practitioners, and providing patients with quality smoking cessation support is no exception. Team approaches, in which the different tasks of the ABC approach are shared between different health professionals and organisations, is one possible solution to this problem. Patient outcomes were not within the scope of this review. However, 12 studies did report patient outcomes in addition to the ABC outcomes of interest and one trend was apparent – that of improved process outcomes but little strong evidence to suggest that the interventions lead to changes in patients’ smoking behaviour. Nevertheless, this lack of evidence is not evidence of lack of efficacy. In most cases, patient outcomes (e.g. quit rates) were not the primary outcomes of interest and some studies were not sufficiently powered to detect what can be relatively small, yet important effects. However, several authors stated largely similar conclusions, that changing patients’ smoking behaviours is much more complicated than changing the behaviour of individual clinicians, and that influencing patient outcomes may require more intensive and sustained efforts, with multiple strategies which are reinforced at higher policy levels.

viii


There is insufficient information to inform an economic analysis, given the multiple intervention types and/or components identified and the range of ABC outcomes reported. This is complicated further by inconsistent and/or partial reporting of the outcomes and the use of non-standardised measurement instruments. The cost-effectiveness of programmes aimed at increasing Health Care Professionals’ delivery of these smoking cessation interventions is not well documented. Estimating resource utilisation and any possible cost off-sets and/or savings to the health care system remains beyond the scope of this report. It is clear that increasing the delivery of effective smoking cessation treatments in primary care settings is an important if not essential step in achieving better patient care, and the literature indicates that several interventions offer much promise. It is known that changes in health care outcomes tend to lag behind those in physician performance, and this trend has been observed here. Possible barriers include patients’ not accepting health care professionals’ recommendations, the socioeconomic and educational status of patients, other environmental influences, and the effectiveness of the clinical interventions themselves. Notwithstanding these limitations, smoking cessation remains an important focus, as stopping smoking is probably the best way an individual can improve his or her health. The literature reviewed suggests that health care professionals who are properly trained, supported, resourced, prompted, given personalised feedback are not subject to disincentives, and for whom the task is made uncomplicated, are more likely to deliver smoking cessation interventions to all their patients who smoke when compared to health care professionals delivering usual care.

ix


Table of Contents Review Team..................................................................................................................i Acknowledgements........................................................................................................i Copyright Statement & Disclaimer .............................................................................i Contact Details ............................................................................................................ ii Executive Summary ................................................................................................... iii

Introduction ............................................................................................................ iii Methods.................................................................................................................. iii Results .....................................................................................................................iv Key findings ............................................................................................................iv Effectiveness ...........................................................................................................vi Conclusions............................................................................................................vii

Table of Contents ........................................................................................................ix List of Tables ............................................................................................................. xii List of Figures.............................................................................................................xv List of Abbreviations and Acronyms ......................................................................xvi Introduction..................................................................................................................1

Objective ..................................................................................................................1 Description of indication/condition .........................................................................1 Description of intervention/technology....................................................................3

Background ...........................................................................................................3 Smoking cessation treatments ...............................................................................4 Guideline implementation .....................................................................................4 Smoking cessation cost-effectiveness ...................................................................5 The issue................................................................................................................5

Structure of report ....................................................................................................5 Methods.........................................................................................................................7

Research questions ...................................................................................................7 Literature search.......................................................................................................8

Bibliographic databases.........................................................................................8 Review databases ..................................................................................................8 HTA Groups ..........................................................................................................8 Clinical Practice Guidelines ..................................................................................8 Literature search notes...........................................................................................9

Assessment of study eligibility ................................................................................9 Appraisal of included studies .................................................................................11

Data extraction ....................................................................................................14 Data synthesis......................................................................................................15

Limitations of the review methodology .................................................................15 Results .........................................................................................................................17

Review of the research question, scope, overview of the study characteristics and results ..............................................................................................................17

x


Review question ..................................................................................................17 Scope ...................................................................................................................17

Study characteristics: overview..............................................................................18 Quality .................................................................................................................19 Population............................................................................................................19 Interventions........................................................................................................19 Comparisons........................................................................................................21 Outcomes.............................................................................................................21 Context ................................................................................................................22

Systematic reviews: introduction ...........................................................................24 Systematic reviews and meta-analyses: results......................................................29 Other reviews .........................................................................................................32 Systematic reviews: summary................................................................................32 Systematic reviews: conclusions............................................................................33 Original primary studies: introduction ...................................................................39 Overview of intervention groupings ......................................................................39

Training ...............................................................................................................39 Multi-component .................................................................................................39 Reminders............................................................................................................39 Financial incentives.............................................................................................40 Audit & Feedback ...............................................................................................40 Other....................................................................................................................40

Original primary study results: by intervention type and level of evidence ........41 Training.......................................................................................................................41

Introduction ............................................................................................................41 Training: results from 15 studies ...........................................................................42 Training: results summary......................................................................................49 Training: conclusions .............................................................................................52

Multi-component Interventions ................................................................................61 Introduction ............................................................................................................61 Multi-component: results from eight studies .........................................................61 Multi-component: summary...................................................................................66

Smoking specific studies .....................................................................................66 General prevention studies ..................................................................................67

Multi-component: conclusions...............................................................................67 Reminders ...................................................................................................................77

Introduction ............................................................................................................77 Reminders: results from five studies......................................................................77 Reminders: summary .............................................................................................80 Reminders: conclusions .........................................................................................80

Financial Incentives ...................................................................................................85 Introduction ............................................................................................................85 Financial incentives: results from four studies.......................................................85

xi


Financial incentives: summary...............................................................................87 Financial incentives: conclusion ............................................................................87

Audit and Feedback ...................................................................................................91 Introduction ............................................................................................................91 Audit & Feedback: results from three studies........................................................91 Audit & Feedback: summary .................................................................................93 Audit & Feedback: conclusion...............................................................................94

Other Interventions....................................................................................................97 Introduction ............................................................................................................97 Other interventions: results from two studies ........................................................97 Other interventions: summary................................................................................97 Other interventions: conclusion .............................................................................98

Body of Evidence......................................................................................................101 Discussion..................................................................................................................107

Differences across outcomes .............................................................................107 Differences across contexts: smoking specific vs. general prevention .............109

Rapid Review: Results and Implications for Practice ..........................................111 Contextual limitations and policy implications.................................................112 Other issues: Patient outcomes..........................................................................112

Economic Considerations ........................................................................................115 Summary and Conclusions......................................................................................117 References .................................................................................................................119 Glossary.....................................................................................................................125 Appendix A: Working Party Membership .......................................................131 Appendix B: Search Strategy ............................................................................133

EMBASE.com...................................................................................................133 The Cochrane Library .......................................................................................135

Appendix C: Included Studies ...........................................................................139 Appendix D: Study Characteristics: All Primary Studies ..............................141 Appendix E: Excluded Studies Annotated by Reason for Exclusion.............157 Appendix F: Quality Checklists for Appraising Interventions ......................175 Appendix G: Data Extraction Tables ................................................................177

xii


List of Tables Table 1: PICO criteria for determining study eligibility.................................7 Table 2: Nature of the evidence .........................................................................8 Table 3: Exclusion criteria and coding............................................................10 Table 4: Application of selection criteria to citations ....................................11 Table 5: NHMRC Dimensions of evidence .....................................................12 Table 6: NHMRC additional levels of evidence and grades

(NHMRC 2008), by question type ....................................................13 Table 7: Quality criteria for different levels of evidence

(NHMRC, 2000b) ...............................................................................14 Table 8: Reporting biases in systematic reviews ............................................15 Table 9: Included intervention types...............................................................20 Table 10: Other intervention types....................................................................21 Table 11: Systematic review characteristics: The effectiveness of

interventions to increase the delivery of effective smoking cessation treatments in primary care settings .................................25

Table 12: Other review characteristics: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings .....................................................................28

Table 13: Systematic review results: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings .....................................................................34

Table 14: Other review results: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings .........................................................................38

Table 15: Intervention groupings and number of trials ..................................40 Table 16: Sub-categories of training as defined by the original study

authors.................................................................................................42 Table 17: Original study results: The effectiveness of Training to increase

the delivery of effective smoking cessation treatments in primary care settings.........................................................................................53

Table 18: Original study results: The effectiveness of Multi-component interventions to increase the delivery of effective smoking cessation treatments in primary care settings .................................68

Table 19: Original study results: The effectiveness of Reminder systems to increase the delivery of effective smoking cessation treatments in primary care settings .....................................................................81

Table 20: Original study results: The effectiveness of Financial Incentives to increase the delivery of effective smoking cessation treatments in primary care settings .....................................................................89

Table 21: Original study results: The effectiveness of Audit & Feedback to increase the delivery of effective smoking cessation treatments in primary care settings .....................................................................95

Table 22: Original study results: The effectiveness of ‘other’ interventions to increase the delivery of effective smoking cessation treatments in primary care settings .....................................................................99

xiii


Table 23: Body of evidence: Effectiveness of interventions ..........................102 Table 24: Study characteristics: all primary studies .....................................141

Systematic reviews...............................................................................................177 Table 25: Anderson et al. (2004) ......................................................................177 Table 26: Dickey et al. (1999) ...........................................................................181 Table 27: Fiore et al. (2008)..............................................................................186 Table 28: Grimshaw et al. (2004).....................................................................192 Table 29: Lancaster et al. (2000)......................................................................196

Original studies ....................................................................................................199 Table 30: Amundson et al. (2003) ....................................................................199 Table 31: Bentz et al. (2007) .............................................................................201 Table 32: Borgiel et al. (1999) ..........................................................................203 Table 33: Cockburn et al. (1992) .....................................................................205 Table 34: Coleman (2007).................................................................................207 Table 35: DePue et al. (2002)............................................................................209 Table 36: Dietrich et al. (1992).........................................................................211 Table 37: Fiore et al. (1995)..............................................................................213 Table 38: Flocke et al. (2006)............................................................................215 Table 39: Goldstein et al. (2003) ......................................................................217 Table 40: Hogg et al. (2008)..............................................................................219 Table 41: Joseph et al. (2004) ...........................................................................221 Table 42: Katz et al. (2004)...............................................................................223 Table 43: Kottke et al. (1992) ...........................................................................226 Table 44: Krist et al. (2008) ..............................................................................228 Table 45: Lemelin et al. (2001).........................................................................230 Table 46: Lennox et al. (1998) ..........................................................................233 Table 47: McRobbie et al. (2008) .....................................................................235 Table 48: McPhee et al. (1991) .........................................................................237 Table 49: Millett et al. (2007) ...........................................................................239 Table 50: Moher et al. (2001) ...........................................................................241 Table 51: Morgan et al. (1996) .........................................................................244 Table 52: Ockene et al. (1994) ..........................................................................246 Table 53: Ockene et al. (1997) ..........................................................................248 Table 54: Ozer et al. (2001) ..............................................................................251 Table 55: Pbert et al. (2006) .............................................................................254 Table 56: Piper et al. (2003)..............................................................................256 Table 57: Richmond et al. (1998) .....................................................................259 Table 58: Roski et al. (2003) .............................................................................261 Table 59: Rothemich et al. (2008) ....................................................................263 Table 60: Secker-Walker et al. (1992) .............................................................266 Table 61: Solberg et al. (2000)..........................................................................269 Table 62: Strecher et al. (1991) ........................................................................271 Table 63: Tremblay et al. (2001) ......................................................................274

xiv


Table 64: Unrod et al. (2007)............................................................................276 Table 65: Wadland et al. (2007) .......................................................................279 Table 66: Young et al. (2002) ...........................................................................281

xv


List of Figures Figure 1 The prevalence of cigarette smokers, ex-smokers and never

smokers in the New Zealand population 1976 – 2006 (census data). These data show that there has been greater success with reducing initiation than with getting existing smokers to quit (Ministry of Health, 2009). ......................................................................1

Figure 2 Prevalence of current smokers (%), by gender and ethnicity: reproduced from the New Zealand Tobacco Use Survey 2006 (Ministry of Health, 2007b). ....................................................................2

Figure 3 Review scope and comparison of the ABC approach with the 5As framework .......................................................................................17

Figure 4 The ‘cube’ depicts the identified interventions, outcomes and contexts....................................................................................................22

Figure 5 The effectiveness of the 15 included studies of training interventions, grouped by the predominant training approach: training HCPs in specific cessation skills vs. training HCPs to design and implement systems. .............................................................51

Figure 6 An illustration of the apparent decline in cessation activities by HCPs with the increasing complexity of the activities. Data sourced from three good quality, moderately large, and relatively recent studies studies as examples: Bentz et al. (2007), Goldstein et al. (2003) and DePue et al. (2002).............................................................108

xvi


List of Abbreviations and Acronyms χ2 Chi squared test 5As ask, advise, assess, assist, arrange 95%CI 95% confidence interval AHRQ Agency for Healthcare Research and Quality (formerly AHCPR) (USA) AHTAC Australian Health Technology Advisory Committee ANCOVA Analysis of Co-variance ANOVA Analysis of Variance ARR Absolute risk reduction BMI Body mass index CBA Controlled before-after trial CCT Controlled clinical trials CE Clinical effectiveness CFPC College of Family Physicians of Canada CHD Coronary heart disease CI Confidence interval CINAHL Cumulative Index to Nursing and Allied Health Literature CME Continuous medical education CMEP Continuing medical educational plan C-RCT Cluster randomised controlled trial EHR Electronic health record eLinkS Electronic linkage system EMR Electronic medical record EPOC Effective practice organisation of care FFS Fee for service FP Family practitioner GEEs Generalised estimating equations GIMC General internal medicine clinic GP General Practitioner HCP Health Care Professional HFA Health Funding Authority (NZ) HMO Health Maintenance Organisation HRA Health risk appraisal HSO Health Services Organisation HTA Health technology assessment ICC Inter-correlation coefficient ICCs Intra-cluster correlation coefficients INAHTA International Network of Agencies for Health Technology Assessment ITS Interrupted time-series ITT Intention to treat analysis MOH Ministry of Health (NZ) NCI National Cancer Institute NHMRC National Health and Medical Research Council (AU) NHS National Health Service (UK) NNT Number needed to treat NRT Nicotine replacement therapy NZ New Zealand NZHTA New Zealand Health Technology Assessment OECD Organisation for Economic Co-operation and Development OR Odds ratio ORP Outcomes Recognition Programme

xvii


PAR Practice assessment report PCTs Primary Care Trusts PEI Patient exit interview PGY Post graduate year RCT Randomised controlled trial RR Rate ratio RR Relative risk SD Standard deviation SI Special intervention SSCS Specialist smoking cessation services THIN The Health Improvement Network TS Time-series design UC Usual care UK United Kingdom UMMS University of Massachusetts Medical School USA United States VA Veterans Affairs clinic setting WHO World Health Organisation

1


Cigarette smoking prevalence 1976 - 2006 (census data)

0

10

20

30

40

50

60

1976 1981 1996 2006Year

Perce

ntage

SmokersEx-smokersNever-smokers

Introduction Objective The purpose of this systematic review is to provide a summary of the evidence pertaining to the effectiveness of interventions aimed at increasing the likelihood of Health Care Professionals (HCPs) initiating elements of the ABC approach in primary care setting. That is, any intervention that aims to increase the documentation of smoking status (Ask), provision of brief advice to stop smoking (Brief advice), and the offer or provision of or referral to specialist smoking cessation services (Cessation support) compared to usual care. A current challenge is to get more smokers to see the importance of stopping as soon as possible and to use the proven effective treatments available. Primary care has a key role to play in helping to meet this challenge, but lacks clear guidance on how to best undertake this task. This systematic review was requested by the New Zealand Ministry of Health’s Tobacco Policy and Implementation Group. The systematic review was requested to inform national policy, assist District Health Boards (DHBs) and Primary Health Organisations (PHOs) at a regional and local level, and ultimately help people who smoke to stop. Description of indication/condition

Smoking tobacco is a major public health problem in New Zealand. Several decades after tobacco smoke was identified as a deadly carcinogen, an estimated 20.7%1 of all New Zealanders still smoke tobacco daily. New Zealand has made progress in reducing initiation rates over time and the quit rate has been increasing also, however only slowly. As a result, the prevalence of smoking amongst the general population has been trending down, as seen in the decrease in prevalence of daily smoking from 23.7 percent in 1996 to 20.7 percent in 2006 (Figure 1). Figure 1 The prevalence of cigarette smokers, ex-smokers and never smokers in the New Zealand population

1976 – 2006 (census data). These data show that there has been greater success with reducing initiation than with getting existing smokers to quit (Ministry of Health, 2009).

1 Prevalence estimates for 2006: 20.7% as reported in the Census and 23.5% as reported in the New Zealand Tobacco Use Survey 2006.

2


The prevalence of current smokers varies considerably by ethnicity. The prevalence of smoking among Māori continues to be high (46%) and the prevalence of smoking among Pacific People is also high (36%) (compared with Asian people 12%, and European/Other ethnic groups, 20%) (Ministry of Health, 2006) (Figure 2). Māori and Pacific smoking rates are falling, but still remain substantially higher than those of non-Māori. A significant proportion of the disparities in health seen between Māori and non-Māori may be accounted for by differences in smoking prevalence in these groups and a greater reduction in smoking prevalence in Māori, compared to non-Māori, would close this inequalities gap (Ministry of Health, 2007b). Other countries such as Canada and Sweden have achieved much greater reductions in prevalence than New Zealand. Some states and provinces within OECD countries have achieved daily smoking prevalence rates of less than 15%, notably California (less than 10%) British Columbia (12%) and New South Wales (14%) (Edwards, Wilson, & Thomson, 2008). Prevalence of current smokers by gender and ethnicity

Figure 2 Prevalence of current smokers (%), by gender and ethnicity: reproduced from the New Zealand

Tobacco Use Survey 2006 (Ministry of Health, 2007b). Around 5000 premature smoking-related deaths still occur in New Zealand every year, both as a consequence of active smoking and through exposure to second-hand smoke. Half of these people die in middle age, losing 13 to 15 years of life on average (Public Health Intelligence, 2004). Smoking is the main cause of lung cancer and chronic obstructive pulmonary disease, and is a major cause of heart disease, stroke and a variety of cancers, including mouth, larynx, oesophagus, stomach, cervix, liver and kidney (Ministry of Health, 2006). Smoking is also associated with respiratory infections, asthma, otitis media (glue ear), and the serious adverse outcomes of diabetes (Ministry of Health 1999). These conditions are the major causes of premature death and illness in New Zealand, and smoking cessation can play a significant role in reducing this burden. The overall health burden associated with smoking is sizeable. A recent 50 year follow up of a British doctors’ smoking cohort demonstrated the increased morbidity associated with early and heavy smoking. The study found that up to 66% of lifelong smokers are likely to die from a tobacco-related disease with half these deaths occurring prematurely (Doll, Peto, Boreham, &

3


Sutherland, 2004). In addition, the harm from smoking interacts with other risk factors such as poor diet and poor housing, thus smoking contributes to socioeconomic and ethnic inequalities in health (Ministry of Health, 2004). Smoking cessation is a life-saving intervention, beneficial at any age. Stopping smoking confers immediate health benefits on those who already have smoking related diseases, and cessation prior to age 35-40 is particularly beneficial. Over time, the excess risk of illness and premature death caused by smoking may reduce to that of never smokers (Peto et al., 2000). Helping people who smoke to stop is a leading national health goal. The New Zealand Health Strategy lists the reduction of smoking-related harm as one of its 13 priorities (New Zealand Ministry of Health & King, 2000). To this end, Clearing the Smoke, the national five-year plan for tobacco control in New Zealand (2004-2009), includes promoting smoking cessation as one of its five objectives (Ministry of Health, 2004) and the recently updated Smoking Cessation Guidelines (Ministry of Health, 2007a) are fundamental to achieving this high-priority objective. The goals of the tobacco control plan are:

� to significantly reduce levels of tobacco consumption and smoking prevalence � to reduce inequalities in health outcomes � to reduce the prevalence of smoking among Māori to at least the same level as among

non-Māori � to reduce New Zealanders’ exposure to second-hand smoke.

The stated vision of the national five-year plan for tobacco control is: “For New Zealand to be a country where smokefree lifestyles are the norm” (Ministry of Health, 2004, p.1). Description of intervention/technology Background A range of interventions to reduce smoking prevalence have been implemented in New Zealand and in many countries around the world. These interventions include2: Upstream Interventions

� Interventions that can address upstream societal determinants of smoking such as poverty, inequality, poor education, unemployment, and discrimination

Population level interventions

� Increasing the unit price for tobacco products � Controls on tobacco product marketing � Mass media campaigns � Smokefree sponsorship � Smokefree environment legislation � School-based health education � Age limits on purchasing/use of tobacco

2 Summarised from Wilson, Thomson and Edwards (2008)

4


Individual-level interventions � Provision of smoking cessation advice and support

Reducing the prevalence of smoking in New Zealand through cessation services is a key priority, as identified in the Ministry of Health’s five-year plan, Clearing the Smoke (Ministry of Health 2004). Smoking cessation initiatives funded by the Ministry in order to achieve this goal include (but are not limited to) the Quitline, the Quit mass media campaign, Aukati Kai Paipa, and Health Care Professionals (HCPs) providing cessation support, particularly in primary care settings (Ministry of Health, 2006). Smoking cessation treatments At the individual level, the New Zealand Smoking Cessation Guidelines (2007) provide updated guidance for HCPs to help people who smoke tobacco. These guidelines supersede all earlier national smoking cessation guidelines and are based on an updated comprehensive literature review that summarises the most recent national and international evidence on best practice in smoking cessation (Ministry of Health, 2008). There are some important changes from earlier guidelines: (1) The Guidelines are structured around a new memory aid – ‘ABC’, which aligns with but now supersedes the previously used ‘5As’3 (Ask, Advise, Assess, Assist, Arrange). The three steps are – ‘A’ – Ask about smoking status ‘B’ – Give Brief advice to stop smoking to all people who smoke ‘C’ – Provide evidence-based Cessation support for those who express a desire to stop smoking. (2) That all people who smoke, regardless of whether they express a desire to want to stop or not, should be advised to stop smoking. (3) That support to stop smoking should always be offered to those people who express an interest in stopping. (4) All reference to the ‘Stages of Change’ model have been removed because new research challenges the usefulness of this model in smoking cessation treatment (West, 2005). Guideline implementation Implicit within the New Zealand Smoking Cessation Guidelines is an assumption that health care workers have the prerequisite knowledge, attitudes and skills to support smokers in ways that maximise the smokers’ chances of stopping smoking permanently (Ministry of Health, 2007a). To this end, the document Implementing the ABC Approach for Smoking Cessation: Framework and work programme has been prepared for the Ministry of Health Tobacco Policy and Implementation Team by Martin Jenkins and published in January 2009 by the Ministry of Health. This implementation document outlines, among other things, the different elements and key points of leverage that will be used to implement and support the ABC approach as 3 The 5As framework of ask, advise, assess, assist, arrange, as recommended by the U.S. Department of Health and Human Services Clinical Practice Guidelines 1996, and based on outcomes from six major clinical trials of physician-delivered smoking interventions conducted in the late 1980s (Glynn, Manley, & Pechacek, 1990).

5


well as the work programme and the high-level implementation plan (Ministry of Health, 2009). Smoking cessation cost-effectiveness Smoking cessation services, in general are reported to be extremely cost-effective. International evidence (Centers for Disease Control and Prevention, 1992; Cummings, Rubin, & Oster, 1989; Tsevat, 1992) shows the overwhelming cost-effectiveness of smoking cessation treatments. In New Zealand, an economic evaluation of the Quitline NRT service reported favourable cost-effectiveness (O’Dea, 2004). The post-NRT Quitline programme was found to be “almost certainly a highly cost-effective programme” (O’Dea, 2004, p.4) (the lack of complete certainty arises from the lack of the pre-NRT Quitline effectiveness data). What is not known is the cost effectiveness of interventions that aim to increase the likelihood of Health Care Professionals (HCPs) routinely initiating these treatments in primary care setting. The issue Despite the known cost-effectiveness of patient-level smoking cessation treatments, a common finding from health services research is that many patients do not receive these treatments. Possible barriers to providing services that operate at levels beyond the control of individual HCPs include financial disincentives, organisational barriers and insufficient resources (Grimshaw, Eccles, & Tetroe, 2004). Many studies of physician and/or patient self-reported smoking cessation support suggest that opportunities are very often missed. Many published studies of smoking cessation practices report that rates of recording patients’ smoking status range widely from as low as 12% to as high as 80% (Flocke, Gordon, & Pomiecko, 2006; Goldstein et al., 2003; Grimshaw et al., 2004; Kottke et al., 1992; Millett, Gray, Saxena, Netuveli, & Majeed, 2007) and the reported delivery of brief advice and cessation support is typically very low. The rates at which HCPs provide the ABCs in New Zealand are not known. General practitioners tend to focus on trying to improve their effectiveness by improving efficacy rather than improving coverage. However, improving coverage may produce a higher return for effort (Glasgow, Lichtenstein, & Marcus, 2003). Spending one minute on prevention with more patients rather than more time on fewer patients may be the more effective strategy (Litt, 2005). Structure of report This review summarises evidence pertaining to the effectiveness of interventions aimed at increasing the likelihood of HCPs initiating elements of the ABC approach in the primary care setting. This report is divided into the following sections. The next section describes the review’s methods and includes the research questions, search strategy, inclusion and exclusion criteria, the data extraction, appraisal and synthesis methods, and the methodological limitations of the evidence review. The results section considers the included appraised studies, reporting first on the systematic reviews and meta-analyses, and then on the original primary research, grouped by intervention type. Study characteristics and findings are reported in separate tables and synthesised in the text, and the body of evidence is reported. The final section overviews and summarises results, briefly discusses key findings, the limitations of the evidence base and identified gaps in knowledge, and presents key conclusions. A glossary and detailed appendices follow, including a list of all the excluded papers annotated by reason for exclusion, and the completed data extraction tables for included papers.

6


7


Methods This review is perhaps different from the norm as it does not consider the effectiveness of interventions designed to increase the chances of quitting in individual smokers per se. Rather, the review considers the effectiveness of interventions delivered within primary care that aim to increase the likelihood of Health Care Professionals (HCPs) influencing and supporting smokers to make a successful quit attempt by asking all patients their smoking status, providing brief advice to smokers and by offering and providing cessation support to smokers willing to quit. Research questions The clinical question to be answered by this review was defined by staff from the Ministry of Health’s Tobacco Policy and Implementation Section working party, in conjunction with the reviewers. The primary research question to be addressed by this review was: What interventions have been successfully implemented in primary care and increase (a) assessment of smoking status (b) the provision of brief advice and (c) the offer of effective smoking cessation treatment (including counselling, pharmacotherapy, or referral for smoking cessation treatment), compared to standard care? For inclusion in the current review, the evidence had to fulfil the PICO criteria outlined in Table 1 and Table 2. These criteria were developed a priori and described in the scoping protocol prepared prior to commencement of the review proper. Table 1: PICO criteria for determining study eligibility Patient population

The population is all health care professionals (HCPs) delivering smoking cessation services in primary-care settings. Particular attention should be given to interventions that have been examined in Māori and Pacific specific settings or contexts.

Intervention Any intervention that aims to influence the documentation of smoking status, provision of brief advice to stop smoking, and offer or provision of, or referral for, smoking cessation treatment.

Comparator The interventions should be compared with ‘no-intervention’ or standard care (i.e. the usual level of care that would normally be provided or undertaken within the setting in the absence of an intervention aimed at increasing the incidence of HCPs initiating ABC).

Outcomes The measurable outcomes should be relevant to the New Zealand ABC strategy. That is: documentation of smoking status, provision of brief advice to stop smoking, and offer or provision of, or referral for, smoking cessation treatment. These outcomes may include: –The incidence of recording of smoking status in medical records ‘A’ –The incidence of patients reporting having been asked their smoking status during a primary care visit, via structured/exit interviews ‘A’ –The incidence of recording of delivery of GPs’ (or other HCPs’) smoking cessation Brief advice –The incidence of recording (or patient report) of referral to other cessation services ‘C’ –The incidence of recording (or patient report) of providing cessation counselling ‘C’ –The incidence of recording (patient report) of prescriptions for nicotine addiction medication ‘C’ – Patient reported receipt of (any) tobacco intervention at a primary care visit ‘ABC’ –The incidence of a positive response in a structured interview on the question of GPs’ (or other HCPs’) adoption/implementation of the New Zealand Smoking Cessation Guidelines (2007) or a component of the ‘ABC’ approach, or comparable smoking cessation activities, – Any combination of the above outcomes

It is important to note that studies not designed to answer the research question were deliberately excluded. The most common examples of such exclusions were studies where the population comprised patients not Health Care Professionals, and/or where the reported outcomes related to patient behaviour change (e.g. quitting smoking, or not) rather than

8


process outcomes, for example, counselling smokers in smoking cessation and supporting patients’ quit attempts. Table 2: Nature of the evidence Publication type Studies published in the English language, including primary (original) research published

as full original reports and secondary research (systematic reviews and meta-analyses) appearing in the published literature.

Study design Those that provide at least Level IV evidence according to the National Health and Medical Research Council (NHMRC) additional levels of evidence and grades (NHMRC, 2008)

Study duration No specified minimum study/intervention duration. Sample size At least 20 evaluable subjects per study arm (or exposed to both treatments). Literature search A systematic method of literature searching and selection was employed in the preparation of this review. Searches were limited to English language material published from 1990 onwards. The searches were completed on 24th, November, 2008. Therefore, studies published after this date were not eligible for inclusion in the systematic review. The following databases were searched: Bibliographic databases

� EMBASE + MEDLINE Platform: EMBASE.com (*) Review databases

� Cochrane Database of Systematic Reviews � Cochrane Central Register of Controlled Trials (CENTRAL) � Database of Abstracts of Reviews of Effectiveness (DARE) � Health Technology Assessment database (HTA) � NHS Economic Evaluation database (NHSEED) � Cochrane Methodology Register (CMR)

Platform: Wiley InterScience HTA Groups

� INAHTA website database: http://www.inahta.org/Search2/?pub=1 � MSAC: http://www.msac.gov.au/ � ANZHSN: http://www.horizonscanning.gov.au/ � NZHTA: http://nzhta.chmeds.ac.nz/ � NICE: http://www.nice.org.uk/ � AHRQ/USPSTF: http://www.ahrq.gov/ � CADTH: http://www.cadth.ca/ � SBU: http://www.sbu.se � KCE: http://kce.fgov.be

Clinical Practice Guidelines � National Guideline Clearing House database: http://www.guideline.gov/

* EMBASE.com combines EMBASE and MEDLINE content into a single, Web-accessible platform (http://www.embase.com), with duplicate citations removed. The single EMBASE.com platform means that EMBASE and MEDLINE can be searched simultaneously.

9


Literature search notes This is an area that is not well indexed and the following notes provide some background and context to the iterative search process. The indexing of health care professionals in primary care is somewhat inconsistent; however, an iterative search process was developed to capture as much of this literature as possible. There is not exact nor clear and consistent indexing for the ABC approach (as anticipated) or the 5As framework. Therefore, there was a reliance on phraseology for these principles potentially used in the titles or abstracts, to focus any retrieval. Broadly, search terms and phraseology included: primary care physician, counselling, consultation, health care delivery, preventative medicine, practice guidelines, health personnel attitudes and doctor-patient relations, smoker, smoking, smokers, tobacco, and subject approaches for diagnosis, prevention, treatment and drug therapy of smoking and/or tobacco dependence. Also, the search was not focussed on any specific pharmacotherapy (nicotine drug therapy and any of its derivative products including gum and patches, varenicline, bupropion, etc.), deeming these to be treatment rather than the interventions leading to treatment; or to any aspects of treatment (such as compliance or acceptance) or patient outcomes, deeming these to be outside the scope of the review question. The search terms, search strategy and number of citations identified are presented in Appendix B. The reference lists of included papers were subsequently scanned to identify any peer-reviewed evidence that may have been missed in the literature search. Hand searching of journals and contacting authors for unpublished research or additional information was undertaken in some instances in this review, but this was uncommon. Whilst grey literature and unpublished material such as conference abstracts were not included in the evidence review, they may be referred to in background sections. Assessment of study eligibility Studies were selected for appraisal using a two-stage process. First, titles and abstracts (where available) identified from the search strategy were scanned and excluded as appropriate. Second, the full text articles were retrieved for the remaining studies and selected for inclusion and appraisal in the review if they fulfilled the study selection criteria Table 3. Double-checking of the eligibility of studies by a second reviewer was undertaken as necessary. Citations were excluded for the following reasons:

10


Table 3: Exclusion criteria and coding Exclusion Code Reason E1 Inappropriate study design

Inappropriate study design: including non-systematic/narrative reviews, exploratory studies with no appropriate comparison groups, case reports, animal studies, short notes, letters, editorials, conference abstracts, and studies not deemed appropriate to the research question or nature of review.

E2 Incorrect population

Incorrect population group: populations belonging to health care systems that do not closely match New Zealand’s or another developed country’s health care systems in terms of organisational aspects and quality of care. Individual smokers (rather than health care professionals). Health care professionals delivering services in secondary/tertiary care.

E3 Incorrect intervention

Incorrect intervention or no intervention or wrong exposure variables. Does not include the correct intervention/s. For example, interventions aimed at increasing patient acceptance of cessation support rather than provision of ‘Ask’ – ‘Brief advice’ – ‘Cessation support’ (ABC) by health care professionals, in primary care contexts.

E4 Inappropriate comparator

Inappropriate comparator: does not include the correct comparator/s. Specifically, (1) studies that do not compare time periods (pre-interventions and post interventions) within the same population, (2) comparison groups not sampled on the same exposure or interventions/ situations (e.g., those who have received a specific intervention) versus those who have not, (3) comparison groups not sampled on relevant outcomes, (4) ecological studies with demographic or other non-modifiable system level variables that cannot be compared or extrapolated to individual levels.

E5 Inappropriate outcomes

Inappropriate outcomes: outcomes that are not congruent with the New Zealand Smoking Cessation Guidelines (2007). Patient- centred rather than provider-centred outcomes. For example, quit rates, relapse rates, patient satisfaction with services.

E6 Non-English language E7 Fewer than 20 patients/subjects E8 Published pre-1990 E9 Full-text not available from any source E10 Other or retrieved for background only There were 464 non-duplicate studies identified by the search strategy. As detailed in Table 4, 130 full text articles were eligible for retrieval after excluding studies from the search based on titles and abstracts. Of full papers retrieved, 88 did not fulfil the inclusion criteria. Therefore, 42 articles were fully appraised and are included in this report (listed in Appendix C). All excluded articles are presented in Appendix E , annotated by reason for exclusion based on the exclusion criteria detailed above. Reasons are presented hierarchically such that only the first reason that applied is reported. Other cited publications (e.g. those providing introductory or background material or relating to the review methodology) are presented in the References.

11


Table 4: Application of selection criteria to citations

Initial search = 464

Retrieved full text = 130 Inappropriate study = 29 Inappropriate population = 7 Inappropriate intervention = 8

Inappropriate comparator = 2

Inappropriate outcomes = 18

Included studies = 42

Non-English = 0

Second Pass Exclusion by full paper

Inappropriate study = 220

Inappropriate population = 80

Inappropriate intervention = 17

Inappropriate comparator = 0

Inappropriate outcomes = 14

Non-English = 0

Inappropriate sample size = 2 Published or data pre 1990 = 0 Full text not available = 1

First Pass Exclusion by title & abstract

- 334

- 88

Inappropriate sample size = 10 Published or data pre 1990 = 8

Full text not available = 0

Other = 0

Other = 6

12


Appraisal of included studies Dimensions of evidence The aim of this review was to find the highest quality evidence to answer the clinical question. In accordance with NHMRC guidance, the following dimensions of evidence were reviewed for each of the included studies (Table 5). It is important to recognise that the value of a piece of evidence is determined by all of these dimensions, not just the level of evidence. Table 5: NHMRC Dimensions of evidence Dimension Reviewers definition Strength of the evidence Level (Table 7 below)

The study design used, as an indication of the degree to which bias has been eliminated by the design alone. The levels reflect the effectiveness of the study design to answer the research question.

Quality

The methods used to minimise bias within an individual study (i.e., other than design per se)

Statistical precision An indication of the precision of the estimate of effect reflecting the degree of certainty about the existence of a true effect, as opposed to a effect due to chance (p-value or, alternatively, the confidence interval).

Size of effect Determines the magnitude of effect and whether this is of clinical importance. The distance of the study estimate from the ‘null’ value and the inclusion of only clinically important effects in the confidence interval.

Relevance of evidence Considers the relevance of the study to the specific research question and the context in which the information is likely to be applied, with regard to a) the nature of the intervention, b) the nature of the population and c) the definition of the outcomes.

The evidence was assessed according to the dimensions outlined in Table 5 above. Each study was also assigned a level of evidence in accordance with the NHMRC (2005) interim levels of evidence (Table 6). The highest level of evidence available is a systematic review of randomised controlled trials, which are considered the study type least subject to bias. Individual randomised controlled trials also represent good evidence. However, comparative observational studies such as cohort and case-control studies or non-comparative case series may often be more readily available. Such studies are often conducted early in the development of a technology, or to detect rare outcomes or outcomes which develop long after an exposure (e.g. cancer, cardiovascular disease). Nevertheless, these lower levels of evidence remain subject to considerable bias.

13


Table 6: NHMRC additional levels of evidence and grades (NHMRC 2008), by question type

Level Intervention 1

I 2 A systematic review of level II studies

II A randomised controlled trial III-1 A pseudo randomised controlled trial

(i.e. alternate allocation or some other method) III-2 A comparative study with concurrent controls:

▪ Non-randomised, experimental trial3

▪ Cohort study ▪ Case-control study ▪ Interrupted time series with a control group

III-3 A comparative study without concurrent controls: ▪ Historical control study ▪ Two or more single arm study4

▪ Interrupted time series without a parallel control group IV Case series with either post-test or pre-test/post-test outcomes

Explanatory notes

1 Definitions of these study designs are provided on pages 7-8 How to use the evidence: assessment and application of scientific evidence (NHMRC 2000b) and in the accompanying Glossary. 2 A systematic review will only be assigned a level of evidence as high as the studies it contains, excepting where those studies are of level II evidence. Systematic reviews of level II evidence provide more data than the individual studies and any meta-analyses will increase the precision of the overall results, reducing the likelihood that the results are affected by chance. Systematic reviews of lower level evidence present results of likely poor internal validity and thus are rated on the likelihood that the results have been affected by bias, rather than whether the systematic review itself is of good quality. Systematic review quality should be assessed separately. A systematic review should consist of at least two studies. In systematic reviews that include different study designs, the overall level of evidence should relate to each individual outcome/result, as different studies (and study designs) might contribute to each different outcome. 3 This also includes controlled before-and-after (pre-test/post-test) studies, as well as adjusted indirect comparisons (i.e. utilise A vs. B and B vs. C, to determine A vs. C with statistical adjustment for B). 4 Comparing single arm studies i.e. case series from two studies. This would also include unadjusted indirect comparisons (i.e. utilise A vs. B and B vs. C, to determine A vs. C but where there is no statistical adjustment for B). Note A: Assessment of comparative harms/safety should occur according to the hierarchy presented for each of the research questions, with the proviso that this assessment occurs within the context of the topic being assessed. Some harms (and other outcomes) are rare and cannot feasibly be captured within randomised controlled trials, in which case lower levels of evidence may be the only type of evidence that is practically achievable; physical harms and psychological harms may need to be addressed by different study designs; harms from diagnostic testing include the likelihood of false positive and false negative results; harms from screening include the likelihood of false alarm and false reassurance results. Note B: When a level of evidence is attributed in the text of a document, it should also be framed according to its corresponding research question e.g. level II intervention evidence, level IV diagnostic evidence, level III-2 prognostic evidence. Note C: Each individual study that is attributed a “level of evidence” should then be rigorously appraised using validated or commonly used checklists or appraisal tools to ensure that factors other than study design have not affected the validity of the results. Source: Hierarchies adapted and modified from: NHMRC 2008; NHMRC 1999; Bandolier 1999; Lijmer et al. 1999; Phillips et al. 2001.

14


Even within the levels of evidence stated above there is considerable variability in the quality of evidence. In accordance with NHMRC guidelines, it was necessary to consider the quality of each of the included studies. NHMRC quality checklists (1999) were employed to appraise included articles (Appendix C). The characteristics and quality of each included study were assessed using a number of quality criteria, as shown in Table 7, with studies rated as good, fair or poor quality. In addition, the AGREE tool (The AGREE Collaboration, 2003) was employed to appraise one clinical practice guideline. In this one instance, the AGREE tool short form was nested within the NHMRC based appraisal framework for consistency. Table 7: Quality criteria for different levels of evidence (NHMRC, 2000b) Study type Quality criteria Systematic review

Was a clinical question clearly defined? Was an adequate search strategy used? Were the inclusion criteria appropriate and applied in an unbiased way? Was a quality assessment of included studies undertaken? Were the characteristics and results of the individual studies appropriately summarised? Were the methods for pooling the data appropriate? Were sources of heterogeneity explored?

RCT

Was allocation to treatment groups concealed from those responsible for recruiting subjects? Was the study double-blinded? Were patient characteristics and demographics similar between treatment arms at baseline? Were all randomised participants included in the analysis? Were the statistical methods appropriate? Were any subgroup analyses carried out?

Cohort How were subjects selected for the ‘new’ intervention? How were subjects selected for the comparison or control group? Does the study adequately control for demographic characteristics, clinical features and other potential confounding variables in the study design or analysis? Was the measurement of outcomes unbiased (i.e., blinded to treatment group and comparable across groups)? Was follow-up long enough for outcomes to occur? Was follow-up complete and were there exclusions from analysis?

Case-control How were the cases defined and selected? How were the controls defined and selected? Does the study adequately control for demographic characteristics and important potential confounders in the study design or analysis? Was measurement of exposure to the factor of interest (e.g. the new intervention) adequate and kept blinded to case/control status? Were all selected subjects included in the analysis?

Adapted from NHMRC (2000) Data extraction Data was extracted onto standardised data extraction forms, and included information regarding study design, patient characteristics, details of the intervention, relevant outcomes, study quality, and relevant results. Data was extracted by two reviewers. Unless otherwise specified, the data that was most adjusted for confounders and/or multiple comparisons are reported. Furthermore, where subgroup analyses are available, these were reported if they were deemed relevant. Completed data extraction forms containing detailed information regarding study characteristics and quality, together with a brief summary of study results, can be found in Appendix G.

15


Data synthesis In addition to the level and quality of evidence of individual studies, the review considers the body of evidence in total. This involves consideration of the volume of evidence and its consistency. For systematic reviews with analyses involving evidence from RCTs, a meta-analysis should be performed when appropriate using the methodology of the Cochrane Collaboration (Mulrow & Oxman, 1997). However, this was not undertaken as the trial characteristics and subject characteristics were not sufficiently homogeneous to justify a meta-analysis. Quantitative pooling was similarly not possible for the other levels of evidence. Data from observational studies is subject to considerable heterogeneity and to biases that vary between studies. The review presents the statistical precision of the estimated effect size, together with a narrative synthesis of the results. Finally, the review will consider the relevance of the evidence, both with regard to the applicability of the population and the intervention as well as the relevance to the New Zealand health care setting. Limitations of the review methodology This review used a structured approach to review the literature. However, there were some inherent limitations to this approach. All types of study are subject to bias, with systematic reviews being subject to the same biases seen in the original studies they include, as well as biases specifically related to the systematic review process. Reporting biases are a particular problem related to systematic reviews and include publication bias, time-lag bias, multiple publication bias, language bias, and outcome reporting bias (Egger et al. 2001). A brief summary of the different types of reporting bias is shown in Table 8. Other biases can result if the methodology to be used in a review is not defined a priori (i.e., before the review commences). Detailed knowledge of studies performed in the area of interest may influence the eligibility criteria for inclusion of studies in the review and may therefore result in biased results. For example, studies with more positive results may be preferentially included in a review, thus biasing the results and overestimating treatment effect. Table 8: Reporting biases in systematic reviews Type of bias Definition and effect on results of review Publication bias The publication or non-publication of research findings.

Small, negative trials tend not to be published and this may lead to an overestimate of results of a review if only published studies are included.

Time-lag bias The rapid or delayed publication of research findings. Studies with positive results tend to be published sooner than studies with negative findings and hence results may be overestimated until the negative trials ‘catch up’.

Multiple publication bias

The multiple or singular publication of research findings. Studies with significant results tend to be published multiple times, which increases the chance of duplication of the same data and may bias the results of a review.

Citation bias The citation or non-citation of research. Citing of trials in publications is not objective, so retrieving studies using this method alone may result in biased results. Unsupported studies tend to be cited often which may also bias results.

Language bias The publication of research findings in a particular language. Significant results are more likely to be published in English, so a search limited to English-language journals may result in an overestimation of effect.

Outcome reporting bias

The selective reporting of some outcomes but not others. Outcomes with favourable findings may be reported more. For example, adverse events have been found to be reported more often in unpublished studies. This may result in more favourable results for published studies.

Adapted from Egger et al. (2001 ).

16


Some of these biases are potentially present in this review. Only data published in peer-reviewed journals is included. No attempt was made to include unpublished material, as such material typically has insufficient information upon which to base quality assessment, and it has not been subject to the scrutiny of the peer-review process. In addition, the search was limited to English-language publications only so language bias is a potential problem also. Outcome reporting bias and inclusion criteria bias are unlikely as the reviewers had no detailed knowledge of the topic literature, and the methodology used in the review and the scope of the review was defined a priori. The review scope was developed with the assistance of Ministry of Health staff to support policy and purchasing relevant to New Zealand. All studies included in this review were conducted outside New Zealand, and therefore, their generalisability to the New Zealand population and context may be limited and needs to be considered. This review was confined to an examination of the efficacy of the interventions and did not consider ethical or legal considerations associated with these interventions. Papers published pre-1990 were not considered as these tended to concern outdated practices. The studies were initially selected by examining the abstracts of these articles. Therefore, it is possible that some studies were inappropriately excluded prior to examination of the full text article. However, where detail was lacking ambiguous papers were retrieved as full text to minimise this possibility. Reasons for exclusion for every article included in the review are presented in Appendix E for transparency. Data extraction, critical appraisal, and report preparation were performed by both reviewers. For a detailed description of interventions and evaluation methods, and results used in the studies appraised, the reader is referred to the original papers cited.

17


Results Review of the research question, scope, overview of the study characteristics and results Review question What interventions have been successfully implemented in primary care that increase ‘A’ assessment of smoking status, ‘B’ the provision of brief advice, and ‘C’ the provision of cessation support (e.g. the offer of effective smoking cessation treatment including counselling, pharmacotherapy, or referral for smoking cessation treatment), compared to standard care? Scope Figure 3 below illustrates the scope of the review and makes clear the distinction between provider (or process outcomes) and patient outcomes. Figure 3 also illustrates how the New Zealand ABC approach aligns with another similar and widely used framework, for example, the 5As (Fiore, Bailey, Cohen, & et al., 1996). Broadly, process outcomes include the adoption and implementation of the New Zealand Smoking Cessation Guidelines (2007), or any recommended component of the ABC approach, or comparable smoking cessation frameworks. The scope specifically excludes patient outcomes, such as acceptance of treatment, quit rates, satisfaction with services received, or self-efficacy. A number of the included studies do nevertheless report patient outcomes alongside the provider outcomes of interest. While patient outcomes are outside the scope of this review, for completeness, quit rates (or abstinence) have been reported in some cases. However, aspects of the included trials that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further. Further details of patient outcomes are available in the original authors’ publications.

Figure 3 Review scope and comparison of the ABC approach with the 5As framework

Ask Brief advice

Cessation

Refer or Provide

Provider

Ask Advise

Assess

Assist

Arrange

The ABC approach Scope of the review: The effectiveness of interventions acting within this segment of the continuum

The 5As framework

Acceptance

Treatment

Cessation

Patient

18


Study characteristics: overview Of the 42 papers identified as eligible, four were systematic reviews, one meta-analysis was extracted from within an evidence-based clinical practice guideline (see other reviews), and 37 were original primary research studies. The five review articles are reported fully, commencing on page 25, and the original research articles follow, commencing on page 41. The 37 original studies differed considerably with respect to study design. Study designs included 14 cluster randomised controlled trials (C-RCTs) and 11 randomised controlled trials (RCTs), all were evidence level II. The remaining 12 studies were of other designs: two were non-randomised controlled clinical trials (evidence level III-2) (Goldstein, et al., 2003; Kottke, et al., 1992); six were controlled before and after trials (comparing outcomes of the same group before and after the implementation of the intervention or ‘within group analysis’: evidence level III-3) (DePue et al., 2002; Fiore et al., 1995; Flocke, et al., 2006; Krist et al., 2008; Ozer et al., 2001; Secker-Walker et al., 1992); three were time-series designs (Amundson, Solberg, Reed, Martini, & Carlson, 2003; Coleman, Lewis, Hubbard, & Smith, 2007; Millett, et al., 2007) (evidence level III-2); and one study (Tremblay et al., 2001) analysed comparative data from two temporally different cross-sectional surveys (evidence level IV). In this case, the absence of a control group precludes attribution of the observed changes to the programme as there may have been other social/environmental changes that took place during the study period that may have influenced HCPs and/or patient behaviours, including the initiation of funding for NRT, mass media campaigns (possibly leading to patients prompting HCPs), and/or law changes to restrict tobacco sales or to limit smoking in public places. While the possible influence of such confounders is expected when employing these study designs, and is not indicative of study quality per se, it is difficult to know how these factors might have influenced the results. Of particular note is the common utilisation of the cluster randomised controlled trial (C-RCT) in effectiveness studies of interventions aimed at changing health professionals’, rather than patients’, behaviours. In clinic settings, if the unit of randomisation is the patient, then there is a high chance of contamination between the intervention and control conditions due to HCPs within a site or clinic inadvertently applying the intervention universally. To counter this possible contamination effect (and possible type two errors), the C-RCT design was used in 14 studies, whereby the unit of randomisation was the clinic or practice (or other higher level unit) not the patient. The 37 original studies also differed widely with respect to study duration. The mean study duration of the randomised trials was 14 months; however study durations ranged widely from three or four months (Cockburn et al., 1992; McRobbie, Hajek, Feder, & Eldridge, 2008) to four years (Ockene et al., 1997; Piper et al., 2003). For non-randomised trials, the mean duration was 30 months with the longest running trial being a 15 year interrupted time series evaluation of contractual financial incentives in the U.K. (Coleman, et al., 2007). Overall, the studies involved more than 3000 health professionals (the majority were general practitioners working in primary care/community settings) including nurse practitioners, physician assistants, and other rooming and administrative staff. The number of patients involved in the studies ranged between a few hundred and 1.6M (electronic patient records), with the majority involving adult patients. Two studies involved adolescents (Ozer, et al., 2001; Pbert et al., 2006), one study involved pregnant women (Secker-Walker, et al., 1992), another study involved patients with diabetes (Millett, et al., 2007), and another study involved patients with established coronary heart disease (Moher et al., 2001).

19


Quality Overall, the methodological quality of the original research studies was fair (NHMRC quality assessment tool, Appendix F). The majority of the studies used a randomised design and while many described the method of randomisation, others did not. Some studies provided data on withdrawals and drop-outs, but not all. In general, studies provided information on their analyses but many did not report on the intention to treat analysis or on confounding factors, although most provided a good level of detail when describing the analyses. Due to the nature of the interventions and the challenges of conducting (pragmatic) randomised trials in clinic settings, the majority of the studies were not blinded or it was not clear who was blinded to the study or question. However, most used blinded research assistants for chart abstraction. One innovative study (Young, D'Este, & Ward, 2002) did achieve blinding of the participants (family physicians) by simultaneously applying an intervention identical in format and intensity to the control group, but about cervical screening rather than smoking. Many of the studies used subjective measures for assessing outcomes, most commonly, patient exit interviews. These are likely to be subject to some degree of recall bias. Tools for measuring the primary outcomes were usually described adequately, but often the reliability and validity were not cited or demonstrated. Population In all of the included studies, the population was Health Care Professionals in primary care settings. HCPs were studied as individuals, in groups, or within an entire clinic or practice. Such groups comprised various different combinations of primary healthcare physicians (the majority were general practitioners), practice-based nurses, other health care professionals, and administrative staff. In terms of the study population, it is important to distinguish between ‘participants’ and ‘patients’. Participants, in this case, are those HCPs on which the intervention is targeted and those whose clinical behaviour is intended to be influenced. Patients provide both the opportunity for the HCP to intervene, and act as part of the assessment process, but patients’ smoking behaviours per se are not the focus of this review. Patients, via an exit survey for example, provide a way to measure and record practitioners provision or cessation activities (e.g. brief advice or counselling). Patients, more often than not, were ‘passive’ and not formally enrolled in a study (although consent to complete interviews and questionnaires was usual). For example, retrospective chart audits of the documentation of patient’s smoking status do not require patient participation per se. Generally, the patient group was ‘all patients’. Interventions Both original papers and systematic reviews collectively evaluated the effectiveness of a range of interventions that were aimed to increase the delivery of effective smoking cessation treatments in primary care settings. The identified intervention types or themes are listed and described briefly in Table 9 and Table 10. Table 10 lists and describes the intervention types as they are grouped and summarised in this review. Table 11 lists seven other intervention types that were also identified in the literature as described in the style of the original authors. Broadly, six of the seven intervention types are sub-categories of Training and are included and noted as training. For the other intervention type, Patient Mediated Interventions, no eligible studies were identified. Few studies reported any theoretical underpinnings for the programme. However, there were exceptions as Lennox, et al. (1998) described the Stages of Change model, Solberg et al. (2000) Continuous Quality Improvement, Kottke, et al. (1992) the Social Learning Model of Consultation and Tremblay et al. (2001) reported on the use of the Precede-Proceed model of programme development. Other authors made references to the 4As and 5As frameworks, but not necessarily the intervention’s theoretical basis.

20


There was notable heterogeneity in the way in which authors described similar interventions, and in the terminology and claimed theoretical underpinnings for some of the interventions. However, some commonality and overlap was also evident within categories and between categories with respect to the essence of the intervention. Overall, the groupings reported here were formed in an attempt to best represent the total body of evidence in a way that is simplified, but importantly, potentially generalisable to New Zealand’s primary care context. Table 9: Included intervention types Intervention Definition †Training Training interventions were interventions in which the principal component was imparting

or transferring new knowledge, skills, and competencies to HCPs, to enable them to provide better care for their smoking patients, by improving the delivery of smoking cessation activities within the practice.

Multi-component Any intervention that includes two or more intervention components. Reminders Any intervention (manual or computerised) that prompts the health care provider to

perform a clinical action. Examples include concurrent or inter-visit reminders to professionals about desired actions such as screening or other preventive services, enhanced laboratory reports or administrative support (e.g., follow-up appointment systems or stickers on charts).

Financial incentives

Financial incentives include any ‘pay-for-performance’ schemes or financial incentives offered for reaching preset clinical performance targets.

Audit & feedback Any summary of clinical performance or health care over a specified period (with or without recommendations for clinical action), when this data is fed back to clinicians, either on an individual or group basis, and additionally, comparisons with nominal or target performance may provided. The information may have been obtained from medical records, computerised databases, from patients, or by observation.

Other Other intervention types. † Including academic detailing, outreach visits, educational facilitators, workshops

21


Table 10: Other intervention types ‡Intervention Definition Academic detailing

Academic detailing is a form of educational outreach that features personal educational visits to clinicians in their own practice setting. Academic detailing provides an opportunity for the educator, or “office practice consultant”, to assess the needs and motivation of the targeted clinician and subsequently tailor the educational intervention to the particular needs, barriers, and motivational readiness of the targeted audience and the specific practice environment (Soumerai, McLaughlin, & Avorn, 1989). This type of face-to-face visit has also been referred to as university-based educational detailing, academic detailing, and educational visiting.

Outreach visits Use of a trained person who meets with providers in their practice settings to provide information. The information given may include feedback on the provider's performance.

Use of local opinion leader

Use of providers explicitly nominated by their colleagues to be "educationally influential”.

Conference Participation of health care providers in conferences, lectures, workshops or traineeships outside their practice settings.

Local consensus process

Inclusion of participating providers in discussion to ensure agreement that the chosen clinical problem is important and the approach to managing it appropriate.

Patient-mediated interventions

Any intervention aimed at changing the performance of health care providers for which information was sought from or given directly to patients by others (e.g., direct mailings to patients, patient counselling delivered by others, or clinical information collected directly from patients and given to the provider).

Educational materials

Distribution of published or printed recommendations for clinical care, including clinical practice guidelines, audiovisual materials and electronic publications.

‡Other intervention types as described in the style of the original authors including interventions identified within general prevention trials or systematic reviews of a mix of smoking specific and other trials, but not necessarily reporting smoking specific outcomes. Collated from: Grimshaw et al. (2004), Oxman et al. (1995) and Soumerai (1989). Comparisons Many of the studies did not clearly describe the comparison used. In the main, the interventions were compared with ‘no-intervention’ or ‘usual care’, but this was often not explicit. In this review, ‘usual care’ is defined as “the usual level of care that would normally be provided or undertaken within the particular setting, in the absence of an intervention aimed specifically at increasing the incidence of HCPs initiating smoking cessation activities in routine practice”. In clinic settings, ‘usual care’ may be variable within and between clinics and it is possible for some contamination of the control groups to occur (for example, staff rotation within practices resulting in HCPs inadvertently delivering the intervention to control group patients), however, this would result in an underestimation of any effect rather than an overestimation of the effectiveness of the programme. Outcomes The identified studies often measured a range of outcomes, using various different tools. Broadly, the outcomes were the performance of any/all elements of the ABC approach in primary care setting – the documentation of smoking status, provision of brief advice to patients to encourage them to stop smoking, and the offer or delivery of cessation support. The majority assessed the performance of health care providers from patient exit interviews and/or chart review. Others reported data from physicians’ mail surveys and telephone interviews designed to assess physicians’ perceptions of their own effectiveness, with regard to giving quitting advice or cessation support, and in some cases, the program’s effectiveness and feasibility (qualitative process evaluation). For instance, in one study (Wadland et al., 2007), clinicians completed survey forms that assessed their characteristics, baseline patient exit surveys assessed clinicians’ use of the 5As framework, and clinicians’ referrals to a Quit-Line were tracked by the Quit-Line staff . Similarly, Young et al. (2002) collected data from patient

22


recall of family physician advice, medical record audit, and from family physician self-report. Studies often did not measure all of the ABC outcomes, and only 24 of 37 (64%) original research studies included any measure of cessation support ‘C’. Context The included studies were conducted in six different countries with the United States being clearly the most represented (n = 26 studies), followed by the United Kingdom (n = 7 studies), Canada (n = 4 studies), Australia (n = 3 studies) and the Netherlands and Scotland both (n = 1 study). Finally, while the population (HCPs) and setting (primary care) were similar in all of the studies, two variations of context were apparent: smoking specific studies (32 studies) and general prevention studies (10 studies). Smoking specific studies were those studies in which the intervention and outcomes (process, patient, or both) related only to HCPs’ smoking cessation activities. General prevention studies were those studies in which two or more preventive manoeuvres were addressed within one trial. For example, the effectiveness of Training was evaluated in the context of a general prevention study reporting outcomes relating to smoking, diet and exercise counselling, and influenza vaccination (Borgiel & Williams, 1999). When relevant smoking outcomes were reported separately and the authors’ discussion and conclusions differentiated between the different outcomes, then the study was included (Amundson, et al., 2003; Dietrich et al., 1992; Flocke, et al., 2006; Grimshaw, Thomas, et al., 2004; Hogg et al., 2008; Lemelin, Hogg, & Baskerville, 2001; McPhee, Bird, Fordham, Rodnick, & Osborn, 1991; Moher, et al., 2001; Solberg, et al., 2000). If the smoking data could not be extracted separately (or the study did not include any relevant smoking outcomes), then the study was excluded. Further, when considering the eligibility of systematic reviews that included an amalgamation of smoking specific, general prevention including smoking, and other trials, such reviews were only included if meaningful smoking specific data were reported. Results overview Results of this review are presented in the following section, and the cube in Figure 4 below illustrates the interventions, outcomes and contexts described.

Figure 4 The ‘cube’ depicts the identified interventions, outcomes and contexts

23


Methodological information and results extracted from the included studies follow, first for systematic reviews, meta-analyses and other-reviews, then for the 37 original research studies. The original included studies are grouped and presented by intervention type, in the order shown in Figure 4. More detailed information is available in the data extraction tables in Appendix G, or alternatively, the reader should refer to the authors’ original paper. Given the necessity to focus on the specific review question, only data relevant to the current review are presented.

24


Systematic reviews: introduction The literature search identified four relevant systematic reviews and one other review type, a meta-analysis by Fiore et al. (2008). The characteristics of the review articles are described in Table 11 and Table 12. The reviews differed considerably with respect to their inclusion criteria, in particular, the different types of study designs included for review, the publication year range, and the process and/or patient outcomes reported. One review included only randomised controlled trials or cluster randomised controlled trials (Lancaster & Fowler, 2000), whereas the other four reviews also included non-randomised controlled clinical trials, controlled before and after trials, and interrupted time-series designs (Anderson & Jané-Llopis, 2004; Dickey, Gemson, & Carney, 1999; Fiore et al., 2008; Grimshaw, Thomas, et al., 2004). The reviews also differed with respect to the interventions; one review (Lancaster & Fowler, 2000) evaluated the effectiveness of a single intervention (training), whereas the other four evaluated the effectiveness of up to nine different types of interventions, alone and in combination. The systematic reviews are presented in this section ordered by level of evidence and then by publication date. The one systematic review that contained only randomised trials is presented first. The other three reviews all contained both randomised trials and other non-randomised designs and were considered level I / III-3– the latter reflecting the same level as the non-randomised trials that were included within the review.

25


Table 11: Systematic review characteristics: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings

Level I evidence Citation Study Population Intervention Comparator Outcomes Lancaster et al. (2000) U.K.

Systematic review of 10 randomised controlled trials (nine included studies from U.S.A. and one from Canada) Quality: Good

HCPs or Practices Included trials in which the unit of randomisation was a healthcare practitioner or practice.

Training Interventions in which HCPs were trained in methods to promote smoking cessation among their patients.

Usual care Rates of performance of tasks of smoking cessation by health care professionals including offering counselling, setting quit dates, giving follow-up appointments, distributing self-help materials, and recommending nicotine gum ‘ABC’.

∆Educational materials Distribution of published or printed recommendations for clinical care, including clinical practice guidelines, audiovisual materials and electronic publications. ∆Education Educational meetings: healthcare providers who have participated in conferences, lectures, workshops or traineeships. Local opinion leaders Use of providers nominated by their colleagues as ‘educationally influential’.

Grimshaw et al. (2004) U.K.

Systematic review of 235 trials reporting 309 comparisons including cluster randomised controlled trials, randomised controlled trials, controlled clinical trials, controlled before and after trials, and interrupted time-series designs. Quality: Good

HCPs or Practices Included trials in which the unit of randomisation was a healthcare practitioner or practice.

∆Educational outreach visits Use of a trained person who met with providers in their practice settings to give information with the intent of changing the provider’s practice.

Usual care Dichotomous process variables, including asking and/or documenting smoking status ‘A’, providing and/or documenting the provision of brief advise ‘B’, and the provision and/or documentation of any component of cessation support ‘C’. Note: some studies also reported quit rates ‘Q’.

∆ = Evaluated within smoking trials

26


Table 11: Systematic review characteristics: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

Citation Study Population Intervention Comparator Outcomes Patient-mediated

interventions New clinical information (not previously available) collected directly from patients and given to the provider, e.g. depression scores from an instrument.

Local consensus processes Inclusion of participating providers in discussion to ensure that they agreed that the chosen clinical problem was important and the approach to managing the problem was appropriate.

∆Audit & Feedback Any summary of clinical performance of healthcare over a specified period.

∆Reminders Patient- or encounter-specific information, provided verbally, on paper or on a computer screen, which is designed or intended to prompt a health professional to recall information.

Marketing Use of personal interviewing, group discussion (‘focus groups’), or a survey of targeted providers to identify barriers to change and subsequent design of an intervention that addresses identified barriers.

∆ = Evaluated within smoking trials

27



Citation Study Population Intervention Comparator Outcomes Education Educational interventions (12 trials) included the distribution of educational materials, educational meetings or outreach visits, and the provision of audit and feedback (to both practitioners in training and/or established practitioners). Office systems Practice-based interventions (four trials) included revision of professional roles, changes in the medical records systems, and the use of quality monitoring systems. Educational facilitator Outreach visits (five trials): when there was a personal visit by a trained person to health practitioners in their work setting.

Anderson et al. (2004) The Netherlands

Systematic review/meta-analysis of 19 trials including randomised or controlled clinical trials, controlled before and after trials and interrupted time-series designs Quality: Good

HCPs or Practices Included trials in which the unit of randomisation was a healthcare practitioner or practice

Multi-Component Multi-component interventions (ten trials): when programmes included more than one type of intervention component within an educational or practice-based intervention, or when combined at least one educational and one practice-based intervention category.

Usual care Measures of practitioners’ behaviour. Screening rates ‘A’ Advice-giving to smokers ‘B’ Also, bio-chemically verified patient quit rates ‘Q’.

28



Citation Study Population Intervention Comparator Outcomes Office tools Health risk assessment questionnaires and appraisals, prompting and reminding (chart stickers, checklists, flow charts, reminder letters), and education (manuals and handbooks).

Dicky et al. (1999) U.S.A.

Systematic review of 33 trials, including one meta-analysis, 13 randomised controlled trials, five controlled clinical trials, ten time-series designs, and four controlled before and after trials Quality: Good

Primary healthcare physicians, practice-based nurses, other health care professionals, and administrative staff.

Teamwork The coordination and delegation of tasks between all staff (clinical and administrative).

Usual care Counselling rates for smoking cessation as reported by patient survey and/or chart review in 13 ‘smoking specific’ trials (including one meta-analysis), and in 12 trials, counselling rates for smoking cessation reported within ‘general prevention’ programmes, and in eight trials, ‘other’ outcomes.

Table 12: Other review characteristics: The effectiveness of interventions

to increase the delivery of effective smoking cessation treatments in primary care settings

Citation Study Population Intervention Comparator Outcomes Training Training included didactic presentation of material, group discussions, and role playing; from formal training during residency to on-site clinician training within the community.

Fiore et al. (2008) U.S.A.

Clinical Practice Guideline Good

Randomised controlled trials. Cluster randomised controlled trials.

Reminders Reminders included any charting system, such as chart reminder stickers or treatment algorithms attached to the chart.

Usual care Outcomes reported based on the “5As” framework [with the New Zealand ABC approach indicated in brackets] – Ask [Ask ‘A’], Advise to quit [Brief Advice ‘B’], Assess willingness to quit, Assist with quitting, Arrange follow-up [Cessation support ‘C’].

29


Systematic reviews and meta-analyses: results Lancaster and Fowler (2000) up dated and replaced their earlier systematic review (Silagy, Lancaster, Gray, & Fowler, 1994), excluding one paper and adding three additional post – 1990 trials. The aim of the review was to assess the effectiveness of interventions that train healthcare professionals (HCPs) in methods for improving the quality of care delivered to their patients who smoke. Lancaster and Fowler (2000) proposed that training HCPs is more effective than no training in increasing the number of smokers who are offered advice about stopping and who subsequently achieve abstinence, and that the effect of training can be enhanced by either providing prompts and reminders to HCPs or encouraging them to offer nicotine replacement therapy as an adjunct to their brief advice. Lancaster and Fowler (2000) systematically searched the literature and subsequently included ten cluster randomised trials. Nine of ten trials reported process outcomes, and all trials reported on patient smoking behaviour as well. The process outcomes included the number of smokers who were counselled, asked to set a date for stopping (quit date), given a follow up appointment, given self help materials, offered nicotine gum, or prescribed a quit date. In all the trials, the unit of randomisation was a healthcare practitioner or practice. Eight of the trials examined the effect of training compared to untrained control groups. In three trials, all the health care professionals received training, and the comparisons examined the effect of using prompts and nicotine gum, separately and combined, as enhancements to training. Almost all the training was provided on a group basis in a tutorial or workshop setting and the duration of training ranged from a single one hour session to a whole day. The participation rates among those invited ranged from less than 10% in private practice settings to 90% when the participants were members of a residency training programme. Lancaster and Fowler (2000) noted the difficulties in pooling cluster level data and decided on a narrative approach to synthesising the results, rather than a formal meta-analysis. Overall, Lancaster and Fowler (2000) found that training increased the smoking cessation activities of HCPs in almost all the studies. Trained professionals were about 1.5 to 2.5 times more likely to perform tasks of smoking cessation than untrained controls, and the effects of training on process outcomes increased if prompts and reminders were used. However, of eight studies that compared patient smoking behaviour between trained professionals and controls, six found no effect of intervention. Thus, there was no strong evidence that training HCPs changed smoking behaviour. In light of this finding (that training improves process but not patient outcomes), Lancaster and Fowler (2000) warn that simply providing training programmes for HCPs, without addressing organisational and contextual factors, is unlikely to be a wise use of resources. In perhaps the most comprehensive review of the literature relating to guidelines implementation strategies to date, Grimshaw et al. (2004) report on 235 trials of interventions aimed at changing practitioners’ behaviour and/or clinic performance, regarding the provision of preventive services in primary care settings. The review was conducted in the U.K. using the Cochrane Effective Practice and Organisation of Care (EPOC) group methodology. The systematic searches identified over 5000 potentially relevant papers, of which, 235 reporting 309 comparisons were included in the final analysis. Included studies comprised cluster randomised controlled trials, randomised controlled trials, controlled clinical trials, controlled before and after trials, and interrupted time series trials that reported measures of practitioners’ behaviour and/or clinic performance. Of these 309 comparisons, 73% evaluated multi-component interventions but only eight papers reported smoking related outcomes, one of those reported smoking related outcomes from a ‘general prevention’ trial. Thirty-six studies aimed to change one prevention behaviour and the

30


remainder targeted several behaviours (to a maximum of six behaviours– e.g. prescribing behaviours, diet, exercise, or smoking cessation counselling). The study population always involved GPs/physicians (or groups of clinicians), and in 14 studies, other healthcare professionals were also involved including nurses and/or administrative staff. Grimshaw et al. (2004) used the classification of professional interventions from the EPOC taxonomy, to group all the identified interventions into 11 broad categories or types based on the intervention component/s that appeared most predominant, and with consideration given to the different terminology often used to describe essentially the same intervention (e.g. academic detailing and educational outreach visits). The 11 broad intervention types identified were (in alphabetical order): audit and feedback, distribution of educational materials, educational meetings, educational outreach visits, local consensus processes, local opinion leaders, marketing, mass media, patient-mediated interventions, reminders, plus multifaceted interventions. The identified interventions had been applied in multiple contexts, all with the common aim of increasing the performance of HCPs in the provision of a range of preventive services to their patients, including smoking cessation advice and cessation support (quit rates and other patient level outcomes were also reported for some studies). Only five of the identified intervention types (audit and feedback, distribution of educational materials, educational outreach visits, multifaceted interventions, and reminders) were represented by smoking specific or general prevention/smoking trials. Drawing from within the overall results, the effectiveness of the interventions implemented in smoking specific or general prevention/smoking trials were reported as follows– Generally effective: reminders, the most frequently evaluated single intervention, moderate effect sizes (i.e. the median absolute difference across post-intervention measures is >10 and ≤ 20%); educational outreach, generally more effective than educational materials, modest effect sizes (> 5% and ≤ 10%); multifaceted interventions, modest effect sizes, however, there appeared to be no evidence of an expected dose-response curve associated with increasing the number of components in an intervention. Mixed effects: audit and feedback, modest effect sizes, although this finding was less robust as only two of the included smoking studies used audit and feedback as an intervention component. Generally ineffective: educational materials, modest effect sizes, while education materials are often important, as a stand-alone intervention, their effects may be short lived. There were no trials of financial incentives implemented as a single intervention. Financial incentives were included in 16 multi-component trials, but none reported smoking outcomes. Grimshaw et al. (2004) concluded that overall, the majority of comparisons reporting dichotomous process data observed modest to moderate improvements in care. However, there was considerable variation in the observed effects both within and across interventions (and the number of included smoking trials was relatively low). Further, the authors point to the imperfect evidence base, in particular the lack of evidence about which strategies are likely to be efficient under different circumstances, and hence the need for decision makers to use considerable judgement about the best use of limited resources. Anderson and Jané-Llopis (2004) systematically searched and reviewed studies that evaluated the effectiveness of educational and practice base strategies to increase the involvement of primary health-care practitioners in the treatment of tobacco dependence. Randomised or controlled clinical trials, controlled before and after trials, and interrupted time-series designs that reported measures of practitioners’ behaviour were appraised. Anderson and Jané-Llopis (2004) conducted a meta-analysis, using a random effects model, of 24 programmes identified in 19 trials. Some studies also reported bio-chemically verified patient quit rates. The included

31


studies focused on the engagement of health-care practitioners’ behaviour in dealing with tobacco or dealing with broader prevention activities that included tobacco, directed to all patients 18 years and older and published in English or Spanish between 1966 and 2001. In 17 trials, the participants were primary healthcare physicians (including five trials within physician vocational or residency training contexts) and in two trials the participants were practice-based nurses. All of the programmes were undertaken in high-income countries (Australia, Canada, the U.K., eight of 19 in the U.S.A, and one in Israel). Interventions evaluated included: educational interventions (12 trials), practice-based interventions (four trials), outreach visits (five trials), and multi-component interventions (ten trials). Educational interventions included the distribution of educational materials, educational meetings or outreach visits, and the provision of audit and feedback (to both practitioners in training and/or established practitioners). Practice-based interventions included revision of professional roles, changes in the medical records systems, and the use of quality monitoring systems. Outreach visits were defined as a personal visit by a trained person to health practitioners in their work setting. Finally, multi-component interventions included more than one type of intervention component within an educational or practice-based intervention, or combined at least one educational and one practice-based intervention category. Anderson and Jané-Llopis (2004) reported that overall, the programmes led to a 15% (95% CI = 7–22) increase in practitioners’ screening rates (mean control rates of 48%); a 13% (95% CI = 9–18) increase in practitioners’ advice-giving rates (mean control rates of 51%), and a 4.7% (95% CI = 2.5–6.9) increase in bio-chemically verified patient quit rates (mean control rates of 16.9%). The authors also reported significant and potentially important differences between the intervention effects on established practitioners compared to practitioners in training. For established practitioners, programmes were effective in changing their screening and advice-giving rates, but not their patients’ quit rates. For practitioners in training, programmes were found to be effective in changing their advice-giving rates and their patients’ quit rates but not their screening rates. The authors suggest that the behaviours of practitioners in training might be more malleable than those of established practitioners. Further, that the provision of educational interventions for practitioners in training in combination with systematic outreach practice based support for established practitioners is likely to be an effective strategy to increase the delivery of smoking cessation treatments in primary care settings. Dicky, Gemson and Carney (1999) systematically searched the literature and reviewed studies investigating the effectiveness of ‘office tools’ and ‘teamwork’ to increase the involvement of primary health-care practitioners in the treatment of tobacco dependence. Office tools included health risk assessment questionnaires and appraisals, prompting and reminding (chart stickers, checklists, flow charts, reminder letters), and education (manuals and handbooks). Teamwork included the coordination and delegation of tasks between clinical and administrative staff. Tasks described within the studies included arranging patient appointments, disseminating and reviewing behavioural risk questionnaires, applying stickers and reminders to patient charts, distributing educational materials, and performing follow-up mail and telephone contacts. The participants were primary healthcare physicians, practice-based nurses, other health care professionals, and administrative staff. The outcome of interest was smoking cessation counselling, as reported by patient survey and/or chart review. However, in an undetermined number of cases, ‘smoking cessation counselling’ may only have amounted to ‘brief advice’. A number of the included studies also reported on provider counselling at the patient level (i.e. effectiveness); however, these outcomes, while considered important, were outside the scope of the review. Counselling rates for smoking cessation were reported for 13 smoking specific trials (including one meta-analysis) and within 12 trials of general prevention programmes. Of the 13 smoking

32


specific studies, 92% reported a statistically significant increase in the provision of smoking cessation counselling (and/or brief advice). Of 12 ‘general prevention’ trials, 58% reported a statistically significant increase in the provision of smoking cessation counselling (and/or brief advice). The authors provided a narrative synthesis of the data and on a case-by-case basis, the necessary caveats (due to the diverse methodology, lack of randomisation and modest scope evident in the included trials). Dicky et al. (1999) reported consistent evidence from a number of studies regarding the effectiveness of both office tools and teamwork for improving the rate of primary care provider behaviour-change counselling. Dicky et al. (1999) concluded that generally, multi-component programmes appeared more effective than single-component programmes, and programmes implementing office tools and teamwork were more effective than office tools or teamwork alone. The common strategy of chart prompts (in particular, electronic tools over manual systems) has been shown to be related to improved counselling rates. The most effective type of teamwork (and/or tools) probably varies according to provider, staff, patient, and practice characteristics; however, sharing the responsibility for office-based smoking cessation activities among all staff appears to be a promising strategy. Other reviews An extensively updated clinical practice guideline released in May 2008 by the Agency for Healthcare Research and Quality (AHRQ) (Fiore, et al., 2008) identified new counselling and medication treatments that are effective for helping people quit smoking. In addition, the review was broader in scope than the original 1996 and 2000 reviews, and introduced new information pertaining to the effectiveness of interventions aimed at changing practitioners’ behaviour in relation to providing smoking cessation interventions in primary care settings. This expanding of the scope resulted in the 24-member guidelines panel (chaired by Michael C. Fiore) assessing in detail more than 300 new articles identified for possible inclusion in a meta-analysis, and an additional 600 articles were examined but were not supported by meta-analyses. Taken as a whole, the 2008 Guideline is based on three literature reviews and a total of 8,700 research articles published between 1975 and 2007. Five of the 300 new articles report process outcomes (e.g. ask, brief advice, cessation support) and the HCP or clinic was the unit of randomisation. Specifically, topic eleven (relevant to this review) was stated as (the) “Effectiveness of systems interventions, including provider training and the combination of training and systems interventions”. The programmes led to absolute increases of between 9% and 30% across a range of ‘ABC’ outcomes. The guidelines Panel concluded that health care delivery administrators, insurers, and purchasers can promote the treatment of tobacco dependence through a systems approach and should, therefore, make tobacco assessment and coverage of treatment a contractual obligation of the health care insurers and/or clinicians who provide services to them. Various institutional policies were recommended to facilitate these interventions, including implementing a tobacco user identification system in every clinic, providing adequate training, resources, and feedback to ensure that providers consistently deliver effective treatments, dedicating staff to provide tobacco dependence treatments, and assessing the delivery of this treatment in staff performance evaluations. Systematic reviews: summary Summary data from the five reviews of the effectiveness of interventions for increasing the delivery of smoking cessation treatments in primary care settings are summarised in Table 13 & Table 14. Collectively, these five reviews summarise the results from well over 200 trials reporting on the effectiveness of interventions aimed at changing practitioners’ behaviour and/or their delivery of preventive services. Of these, 56 non-duplicate trials (of 75) related specifically to the delivery of smoking cessation activities in primary care settings. Within this

33


scope, the reviews investigated the effects of training, audit and feedback, distribution of educational materials, educational outreach visits, multifaceted interventions, office tools and teamwork as they related to improving the quality of care that HCPs delivered to their patients who smoke. Lancaster and Fowler (2000) reported that trained health care professionals were about 1.5 to 2.5 times more likely to perform tasks of smoking cessation support than untrained controls and the effects of training on process outcomes increased if prompts and reminders were used. Grimshaw et al. (2004) categorised reminders, educational outreach, and multifaceted interventions as generally effective; audit and feedback as demonstrating mixed effects; and educational materials (as a stand-alone intervention) as generally ineffective. Financial incentives were included in 16 multi-faceted trials, but none reported smoking outcomes. Anderson and Jané-Llopis (2004) found that educational and practice-based strategies increased the involvement of primary health-care practitioners in the treatment of tobacco dependence. Overall, the programmes led to a 15% increase in practitioners’ screening rates, a 13% increase in practitioners’ advice-giving rates, and a 4.7% increase in bio-chemically verified patient quit rates. Dicky, Gemson and Carney (1999) reported that of 13 smoking specific studies of office tools and teamwork, 92% reported a statistically significant increase in the provision of smoking cessation counselling, and of 12 general prevention studies, 58% reported similar results. Dicky et al. (1999) concluded that generally, multi-component programmes appeared more effective than single-component programmes, and programmes implementing office tools and teamwork were more effective than office tools or teamwork alone. In particular, the common strategy of chart prompts (more so, electronic over manual systems) was shown to be related to improved counselling rates. Fiore, et al. (2008) found that the programmes studied led to increases of between 9% and 30% across a range of ABC outcomes. Systematic reviews: conclusions Taken together, these five reviews provide extensive information from an enormous range of interventions, comparisons, outcomes, settings, contexts and study designs derived from upwards of 5000 potentially relevant papers (this number includes general prevention and non-smoking trials). Overall, at least for smoking specific process outcomes, the results were positive. However, a number of studies found no corresponding effect of the intervention on patients’ smoking behaviours. A common theme in the literature is that simply providing training and/or other multi-component programmes for HCPs, without addressing policy level, organisational and other contextual factors, is unlikely to be a wise use of health care resources. The evidence suggests that in order to maximise the potential of such programmes, a ‘comprehensive package’ needs to be delivered.

34


Table 13: Systematic review results: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings

Lancaster and Fowler (2000). Systematic review including ten randomised controlled trials. To assess the effectiveness of interventions that train healthcare professionals in methods for improving the quality of care delivered to patients who smoke. Results summary Cessation support: Trained healthcare professionals (HCPs) were about 1.5 to 2.5 times more likely to counsel patients about smoking than un-trained HCPs, and to initiate other interventions such as setting a quit date, suggesting a follow-up appointment, and offering self-help materials or nicotine gum. Prompts and reminders to HCPs to deploy smoking cessation techniques were used in addition to a training programme in three trials, and in each, the prompts increased the frequency of health professional intervention. Study *Ask/Brief advice *Cessation support N.R.T. †Abstinence Cohen 1989 (primary care) U.S.A.

66% prompted vs. 27% control, significant.

Set a quit date: 14% prompted vs. 3% control, significant.

Not reported 7.9% prompted vs. 1.5% control, significant.

Cohen 1989 (dental) U.S.A.

29% prompted vs. 18% control, significant.

Set a quit date: significant. Not reported Not significant

Cummings 1989 (Health Maintenance Organisation) U.S.A.

Asked to quit: significant.

Discuss smoking: 50.1% trained vs. 44.9% control, significant. Arrange a follow-up appointment: 15% trained vs. 5% control, significant.

Both groups ≈ 10%, no significant difference between groups.

8% trained vs. 7.1% control not significant.

Cummings 1989 (private practice) U.S.A.


Discuss smoking: 64% trained vs. 44% control, significant. Arrange a follow-up appointment: significant.

19% trained vs. 13% control, not significant.

6.7% trained vs. 8.2% control not significant.

Kottke 1989 U.S.A.


Set a quit date: significant. Arrange a follow-up visit: not significant.

Not reported Both groups ≈ 5%, no significant difference between groups.

Lennox 1998 Scotland

Ask: significant. Discussion of smoking: no significant difference between groups but rates high, ≈ 70%.

Not reported Not reported 3.6% intervention vs. 4.7% control, no significant difference (at 14 months).

Sinclair 1998 Scotland

Ask: significant. Discussion of smoking: 85% trained vs. 62.3% control, significant.

Not reported Not reported 12% intervention vs. 7.4% control, no significant difference (at 9 months).

Strecher 1991 U.S.A.

Advise smokers to quit: 73% trained vs. 58% control.

Write a quit day prescription: significant. Give self help materials: not significant. Scheduling follow-up: not significant.

Not reported 1.7% intervention vs. 5.7% control, no significant difference (at 6 months).

Wang 1994 Taiwan

Not reported Not reported Not reported 28.6% intervention vs. 4.3% control, significant difference but not for the prompt only group, 8.3% (self-reported at 6 months).

35


Table 13: Systematic review results: The effectiveness of interventions to

increase the delivery of effective smoking cessation treatments in primary care settings (continued)

Grimshaw et al. (2004). Systematic review (of level II, III and III-3 studies), all interventions, smoking specific and general prevention studies including smoking. Summary of the effectiveness of those interventions applied in smoking specific trials and/or general prevention trials reporting smoking outcomes. Intervention †Effect size Comment

Reminders

Moderate

The most frequently evaluated single intervention.

Educational outreach

Modest

Generally more effective than educational materials.

Gene

rally

effec

tive Multifaceted

interventions Modest

The combination of educational materials, educational meetings and reminders appears more effective than educational materials and educational meetings alone. However, overall (smoking and general prevention), there appeared to be no evidence of an expected dose-response curve associated with increasing the number of components in an intervention.

Mixe

d effe

cts Audit and feedback Modest

This finding was less robust as only two of the included smoking studies used audit and feedback as an intervention component.

Gene

rally

ineffe

ctive


Modest

While education materials are often important, as a stand-alone intervention, their effects may be short lived.

†Effect sizes: the median absolute difference across post-intervention measures (all) – and largely consistent with the limited number of smoking trials. Modest = effect sizes > 5% and ≤ 10% Moderate = effect sizes >10 and ≤ 20% Note: also reported …

Results of single interventions and multi-faceted interventions from all included cluster randomised controlled trials: including smoking specific, general prevention including smoking and trials not including any smoking outcomes. Intervention Number of

trials Range of observed effects*

*Median effect size

Educational materials 5 +3.6 to +17.0 +8.1 Audit and feedback 5 +1.3 to +16.0 +7.0 Reminders 14 -1.0 to +34.0 +14.1 Multi-faceted including educational outreach

13 -4.0 to +17.4 +6.0

*Absolute percentage improvement in guideline compliance

36


Table 13: Systematic review results: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

Anderson et al. (2004). Systematic Review, education, outreach, office systems. Systematic review/meta-analysis of 19 trials; 16 RCTs, two controlled clinical trials (CCTs) and one an interrupted time-series design (ITS). Outcomes Mean control

rates ‡Intervention Weighted mean proportion difference (overall)

(95% CI)

Ask ‘A’ 48% + 15%

(7–22)

Advice ‘B’ 51% + 13% (9–18)

∆Quit rate ‘Q’ 16.9% + 4.7% (2.5–6.9) The literature searches identified 696 papers, of which 58 were retrieved and 18 were included in the final analysis. Five of 18 trials were conducted prior to 1990 and ten of 18 trials reported smoking related outcomes within trials of ‘general prevention’ programmes. Three trials were conducted pre-1990 and were general prevention programmes. Eight trials were conducted post-1990 and were tobacco specific trials and reported ‘ask’ and or ‘brief advice’ (two trials reporting only quit rates). For established practitioners, programmes were effective in changing their screening and advice-giving rates, but not their patients’ quit rates. For practitioners in training, programmes were found to be effective in changing their advice-giving rates and their patients’ quit rates but not their screening rates. Overall, the programmes led to a 15% (95% CI = 7–22) increase in practitioners’ screening rates (mean control rates of 48%); a 13% (95% CI = 9–18) increase in practitioners’ advice-giving rates (mean control rates of 51%), and a 4.7% (95% CI = 2.5–6.9) increase in bio-chemically verified patient quit rates (mean control rates of 16.9%). These findings suggest that the behaviours of practitioners in training might be more malleable than those of established practitioners. Generally, multi-component programmes were more effective than single-component programmes. Programmes combining educational and practice-based measures were more effective than educational or practice-based alone. The regression analysis found a greater effect for outreach as opposed to non-outreach programmes, explaining 27% of the variation in effect size, p < .001. Tobacco-specific programmes were more effective as opposed to general prevention programmes, explaining 22% of the variation in effect size, p < .01 ∆While patient outcomes are outside the scope of this review, for completeness, quit rates have been reported here. However, these aspects of the included trials, as they relate specifically to patient outcomes, have not necessarily been appraised or synthesised further. Further details of patient outcomes are available in the original authors’ publications. ‡Additional benefit of intervention compared to control. Overall, including practitioners in training and established practitioners.

37


Table 13: Systematic review results: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

Dicky et al. (1999). Systematic review, office tools and teamwork, smoking specific & general prevention studies. Systematic review of 33 trials: seven smoking specific trials of ‘office tools’; 17 general prevention trials of ‘office tools’; six smoking specific trials of ‘teamwork’; three general prevention trials of ‘teamwork’ interventions. Outcomes Number of trials Percentage of trials reporting a statistically significant

increase in the provision of smoking cessation counselling.

†Counselling ‘C’ Smoking specific trials

13 92%

‡Counselling ‘C’ General prevention including smoking

12 58%

One Meta-analysis, 13 randomised controlled trials, five controlled clinical trials, ten time-series designs and four pre-test/post-test trials. †Counselling rates for smoking cessation as reported by patient survey and/or chart review in 13 ‘smoking specific’ trials (including one meta-analysis). ‡Counselling rates for smoking cessation reported within ‘general prevention’ programmes.

38


Table 14: Other review results: The effectiveness of interventions to increase the delivery of effective smoking cessation treatments in primary care settings

Fiore et al. (2008). Clinical Practice Guideline, training and charting (reminders). †Results for topics meta-analyzed for the 2008 Guideline update: topic– system features, five comparisons. Comparison 1 Effectiveness of clinician training combined with charting on asking about smoking status (Ask) (n = 3 studies) Intervention Number of arms Odds Ratio (95%C.I.) Estimated rate (95% C.I.)

No intervention 3 1.0 58.8 Training & Charting 3 2.1 (1.9–2.4) 75.2 (72.7–77.6)

Comparison 2 Effectiveness of clinician training on rates of providing treatment (Assist/Cessation) (n = 2 studies)

Intervention Number of arms Odds Ratio (95% C.I.) Estimated rate (95% C.I.)

No intervention 2 1.0 36.2

Clinician training 2 3.2 (2.0–5.2) 64.7 (53.1–74.8)

Comparison 3 Effectiveness of training combined with charting on setting a quit date (Assist/Cessation) (n = 2 studies) Intervention Number of arms Odds Ratio (95%C.I.) Estimated rate (95% C.I.)


Comparison 4 Effectiveness of training combined with charting on providing materials (Assist/Cessation) (n = 2 studies) Intervention Number of arms Odds Ratio (95%C.I.) Estimated rate (95% C.I.)


Comparison 5 Effectiveness of training combined with charting on arranging for follow-up (Arrange/Cessation) (n = 2 studies) Intervention Number of arms Odds Ratio (95%C.I.) Estimated rate (95% C.I.)

No intervention 2 1.0 6.7 Training & Charting 2 2.7 (1.9–3.9) 16.3 (11.8– 22.1)

†Based on the “5As” framework [with the New Zealand ABC approach indicated in brackets] – Ask [Ask ‘A’], Advise to quit [Brief Advice ‘B’], Assess willingness to quit, Assist with quitting, Arrange follow-up [Cessation support ‘C’]. Note: also reported on smoking abstinence outcomes.

39


Original primary studies: introduction The searches identified 37 eligible primary research studies and the summary characteristics of all 37 included studies are presented in Appendix D . Heterogeneity of the studies in terms of population, interventions assessed, control groups, as well as outcome measures and follow-up periods, precluded the conduction of a meta-analysis. Therefore, throughout the following section, the study results are grouped according to the type of intervention, listing the more frequently reported first, and within the groupings, accordingly from higher to lower level of evidence and finally according to publication date order. The intervention groupings are briefly overviewed below and Table 15 presents the intervention types along with the number of eligible studies included. Then, for each of the groupings, the intervention is first briefly described, followed by a narrative description of each of the included studies, followed by a narrative synthesis/summary of the studies reported within the intervention group, and finally, the data Tables 30-66. The body of evidence is summarised in Table 23. Overview of intervention groupings Training In 15 of the 37 studies (40.5%), the intervention involved training health care professionals to increase their delivery of recommended treatments or to implement a particular treatment strategy (Borgiel & Williams, 1999; Cockburn, et al., 1992; DePue, et al., 2002; Dietrich, et al., 1992; Goldstein, et al., 2003; Hogg, et al., 2008; Joseph et al., 2004; Kottke, et al., 1992; Lennox, et al., 1998; McRobbie, et al., 2008; Morgan et al., 1996; Ockene, et al., 1997; Secker-Walker, et al., 1992; Solberg, et al., 2000; Unrod et al., 2007). Multi-component Eight of 37 studies (21.5%) combined training with another intervention component or components to form a multi-component intervention approach (Flocke, et al., 2006; Katz, Muehlenbruch, Brown, Fiore, & Baker, 2004; Lemelin, et al., 2001; Ozer, et al., 2001; Pbert, et al., 2006; Strecher et al., 1991; Tremblay, et al., 2001; Young, et al., 2002). Adjunct components included a single office-based tool in two studies – with an algorithm (Ockene et al., 1994) or with a reminder system (Pbert et al., 2006) through to many components. For example, Katz et al. (2004) evaluated a guideline-based intervention with five components (a tutorial for intake clinicians, group and individual performance feed-back for intake clinicians, use of a modified vital signs stamp, an offer of free nicotine replacement therapy, and proactive telephone counselling) and Lemelin et al. (2001) evaluated a programme with seven intervention components and/or strategies (audit and ongoing feedback, consensus building, opinion leaders and networking, academic detailing and education materials, reminder systems, patient-mediated activities, and patient education materials). Reminders Five of the 37 studies (13.5%) evaluated the effectiveness of interventions that comprised electronic reminders (Krist, et al., 2008; McPhee, et al., 1991) and non-electronic reminder systems (Fiore, et al., 1995; Piper, et al., 2003; Rothemich et al., 2008). Typically, reminder systems involve non-medical or rooming staff (e.g. nurses or medical assistants or receptionists) routinely recording patients’ smoking status in the electronic medical record or in the patient notes (either by a written note, expanded ‘vital signs’ stamp or modified stationery). Thus, the GP or Physician is actively alerted if his or her patient is a smoker and prompted to take action– i.e. to provide brief advice and/or offer further cessation support.

40


Financial incentives Four studies (11%) evaluated financial incentives: two studies evaluated financial incentives within the context of smoking specific trials and two within the context of general prevention trials, including smoking cessation outcomes. One moderately large cluster randomised controlled trial (n = 40 clinics) was conducted in the United States (Roski et al., 2003) and three large population based interrupted time series studies were also included: two recent studies in the United Kingdom (Coleman, et al., 2007; Millett, et al., 2007) and one similar study in the United States (Amundson, et al., 2003). The intervention, in all cases, was some form of financial incentive scheme, whereby bonus payments were paid to individual practices or medical groups based on the performance of specified smoking cessation activities, as measured against predetermined targets. Thus, underpinning all four studies was the concept of ‘pay-for-performance’. Audit & Feedback Three of 37 studies (8%) evaluated audit & feedback; two of the studies were smoking specific (Bentz et al., 2007; Wadland, et al., 2007) and one was conducted within the context of a general prevention trial (Moher, et al., 2001). Feedback reports rated individual HCPs’ performance in asking, advising, assessing, and assisting with tobacco cessation, and then comparisons were made with a clinic average, and/or an achievable benchmark of care (defined as the average performance of the top 10% of providers being measured). Other Finally, two studies (5.5%) evaluated other intervention types that did not easily fit into the above categories. One randomised trial evaluated the effectiveness of reinforcement contact from the medical practitioner trainer as compared to training only (Richmond, Mendelsohn, & Kehoe, 1998) and one randomised controlled trial evaluated an algorithm as a stand-alone intervention (the provision of intervention algorithm at each patient visit compared to a no-algorithm condition) (Ockene et al., 1994). Table 15: Intervention groupings and number of trials

Intervention type Number of studies Training 15 Multi-component 8 Reminders 5 Financial incentives 4 Audit and feedback 3 Other 2 Total 37

41


Original primary study results: by intervention type and level of evidence Training Introduction Training interventions were interventions in which the principal component was imparting or transferring new knowledge, skills, and competencies to HCPs to enable them to provide better care for their smoking patients by improving the delivery of smoking cessation treatments within the practice. For the purpose of this review, Training is categorised distinctly from ‘educational materials’ or simply education. Educational materials, as an intervention, refers to the distribution of published or printed recommendations for clinical care including clinical practice guidelines, audiovisual materials and electronic publications. Education also includes professional development, whereby the HCP chooses to (or is required to) advance his or her knowledge by higher level study and/or reading journal articles, clinical practice guidelines, and other materials intended to lead to a change in clinical behaviour. Fifteen studies involved training as the principal component of the intervention. In addition, all of the eight ‘multi-component’ studies also included training as a main ingredient. However, these eight studies are reported separately under Multi-component interventions, as it was not possible to determine the relative contributions of the training and the adjunct intervention components separately. In total, 11 of 15 studies (73%) were randomised controlled trials or trials that used cluster randomisation (level II evidence); two studies were non-randomised clinical trials (level III-2 evidence), and two studies used a pre-test/post-test study design (level III-3 evidence). Three trials of training interventions were conducted within the context of ‘general prevention’ programmes; 11 of 15 trials (73%) involved training HCPs ‘on-site’ in their own clinics and 27% of studies involved off-site training. In total, studies of training-based interventions that aimed to increase the delivery of effective smoking cessation treatments involved more than 3000 HCPs across five countries. Typically, study authors defined their training interventions using the terminology that best described the focus or method or content of the training that they believed underpinned the intervention (e.g. tutorial/lectures, outreach facilitation, workshops). For the purpose of clear presentation, all studies that involve training of any type (as the principal component of the intervention) have been grouped together under the heading Training, and one of the broad descriptors from Table 16 has been added to indicate the most appropriate sub-category or type, as defined/described by the original study author/s (however, the interventions were not adequately described in all cases and the differences between the types of training was not always entirely clear). In some cases, the intervention involved training HCPs via a combination of modes, and these interventions have been categorised by the main or featured type or mode of training, or that highlighted by the study author/s.

42


Table 16: Sub-categories of training as defined by the original study authors

Intervention type Training type Number of studies

Academic detailing 7 Tutorial/lecture/workshops 3 Educational facilitator; Outreach facilitation; Outreach visits 2 Continuing Medical Education (CME) 2 Continuous Quality Improvement (CQI) 1

Training

Total 15 Training: results from 15 studies Outreach facilitation, evidence level II. In a large cluster randomised trial including 54 volunteer primary care practices comprising approximately 160 physicians, Hogg et al. (2008) trialled outreach facilitation in a general prevention study (including the measurement of 14 preventive manoeuvres). Outcomes included patient reports of being asked about smoking status, the provision of smoking cessation counselling, and the prescription of nicotine replacement therapy (NRT) as determined by patient chart audit (3049 patients). During the 11.5 month intervention period, two nurses were assigned 13 or 14 practices each and the nurses conducted outreach visits to each practice on average once a month for an average of 46 minutes. The facilitator delivered training in various ways of integrating preventive care into routine/episodic care visits and shared tools used for improving the quality of preventive care delivery, such as prevention flow sheets, chart flags, sticker reminders, electronic reminders, and patient care records. The intervention included periodic follow-up and consensus building every three to six weeks. This outreach facilitation program did not produce improvements in the delivery of preventive care. The authors reasoned that this lack of effect may have been due to the intervention simultaneously facilitating a high number of preventive manoeuvres, the relatively short intervention period, and the large number of practices assigned per facilitator. Resultantly, the facilitators delivered a weak intervention. Lectures, evidence level II. In a cluster randomised trial of 91 GPs from three east London primary care trusts (55 sites), McRobbie et al. (2008) evaluated the impact of a 40-min interactive training session on referrals received by their local specialist smoking cessation services (SSCS), as compared to a control group of GPs who received referral guidance by post only. At three months follow-up, and after adjusting for other factors, the number of patients referred to the SSCS was five times greater among GPs in the intervention arm that the control arm. However, the effect of the intervention dropped off noticeably after the first month post intervention (RR 7.9, 95% CI 2.5 – 24.9) but still remained significant at three months. The authors concluded that a brief training session can significantly increase GPs’ referrals to smoking cessation services but that it is likely that the intervention would need to be reinforced periodically and/or combined with prompts and reminders to maintain effectiveness over time. The authors also acknowledge the study’s limitations, including a relatively short three month follow-up and the participation of only 24% of invited GPs (‘the enthusiastic minority’) and the ‘expert’ status of the educators. Academic detailing, evidence level II. In an attempt to test a training intervention delivered by a health educator using an academic detailing approach, Unrod et al. (2007) conducted a randomised controlled trial of six months duration. Physicians and their patients in four largest metropolitan boroughs in New York City were randomised to either the intervention or the control group. Included were 70 physicians. Patients were 518 smokers (eligibility criteria was

43


that be at least 18 years old, have smoked in the past 7 days, and have smoked more than 100 cigarettes in their lifetime), they were paid $20 for completing initial assessment and $10 for the follow up interview. The intervention aimed to increase physicians’ adherence to the Public Health Service Clinical Practice Guidelines on Tobacco Use and Dependence 5As. These are ask about smoking status; advise smokers to quit; assess readiness to quit; assist in quitting; and arrange follow-up. The training was conducted in a 40-minute visit by a health educator to the physician’s office to provide individual physician training in brief smoking cessation counselling based on the 5As framework and following an ‘academic detailing’ approach’. Physicians in the control condition were not given any training and were instructed to continue their usual smoking cessation practices. At six months smoking cessation outcomes, physicians’ performance of the 5As were assessed through the patient exit interviews of smokers. Overall, the physicians who received the intervention were more likely to perform the 5As than the control. The rates at which physicians advised patients to quit, assessed readiness to quit, assisted in quitting, and arranged follow-up were all significantly increased (p < .001). Per patient report, twice as many intervention as control physicians assessed patients’ readiness to quit (OR 5.06; 95% CI 3.22, 7.95). Furthermore, more intervention than control physicians advised their patients to quit smoking (OR 2.79; 95% CI 1.70, 4.59), two to four times as many intervention as control physicians provided quit smoking assistance at setting goals (OR 4.31; 95% CI 2.59, 7.16), or provided written materials (OR 5.14; 95% CI 2.60, 10.14), or discussed smoking cessation medication (OR 4.72; 95% CI 2.90, 7.68), or referred patients to a quit-smoking programme (OR 6.48; 95% CI 3.11, 13.49). Intervention physicians arranged a follow-up to discuss smoking at five times the rate of control physicians (OR 8.14; 95% CI 3.98, 16.68). Note: The effect on cessation rates was only modestly higher among patients who received the intervention (quit rate of 12% which did not reach statistical significance), but was consistent with quit rates achieved by other interventions in primary care (as indicated from other references in the study). However, the intervention did result in longer quit attempts, which the authors stated is one of the strongest determinants of cessation success Limitations of the study include the use of patient exit interviews which rely on patient recall, exposing the study to recall bias due to the possible inaccuracy in recall among smokers. Another limitation is the lack of an objective measure or chemical test performed to confirm the smoking status of participants. Academic detailing, evidence level II. Joseph et al. (2004) aimed to test the effect of modest intensity, practical systems changes that might increase the delivery of smoking cessation treatments. In a cluster randomised controlled trial with 12 months follow-up, 20 Veterans Affairs medical centres including approximately 200 HCPs were randomised to intervention or control conditions. The participants were all the physicians, nurses, psychologists, and pharmacists of the included 20 Veterans Affairs medical centres, and data from 5678 patients (both patient survey and medical chart data). The intervention, academic detailing, initially involved the off-site training of one facilitator and one key advocate per site by the principal investigator/interventionist, followed by two to three days spent visiting each site. The intervention aimed to identify barriers and prioritise goals specific to each site and to design strategies to increase delivery of smoking intervention to all smokers. The recommended strategies included the use of non-physician personnel for documentation, use of electronic reminders, adaptation of note templates to include smoking status, increasing the availability of NRT, and improving the identification of smoking status in the medical records. The intervention involved recommending the various strategies for implementation only, the

44


responsibility for implementation remained with the individual site. Individual sites proceeded to implement these strategies and individual sites may have implemented strategies in different combinations, or not at all. Implementation was reported as ranging from ‘moderate to significant progress’ towards these goals; however, the exact degree of implementation of the recommended strategies at the clinic level was not known. The intervention did not increase the participant report of being asked about smoking status or receipt of counselling. It did increase the rate of identification of smoking status in the medical record (p = .0001). The intervention also increased the documentation of advice, but this did not reach statistical significance. Similarly, site-level data showed an increase in the prescription of NRT, but this did not reach significance. The intervention did not increase the rate of counselling to stop smoking. The authors concluded that a longer implementation period may have been required to increase the effectiveness of the intervention and that improvements in smoking cessation processes are likely to require more intensive intervention in this population. The authors also note the uniformly lower rates of ABC as recorded in the medical records, as compared to the rates reported by the patient surveys. Joseph et al. (2004) propose that this may be explained by a greater likelihood that smokers recall receipt of counselling versus the likelihood that providers actually record counselling activity given. Continuous quality improvement (CQI), evidence level II. In a large cluster randomised controlled trial (the IMPROVE project), Solberg et al. (2000) tested the hypothesis that a managed care organization, working through an iterative process of change called continuous quality improvement (CQI), could stimulate its private primary care clinics to develop better delivery systems for eight clinical preventive services (including smoking cessation, ‘Ask’). All HCPs from 44 primary care clinics in greater Minneapolis-St. Paul (including approximately 500 primary care Doctors) were randomised to one of two groups; CQI or control. In the intervention group, each clinic was assisted to use CQI multidisciplinary teams to develop and implement system changes. The CQI support comprised a leader and facilitator for each clinic's team, an initial 6-hour conference overview of CQI methods and prevention systems, six 4-hour workshops over 6 months for experiential training in an improvement process to build prevention systems (reinforced with modular manuals), consultation by telephone (every 6 weeks) and by visit (every 3 months), bimonthly breakfast meetings, and a quarterly newsletters to clinic staff. The control clinics received no intervention and continued to provide ‘usual care’. In this trial, CQI methods did not result in clinically important increases in preventive service delivery rates. After 22 months of intervention, rates of asking patients their smoking status had increased from baseline rates by 8.6% in the control group and 15% in the CQI group; however this increase was not statistically significant. In fact, only one of the eight selected preventive services (pneumococcal vaccine) had a significantly greater increase in the intervention group (17.2% absolute increase from baseline to follow-up compared with a 0.3% absolute increase in the control group; p = .003). The authors described the continuous quality improvement intervention as a failure. At the time of the follow-up survey and audit, six clinics had not implemented any system changes and the (suggested) critical process of systematically prompting or reminding the clinician about missing prevention services during a visit were only in place for a few services, in a few of the intervention clinics. If office systems are an important vehicle for improvement in the delivery of clinical preventive services, then, as Solberg et al. (2000) suggest, the critical unanswered question is “How might these systems be established and maintained in typical practice settings without research-funded support”? Continuing medical education (CME), evidence level II. Borgiel and Williams (1999) randomised 57 physicians in Ontario, Canada to one of two groups, the Continuing Medical Education Plan (CMEP) or the Practice Assessment Report (PAR) only group (control). For

45


the intervention group, peer mentors provided structured information for identifying relative strengths and weakness in the practice. Mentors, who were senior and respected physicians in the community, facilitated educational workshops and followed-up with a phone call at two or three months to review the physician’s commitments and subsequent educational activities. Physicians in the control group received a printed report of their practice’s performance only. At 12 months follow-up, comparing pre-test and post-test matched questionnaires and chart audits for 2395 patients revealed no statistically significant differences between groups on any of the study outcomes. Borgiel and Williams (1999) concluded that while some literature supports the use of trained physician-educators or opinion leaders for modifying practices, mentors may be more effective in educating physicians about specialty areas of practice, rather than offering broad general consultations. Workshops, evidence level II. Lennox et al. (1998) conducted a pragmatic cluster randomised controlled trial to assess the impact of a training intervention on health care professionals and their smoking subjects at one year follow-up. Approximately 140 GPs, nurses, and other staff from 16 practices (and a random sample of their patients) were recruited from all general practices in Aberdeen city, Scotland. Patients were not formally recruited into the study, and may or may not have attended their practice during the follow-up period. The intervention was an off-site, one-day ‘Stages of Change’ training workshop (including training in motivational interviewing) for primary health care teams. The workshops were delivered and run by a health promoter and a GP, and used an interactive approach. The workshops aimed to address three commonly stated barriers to HCPs delivering effective health promotion in primary care settings, specifically, a lack of training, low self-efficacy, and a perceived lack of time. Smoking patients in the intervention group appeared more likely than those in the control group to recall smoking having been mentioned in a consultation during the 14 months follow-up, but the differences did not reach statistical significance. Retrospective power calculations verified that the lack of statistical significance for GPs was not due to lack of power; however, for nurses and other HCPs, this may have been the case. There was no evidence to suggest improved patient outcomes. The authors concluded that changing health professionals’ attitudes and self-efficacy does not guarantee sustained changes in preventive behaviours and organisational factors are also important. Continuing medical education (CME), evidence level II. Ockene et al. (1997) undertook a randomised controlled trial to evaluate a community-based intervention for smoking cessation over four years and to report on the smoking intervention practices of physicians and their offices in the intervention and comparison communities. The aim was to determine the effect of the physician training activities on physician behaviour. The programme was the Community Intervention Trial for Smoking Cessation (COMMIT), a large randomised trial of continuing medical education in the USA conducted by the National Cancer Institute (1987). The trial included 11 matched pairs of communities throughout North America randomly assigned to the intervention or the comparison condition. Eight-hundred and sixty nine physicians participated in the study (465 in the intervention group, and 404 in the control group). Physicians in the intervention communities participated in continuing medical education (CME), training for office staff focused on tobacco control and an office intervention system. The intervention involved three levels of training activities for physicians (basic, comprehensive, and “train the trainers”), and one type of training for staff in physicians’ offices. The protocol provided guidelines for the length and content of training, but the frequency and format of training varied according to each community’s needs. The basic training included presentations on tobacco and the use of the National Cancer Institute “4As” (Ask, Advise, Assist, Arrange) counselling model. Outcomes measured were smoking control attitudes and practices reported by primary-care physicians, smoking policies and practices of

46


30 randomly selected medical offices in each community, and patient reports of physician intervention activities. To assess reported smoking intervention practices of the physicians in both groups, a 14-item survey was mailed to each targeted physician Based on the findings from this study, physicians in the intervention communities were more likely to attend training than those in the comparison communities (53% vs. 26%, respectively p < .0005). More smokers from the intervention communities reported receiving pamphlets about smoking than those in the comparison communities (p = .026). No other differences were found between conditions, and over 70% of all smokers reported being asked about their smoking, although less than half of light-to-moderate smokers had been advised to stop smoking while over half of heavy smokers reported such advice. Across both conditions, all of the physician intervention activities were reported more frequently by heavy smokers (p < .001). Academic detailing, evidence level II. Morgan et al. (1996) conducted a two-year randomised controlled trial and recruited primary care practices to compare usual care with brief quit-smoking advice and counselling for midlife and older smokers (aged between 50 and 74 years). Intervention practices received on-site training to implement a modified National Cancer Institute 4As framework. The training program was designed for brief office-based presentation using the principles of pharmaceutical and academic detailing. Of 49 eligible primary care practices, 39 practices indicated a willingness to participate in the brief office-based training program. Entire practices (physicians and key non-physician office staff) were randomised to the intervention group (immediate intervention n = 23) or control group (delayed intervention n = 26). Training involved praising patients for previous quit efforts, providing personalised feedback linking smoking to symptoms, discussing the health benefits of quitting for older smokers, and giving a clear message to stop smoking. Control practices were instructed to provide usual care to their older smokers over the accrual and follow-up period. The study measured the physicians’ perceptions of effectiveness giving quitting advice, and program’s effectiveness and feasibility (via physician self-report questionnaires). Provider adherence was also measured which included eight areas, two of them corresponding to B (brief advice) and C (cessation support). Results from the patient-reports of interactions with physicians at two to four weeks follow-up (for the intervention group) showed that 88% of patients reported getting clear physician quitting advice since enrolment B and 95% reported receiving the Clear Horizons guide (a copy of a smoking cessation guide tailored to older smokers). Only 35%, however, reported having received a smoking follow-up letter. Results from patient-reports of physicians interactions and quit rates at six-month follow-up, by study group, showed that there were significant differences between Immediate and Delayed practices with respect to receiving advice to quit B and discussions regarding smoking, setting a quit date and nicotine gum use C. There was a higher provider adherence levels reported by patients for trained physicians. The majority of the intervention group physicians continued to address smoking and cessation with patients during the six months following the initial visit. Six-month abstinence rates were nearly doubled by training physicians and staff to deliver a smoking cessation intervention tailored to midlife and older adults. Educational facilitator, evidence level II. Cockburn et al. (1992) randomly selected 264 general practitioners from the Melbourne metropolitan region and allocated them to receive one of three approaches for the trial distribution of a smoking intervention programme: a quit smoking intervention kit. The three approaches differed in intensity from the most intensive approach, personal delivery and presentation of a smoking cessation kit by an educational facilitator (visit lasted an average of 12.8, SD 5.8 minutes) with a follow up visit six weeks later, to delivery to the receptionist by a friendly volunteer courier (unannounced visits) with a

47


follow up phone call six weeks later, and least intense, postal delivery with a follow up letter six weeks later. In the most intensive approach, the educational facilitators made appointments to see general practitioners in the same way as pharmaceutical representatives, and on average the visits lasted approximately 13 minutes. The primary outcome was the measured use of components of the kit at four months follow-up. Specifically, GPs offering their patients simple advice to quit, and/or giving the patient a comprehensive self help booklet and/or a ‘help’ card, and/or cessation advice within a structured framework together with a ‘contract’ card whereby the patient makes a written contract with the general practitioners to quit smoking by a certain date. Additionally, GPs’ smoking cessation activities were assessed via a self-report questionnaire which contained a number of subscales. Cockburn et al. (1992) reported a trend only towards general practitioners in the ‘educational facilitator’ approach being more likely to have used at least one of the ‘minimal intervention’ components of the kit (ask and/or providing brief advice) (p = .07). However, there was a significant difference in the mean number of contract cards used by general practitioners in each group. GPs in the educational facilitator group used significantly more contract cards (6.54; SD 12.59) than general practitioners in the other two groups (‘Courier’ 3.79; SD 10.03 and ‘Post’ 1.92; SD 6.33), p = .02. While the resultant increase in the delivery of smoking cessation assistance to patients was statistically and probably clinically significant, the authors reported that the educational facilitator approach, does not seem to be a realistic option due to the high cost of marketing one-on-one to GPs. Academic detailing, evidence level II. To test the impact of physician training and a facilitator assisted office system intervention on cancer early detection and preventive services, Dietrich et al. (1992) randomised office based family physicians and general internists in 98 practices in New Hampshire and Vermont, using a 2x2 factorial design. The two interventions were compared singularly and in combination, with a control group. Both the education and the facilitated office system interventions occurred on site in the community practices. As such, both interventions could be described as academic detailing or educational outreach, although the content was distinctly different (i.e. education about preventive services versus education and assistance to select and implement office systems and tools). Specifically, the ‘education’ intervention comprised a day long, expert-led, group education meeting for physicians discussing smoking cessation advice for 120 minutes and other services for 30 to 60 minutes each. The ‘office system’ intervention comprised assistance from a project facilitator in the design and implementation of office routines that support provision of cancer early detection and preventive services, including team work (division of responsibilities for providing services), preventive care flow sheets, external identifiers for smokers' charts, health education posters and brochures, and patient held health diaries. The study outcomes were the proportion of eligible patients who reported receiving appropriate services during the previous year, at baseline and at 12 month follow-up, as measured by 2595 patient exit questionnaires (at least 446 identified smokers). Overall, the office systems intervention appeared more effective than the education intervention. However, for some outcomes, the combination of two interventions was no better than one alone, and for others, the interventions had no apparent effect. For ‘advice to quit smoking’, incidence rates did increase significantly for the office system only group (p = .017) but not for the combined intervention group. The authors speculated that this was possibly because the education sessions addressed controversies and perhaps raised doubts about some services. Dietrich et al. (1992) concluded that their results provide evidence that facilitator assistance in implementing an office system can increase the provision of preventive services. Academic detailing, evidence level III-2. In a non-randomised clinical trial of 24 months duration in two geographically different communities, Goldstein et al. (2003) implemented an

48


academic detailing intervention within a community-based sample of 259 primary care physicians (n = 88 intervention vs. n = 171 control) who provided primary care for at least 25% of their patients. The intervention, academic detailing, comprised a multi-component office-based assessment of physicians’ baseline knowledge and attitudes, identifying clear goals, and the use of supportive written materials. Analysing data from 4295 adult smokers’ self-report questionnaires, Goldstein et al. (2003) found that the percentage of patients who reported that their physician addressed smoking was similar between control and experimental groups. There was a borderline significant difference in intervention patients’ reports of cessation support at 18 months, as compared to the control group (p = .06). The authors concluded that although the percentage of patients who reported that their physician addressed smoking was similar between control and experimental groups, the percentage that quit after receiving counselling differed between the groups (18% control vs. 30% intervention; p = .03). Goldstein et al. (2003) suggested that the data indicated an increase in the effectiveness of the smoking cessation interventions actually delivered to individual patients. Academic detailing, evidence level III-2. In a non-randomised clinical trial, Kottke et al. (1992) tested the effect of an academic detailing/organisational support intervention on rates of smoking cessation advice given to patients in clinics identified within an independent practice association-type managed health care plan. Non-volunteer primary care clinics were the unit of analysis (n = 29), and included 142 primary care physicians and other support staff. The intervention involved repeated face-to-face contacts to recruit and train clinic personnel to implement a systematic smoking intervention program. The intervention, based on the “social learning model of consultation” (Brown & Schulte, 1987), included on average 5.7 site visits, 24 telephone calls, and 5.6 mailings per practice site over the 18-month period. The facilitators provided information about feasible and effective system changes (including allocating/defining the responsibility for identifying smokers), help with problem solving, assessment of program progress, and feedback and reinforcement. The programme was tailored to the unique attributes of each clinic. The intervention significantly increased the rates at which a population of primary care clinics identified their patients who smoked and advised them to quit smoking. The rate at which specific help was offered was notably higher in the intervention group, but it failed to reach statistical significance. The authors reported that not all of the clinics were pro-active in implementing specific smoking cessation interventions. At two years follow-up, just under half of the 29 clinic sites showed very little evidence of implementing a systematic programme. Despite this low rate of implementation, the intervention still demonstrated a statistically significant change in provider behaviour. Academic detailing, evidence level III-3. DePue et al. (2002) trialled an academic detailing approach in 14 community health centres in Rhode Island, across 14 separate sites. The pre-test/post-test study involved making on-site training available to 27 physicians, 53 nurses, one social worker, and 70 other staff (including medical assistants, dieticians, counsellors, receptionists, and dentists). In total 151 HCPs were eligible for training from six large, three medium, and five small sites, and follow-up data was collected by chart abstraction (1591 patient charts) at 12 months. The intervention involved three one-hour on-site provider training sessions, reviewing effective tobacco interventions; including chart prompts, ancillary staff involvement, physician audits with feedback, skills training (tobacco counselling skill building), and patient education materials. The trainer educated staff and made strong recommendations for a range of system changes (including systems for identification and documentation, team approaches, chart prompts). The trainer also facilitated staff goal setting, office system design and problem solving, and also facilitated selected staff to develop a plan to implement the selected practice changes. The authors note that the intervention, academic

49


detailing, involves experts training staff and making recommendations but with non-research staff applying the recommendations and implementing change at the site level. At follow-up, DePue et al. (2002) found that only a few practices had succeeded in implementing the changes as identified, because of competing demands. While the mean ‘ask’ rate increased significantly from 30% to 44% (OR 3.00; 95% CI 1.19 – 7.55), the ‘advise’ rate did not. The authors point out that since the premise behind documenting smoking status is to prompt provider advice, an intervention is essentially ineffective if ‘advise’ rates do not significantly increase over time. DePue et al. (2002) conclude that while academic detailing shows promise, their intervention may not have been intensive enough to effectively change these systems and beliefs. A more sustained effort may be needed to change provider behaviour, with repeated offerings over time, and with multiple messages from multiple sources both at local and higher levels. Workshops, evidence level III-3. Secker-Walker et al. (1992) assessed a single-component educational training intervention for residents in obstetrics and family practices, which aimed to increase the delivery of smoking cessation advice during prenatal care. The authors report data from a controlled before and after study (CBA) of the effectiveness of the training programme, nested within a concurrent randomised controlled trial. The aim of the randomised controlled trial (not reported here) was to assess the effectiveness of the training intervention and basic care protocols on patient outcomes, that is, quit rates. In both groups, rooming staff prompt the HCP by placement of a protocol prompt sheet in the patient notes (as administrative or rooming staff identify patients who smokers). The CBA study assessed the frequency and quality of the prompted HCPs’ advice by comparing responses from exit interviews by pregnant women taking part in the trial. Data from the first 20 residents is reported. Women were asked whether their doctor talked about smoking or delivered advice to stop smoking or advised to set up a quit date. Of women in the intervention group 98% and 96% reported they were asked about smoking ‘A’ (first and second ante-partum visit respectively), compared to 95% and 66% of women in the basic care group (first and second ante-partum visit respectively) (p = .0001). Similarly, those who reported that they were given advice to stop smoking in the intervention group were 96% and 91% (first and second ante-partum visit) as compared to 92% and 52% of women in the basic care group (first and second ante-partum visit respectively) (p = .0001). For cessation support, only “Advised you to cut down” reached statistical significance for between-group difference, intervention versus control group at first ante-partum visit (p < .05). Training resulted in significant changes in the advice provided by the residents, with greater emphasis on gaining a commitment to smoking behaviour change but not in the average time providing the advice, approximately three minutes. The women’s responses to the exit interviews after the first and second ante-partum visits showed high levels of adherence by the residents to both the intervention and basic care protocols, which remained steady throughout the time of the trial (two years). The study was still progressing at the publication date so involved first 20 residents in the analysis only. Training: results summary Summary data from the 15 studies of the effectiveness of training interventions for increasing the delivery of effective smoking cessation treatments in primary care settings are summarised below and in Table 17. Training interventions were interventions in which the principal component was imparting or transferring new knowledge, skills, and competencies to health care professionals to enable them to provide better care for their smoking patients by increasing the delivery of smoking cessation treatments and activities within the practice. Training appears to be an essential

50


component of programmes aimed at changing health care professionals’ behaviour, with respect to delivering preventive services in general, and with regard to the delivery of smoking cessation interventions to all patients who smoke. The characteristics of the training programmes varied widely. Three main themes emerged within the results and these themes related to (1) the effectiveness of training interventions over time, (2) differences in the content and focus of the training and (3) differences in the settings in which the training was delivered. 1) The effects of training typically diminish over time. While training appears to be essential to initiating behaviour change, the provision of a ‘one-off’ training programme (whether delivered in one session or more than one session and without appropriate reinforcement and system level support) does not appear to guarantee sustained effects on clinical practice. Several authors comment on the tendency for the effects of training to diminish over time, and in some cases, the targeted cessation activities quickly regressed to baseline or near baseline levels (DePue, et al., 2002; Joseph, et al., 2004; Kottke, et al., 1992; McRobbie, et al., 2008; Solberg, et al., 2000). These authors all reported declining rates of participants performing the tasks of smoking cessation and/or poor implementation of systems at follow-up. It appears that when HCPs are left unaided by outside researcher assistance, the individual’s behaviour change or collective effort to set up office systems diminishes over time, and at longer follow-up periods, implementation of the changes is often poor or non-existent. For example, Kottke et al. (1992) reported that two years after their academic detailing intervention, under half of the intervention clinics showed very little evidence of implementing a systematic programme. McRobbie et al. (2008) reported a drop in the rate of referrals to a specialty cessation services by trained HCPs after only one month (although between group differences remained significant at three months). However, not all of the studies showed this tendency. Dietrich et al. (1992) found that increases in rates of HCPs providing brief advice (for the systems training group only) remained statistically significant at 12 months, following a one day, expert-led, group education meeting. Ockene et al. (1997) evaluated Continuing Medical Education, and at 48 months follow-up, some benefits were still evident for some components of cessation support. 2) Content: cessation skills versus systems training. Two distinct categories of training content or focus emerged. Specifically, training in the design and implementation of office systems or training GPs and other HCPs in the skills of delivering smoking cessation treatments (for example, specific patient centred counselling techniques such as motivational interviewing). For example, in eight studies (Borgiel & Williams, 1999; DePue, et al., 2002; Dietrich, et al., 1992; Hogg, et al., 2008; Joseph, et al., 2004; Kottke, et al., 1992; Ockene, et al., 1997; Solberg, et al., 2000) the main focus of the intervention was training physicians how to develop and implement an appropriate office system or systems for the routine assessment of patients’ smoking status and the subsequent provision of cessation services. In seven studies (Cockburn, et al., 1992; Goldstein, et al., 2003; Lennox, et al., 1998; McRobbie, et al., 2008; Morgan, et al., 1996; Secker-Walker, et al., 1992; Unrod, et al., 2007), the facilitator trained some or all of the office staff in specific cessation skills or how to actually deliver such cessation services to their patients. Thus, in all the studies, the fundamental task of the facilitator was to train HCPs but the point of intervention differs between the system level and the provider-patient interaction level. Most training interventions involved the use of a specially-trained person (usually one of the research team) who met with providers in their practice settings to provide information of greater or lesser complexity. This type of face-to-face visit has been referred to in the literature

51


in a variety of ways including university-based educational detailing, academic detailing, educational visiting, Continuing Medical Education (CME), Continuous Quality Improvement (CQI), educational facilitation, outreach facilitation, outreach visits, and tutorials and lectures. More complex training involved assessing the needs and motivational readiness of the targeted clinician, assessing the particular barriers and enablers of the practice setting, and subsequently tailoring the educational intervention to best match the identified needs. Less complex educational interventions simply aimed to train the targeted clinician in practical skills (e.g. specific counselling skills of using an electronic referral system). Figure 5 overviews the effectiveness of the 15 studies of training interventions, grouped by the predominant training approach or content. That is, either the training of HCPs (either individually or in groups) in the effective delivery of smoking cessation interventions or the training of HCPs in the development and implementation of system changes intended to increase the clinic-wide delivery of smoking cessation interventions. As illustrated, five of seven (71%) trials of cessation skills training resulted in statistically significant between group differences in process outcomes (generally beyond simple ‘Ask’ rates). In contrast, only two of eight (25%) trials of training HCPs in system changes resulted in such differences in process outcomes.

012345678

Cessation skills Systems

Cessation skills training versus training in system design and implementation

Not significant 2 6Significant 5 2

Figure 5 The effectiveness of the 15 included studies of training interventions, grouped by the predominant training approach: training HCPs in specific cessation skills vs. training HCPs to design and implement systems.

For example, Joseph et al. (2004) evaluated off site training of one facilitator and one ‘key advocate’ per site to identify barriers and goals and implement appropriate system changes. The researchers made recommendations only as to the possible systems that could be implemented, but the responsibility for implementation was left to the individual sites. While there was a significant increase in the documentation of patients’ smoking status in the medical records, there was no such difference as measured by patient survey and no other between-group differences were observed (including quit rates). The authors concluded that more intensive interventions were required to implement change. In contrast to this systems approach, Unrod et al. (2007) demonstrated that training individual physicians in brief smoking cessation counselling skills can significantly increase primary care physicians’

52


implementation of the 5As and result in modest effects on patients’ smoking outcomes. Likewise, McRobbie et al. (2008) demonstrated statistically significant increases in referral rates following a 40-min interactive training session by ‘expert’ educators that addressed the rationale and skills for referral of smokers for treatment. While the literature supports the use of trained educators for modifying clinical behaviour, educators may be more effective in educating health care professionals about their specialty areas of practice (e.g. rooming staff recording patients smoking status or nurses providing brief advice or GPs counselling patients) rather than broad and generalised training in system changes. 3) Setting: on-site versus off-site training Common to 11 out of the 15 studies (73%) of training was the use of ‘educational outreach’ interventions that feature personal educational visits to clinicians in their own practice settings, rather than ‘off-site’ training workshops or conferences. On-site training appears more effective than off-site training and this is possibly a reflection of both the likelihood of participation and the type of training. Participation rates for on-site training were typically higher than for off-site training (e.g. DePue et al. 2002 [67%]; Goldstein et al. 2003 [80%]; Unrod et al. 2007 [90%], versus a mean participation rate of less than 50% for off-site training), and training that occurred off-site tended to focus more on system changes rather than training individuals in counselling skills. However, there were exceptions, and there is insufficient evidence to delineate the possible contributions of different programme contents and settings. Training: conclusions Training healthcare professionals is more effective than no training in increasing the numbers of smokers who have their smoking status recorded in the medical records, are offered brief advice about stopping, and subsequently receive at least some cessation support. It appears that the most effective form of training is that of training HCPs in the skills of actually delivering smoking cessation treatments to their patients (i.e. training in effective counselling techniques or methods). Training HCPs to design and implement systems and to initiate system-wide changes does not appear to be a good use of resources. Thus, training will need to be focused on skills, be practice-based, utilise existing opportunities, and be supplemented by other interventions or systematic processes (for example reminder systems). It appears that training-based interventions perform best when they are implemented alongside other interventions.

53


Table 17: Original study results: The effectiveness of Training to increase the delivery of effective smoking cessation treatments in primary care settings

TRAINING Training: Level II evidence Hogg et al. (2008) – Training, outreach facilitation, general prevention, 14 clinical preventive services. *Outcomes Control Intervention p value

Mean difference between groups

95% CI

Ask ‘A’ 33.31% 35.0% p = .64 1.7% -5.6 – 8.9 Advice ‘B’ – – – – Not reported †Cessation ‘C’ (NRT) 13.5% 20.4% p = .16 6.9% -2.7 – 16.5 ‡Cessation ‘C’ (counselling)

38.1% 41.9 p = .3 3.7% -3.6 – 11.0

*This general prevention trial included outcomes that related to 14 clinical preventive services including ‘asking’ smoking status and cessation support. The smoking cessation outcomes only are reported here, as other patient outcomes are outside the scope of this review. Aspects of the study that relate specifically to these other patient outcomes have not necessarily been appraised or synthesised further, however, details of these patient outcomes are available in the original publication. †Prescription of nicotine replacement therapy as assessed chart review. ‡Providing smoking cessation counselling as reported by 3049 patient exit interviews. McRobbie et al. (2008) – Training, one 40-min interactive training session, smoking specific. Outcomes Pre-test

†Post-test

p value Incidence rate ratio RR and 95% confidence interval (95% CI)

Ask ‘A’ – – Not reported Advice ‘B’ – – Not reported Cessation ‘C’ 0.6(control) /1.0(exp) 1.8(control) /6.4(exp).

p < .001 4.9 (1.7 to 14.7)

†The mean number of referrals per GP as recorded by the local specialist smoking cessation services (SSCS) at three months post intervention.

54


Table 17: Original study results: The effectiveness of Training to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

TRAINING Training: Level II evidence Unrod et al. (2007) – Training, academic detailing, smoking specific. *Outcomes Did your doctor:

Control Intervention

p value Odds ratio (95% CI)

Ask ‘A’ “ask whether you smoke?” 47.4 61.2 Not significant Advice ‘B’ “advise you to quit smoking?”

53.0% 76.8% p = .0001

2.79 (1.70, 4.59)

Assisted ‘C’ (any) Assess: “assess whether you are ready to quit?” Assist: “help you set goals about quitting?” Assist: “give you written materials?” Assist: “refer you to a quit-smoking program?” Assist: “talk to you about quit-smoking medications?” Arrange: “make a follow-up appointment to discuss smoking?”

36.8% 20.2% 6.9% 4.5% 24.7% 9.7%

76% 55.1% 32.3% 23.2% 61.6% 47.5%

p < .0001

5.06; (3.22, 7.95) 4.31; (2.59, 7.16) 5.14; (2.60, 10.14) 4.72; (2.90, 7.68) 6.48; (3.11, 13.49) 8.14; (3.98, 16.68)

†Quit rate 7- day point-prevalence abstinence Longest quit attempt in days: M (SD)

8% 12.4 (29.6)

12% 18.4 (36.7)

p = .078 p = .05

1.77 (0.94, 3.34)

*Physician performance of the 5As – measures of smoking behaviour from patient exit interviews “Did your doctor…?” †Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. M – mean SD – Standard deviation Joseph et al. (2004) – Training, academic detailing, smoking specific. Medical record review Patient survey Outcomes Control Intervention p value Control Intervention p value Ask ‘A’ 60.7% 67.0% p = .0001 74.3% 76.0% p = .71 n.s. Advice ‘B’ 69% 61% p = .11 n.s. 71.8% 73.9% p = .60 n.s. Cessation ‘C’ (NRT)† 27.94 31.26 p = .07 n.s. 56.4% 61.2% p = .32 n.s. † Mean number of patients/1000 receiving medication.

55



TRAINING Training: Level II evidence Solberg et al. (2000) – Training, Continuous Quality Improvement (CQI), general prevention, eight clinical preventive services. *Outcomes Control (% increase) Intervention (% increase) p value †Ask ‘A’ 8.6% 15% n.s. *This general prevention trial included outcomes that related to eight clinical preventive services including ‘asking’ smoking status. The smoking cessation outcome only is reported here, as other patient outcomes are outside the scope of this review. Aspects of the trial that relate specifically to these other patient outcomes have not necessarily been appraised or synthesised further, however details of these patient outcomes are available in the original publication. †The proportion of patients who were up-to-date within three months of an index primary care visit on the basis of chart audit. Borgiel et al. (1999). Training, mentor-led education, general prevention. *Outcomes Control Intervention p value Change †Ask ‘A’ and/or Advice ‘B’ 79.1 % 83.1 % n.s. 4% Cessation ‘C’ n.s. Note: no statistically significant differences between groups were observed on any or the study outcomes. *Smoking outcomes reported within the context of a ‘general prevention’ trial, including appropriate prevention/charting of vaccinations, blood pressure, influenza vaccine, discussed smoking, discussed cholesterol testing, asked about alcohol use, papanicolaou smear, breast examination, and referred for mammography. Further details of patient outcomes are available in the original authors’ publication. †The percentage of 2395 patients reporting that their physician discussed smoking. Lennox et al. (1998) – Training, education workshops, smoking specific. *†Outcomes Control Intervention p value Mean difference between

groups (95% CI) GPs Ask/Brief advice 74.9% 79.4% p = .09 4.5% (-0.71-9.71) n.s. Nurses Ask/Brief advice 76.2% 83.2% p = .19 7.0% (-3.58 -17.58) n.s. Other HCPs Ask/Brief advice

68.6% 73.7% p = .63 5.1% (-15.69-25.86) n.s.

‡Continuous abstinence 4.7% 3.6% p = .26 -1.1% (-3.03-‘-0.83) n.s. *Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. †Health professionals’ behaviour as assessed by postal questionnaire (of smoking patients) at eight and 14 months follow-up: recall of smoking being mentioned in a consultation since the time of the workshop. ‡ Continuous abstinence between eight and 14 months.

56



TRAINING Training: Level II evidence Ockene et al. (1997) – Training, Continuing Medical Education (CME), smoking specific. ∆Outcomes: patient reports Control Intervention Difference between

intervention and comparison

p value

Ask ‘A’ Patients asked about smoking status Light-to-moderate smokers Heavy smokers

73% 72.3% 77.5% 77.6%

-0.8 0.1

p < .001

Advice ’B’ advised to stop smoking Light-to-moderate smokers Heavy smokers

44.2% 41.6% 54.0% 55.5%

-2.6 1.5

p < .001

Assisted C (any) Given pamphlets Light-to-moderate smokers Heavy smokers Set a quit date Light-to-moderate smokers Heavy smokers Recommend nicotine replacement Light-to-moderate smokers Heavy smokers See your physician to stop smoking Light-to-moderate smokers Heavy smokers

21.0% 23.6% 27.8% 30.7% 9.9% 10.1% 14.2% 15.1% 19.8% 20.4% 29.5% 30.6% 15.3% 16.5% 23.4% 23.6%

2.6 3.0 0.1 0.9 0.6 1.1 1.3 0.3

p < .001

Outcomes: physicians reported practices by study condition Smoking intervention activities Post intervention survey (%) Intervention

(n = 465 Comparison (n = 404)

p value

†‡Advise to stop smoking ‘B’ 98.4 94.0 p = .009 †‡Set quit date ‘C’ 21.8 14.4 p = .004 Attended training in smoking cessation in past 4 years

52.6 26.1 p = .001

∆ Data represents weighted population estimates. Percentage equals mean percentage with the community as the unit of analysis. † item included in tobacco action score; range (0-10); each item awarded 1 point. ‡ respondents who answered that they did this “most” or “all of the time”.

57



TRAINING Training: Level II evidence Morgan et al. (1996) – Training, office-based physician training, smoking specific. ∆Outcomes: patient reports Control Intervention

p value Odds ratio

(95% CI) Ask ‘A’ Not reported Advice ‘B’ receiving advice to quit (n=566 patients)

66.4% 79.1% .001< p < .01 Not reported

Assisted ‘C’ (any) Discussions regarding smoking (n=564 patients) Setting a quit date (n=475 patients) Nicotine gum use (n=480 patients)

58% 12.4% 23.8%

70.8% 54.9% 61.8%

.001< p < .01 p < .001 p < .001

Not reported

Quit rates: (n = 659 patients)

8.16% 15.41% p < .005 Not reported

∆Patient report of physician interaction and quit rates at six-month follow-up: percentages represent percentage of patients Cockburn et al. (1992) – Training, educational facilitator, smoking specific. Outcomes Intervention

group: Educational facilitator

Comparator group (a): Courier delivery

Comparator group (b): Postal delivery

p value

Ask ‘A’ p = .07 n.s. Advice “B’ p = .07 n.s. Cessation ‘C’

6.54; SD 12.59

3.79; SD 10-03 1.92; SD 6.33 † p = .02

†Statistically significant result between intervention group and either of the comparison groups. Dietrich et al. (1992) – Training, academic detailing, general prevention, cancer early detection and preventive services. *†‡Outcomes Control Education

only Office systems only

Education & Office systems

Odds ratio (95% CI)

Ask ‘A’ – – – Not reported Baseline 73% 79% 83% 78% Advice ‘B’ 12 months 67% n.s. 73% n.s. §84% 80% n.s.

Cessation ‘C’ (any) – – – Not reported *Quit rate ‘Q’ – – – Not reported *Further details of patient outcomes are available in the original authors’ publication. †Proportion of eligible patients who reported receiving services during the previous year at baseline and at 12 month follow-up (n = 2595 patient exit questionnaires). ‡Smoking outcomes reported within the context of a ‘general prevention’ trial of nine preventive services. §For office system only vs. control p = .017

58



TRAINING Training: Level III-2 evidence Goldstein et al. (2003) – Training, academic detailing, smoking specific.

12 months 18 months 24 months *†‡§∆Outcome A B C A B C A B C Control 54.7 % 56.7% 17.8% 54.0% 55.9% 26.8% 52.0% 52.2% 8.9% Experimental 54.2 % 54.0% 17.2% 61.2% 60.9% 35.3%,

p = .06 53.2% 52.8% 10.1%

*Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. There was a borderline significant effect on self-reported smoking quit rates for those residing in intervention areas versus control areas (OR = 1.35; 95% CI .99 –1.83; p = .057). Although the percentage of patients who reported that their physician addressed smoking was similar between control and experimental groups, the percentage who quit after receiving counselling differed between the groups (18% control vs. 30% exp; p = .03). †‘A’ The proportion of patients who reported that their physician talked to them about their smoking habit. ‡‘B’ The proportion of patients who reported that their physician advised them to quit smoking. §‘C’ The proportion of patients who reported that their physician arranged a follow-up visit to discuss smoking. ∆There were no statically significant differences between groups for any of the study process outcomes. There was a borderline significant difference in intervention patients’ reports of cessation support at 18 months, as compared to control (p = .06). Kottke et al. (1992) – Training, organisational/outreach educational workshops, smoking specific. Outcomes Control Intervention p value †Ask ‘A’ 26.0% (±12.2%) 39.8% (±12.3%) p < .05 ‡Advice ‘B’ 26.4% (±14.6%) 40.5% (±12.1%) p < .05 §Assisted ‘C’ 13.4% (±11.1%) 22.8% (±11.5%) p < .1 n.s. †Rates of smoking cessation ‘asking’ to patients who had visited their clinic in the previous six months as measured by 24 question patient surveys ‘A’. ‡Rates of smoking cessation ‘advice’ to patients who had visited their clinic in the previous six months as measured by 24 question patient surveys ‘B’. §The patient was offered specific help if he/she expressed interest in quitting ‘C’. Training: Level III-3 evidence DePue et al. (2002) – Training, academic detailing/outreach, smoking specific. *Outcomes Pre-test Post-test p value Odds ratio (95% CI) Ask ‘A’ 30% 44% 3.00 (1.19-7.55) Advice ‘B’ 19% 26% n.s. 1.29 (0.69-2.43) Assist or Arrange ‘C’ 9% & 6% no change n.s. Not reported †Health check vs. medical 2.43 (1.53-3.87) *Documentation in the last visit (percentage recording): ‘Ask’ notation of smoking status. ‘Brief advice’ was considered any notation of advice or discussion about quitting smoking. ‘Cessation’ (reported here as a composite of ‘assist’ and/or ‘arrange’), based on 1591 patient charts. †Odds ratios for documentation in the last visit for ‘Advice’ at a routine yearly physical or health check ( non-acute visits), compared to during other medical visits.

59



TRAINING Training: Level III-3 evidence Secker- Walker et al. (1992) – Training, reports data for ‘before and after’ trial of training for clinicians (residents) as part of an ongoing randomised controlled trial for pregnant women (patient outcomes). Outcomes (first 20 residents) Pre-training Post-training

p value t

Content and style of smoking cessation advice

2.15 (SD 0.88) 6.05 (SD 0.69) p < .0001 -16.28

Average time for delivery of the advice

3.14 (SD 1.21) min 3.09 (SD 0.82) min Not reported (but stated not significant)

Not reported

*Exit interview responses First ante-partum visit Second ante-partum visit Basic care group

(n=125 women) Intervention group (n=125 women)

Basic care group (n=102 women)

Intervention group (n=102 women)

No. % No. % No. % No. % Ask ‘A’ Did your doctor… talk about smoking? 119 95 122 98 67 66** 99 96 Advice ‘B’ Advise you to stop or stay quit?

115

92

120

96

53

52**

94

91

*Exit interview responses from current smokers

First ante-partum visit Second ante-partum visit


Intervention group (n=125 women


Intervention group (n = 88 women)

No. % No. % No. % No. % Assisted ‘C’ (any) Did your doctor… Ask to set a quit date Advise you to cut down Ask you for a cut down date

18 73 5

14* 58*** 4*

100 50 37

80 40 30

7 41 7

7* 41 7*

63 42 37

74 49 44

*Number and percent of positive responses **p = .0001 for the 2 x 2 contingency tables comparing positive and negative responses of intervention and basic care group for this question at this visit. ***p < .05 for the 2 x 2 contingency tables comparing positive and negative responses of intervention and basic care group for this question at this visit.

60



TRAINING Training: Level III-3 evidence Secker- Walker et al. (1992) – Training, reports data for ‘before and after’ trial of training for clinicians (residents) as part of an ongoing randomised controlled trial for pregnant women (patient outcomes). Adherence to intervention and basic care protocols after one and two years of the trial (percent positive responses to each question).

First ante-partum visit Basic care group Intervention group 1 yr (n=64) 2 yr (n=61) 1 yr (n=62) 2 yr (n=63)

Exit interview question

No. %

No.

%

No.

%

Did your doctor: Ask “A” Talk about smoking?

61

95

58

95

61

98

Advice “B” Advise you to stop?

57

89

58

95

59

95

Assisted “C” (any) Ask you to set a quit date

9

14

9

15

49

79

There are no significant differences within groups.

61


Multi-component Interventions Introduction Multi-component interventions were those in which two or more intervention components were delivered together within a programme. In all eight studies of multi-component interventions, training was one of the principal intervention components. The studies evaluated programmes with between two (Pbert, et al., 2006) and seven adjunct components (Lemelin, et al., 2001). Four studies combined audit and feedback with training (Flocke, et al., 2006; Katz, et al., 2004; Lemelin, et al., 2001; Young, et al., 2002), four studies combined reminders with training (Katz, et al., 2004; Lemelin, et al., 2001; Ozer, et al., 2001; Strecher, et al., 1991), and four studies combined education materials and other practice resources with training (Flocke, et al., 2006; Lemelin, et al., 2001; Tremblay, et al., 2001; Young, et al., 2002). Additionally, two studies added an algorithm (Pbert, et al., 2006; Tremblay, et al., 2001), and two studies involved support from external human resources in the form of proactive telephone counselling (Katz, et al., 2004) and a health educator (Ozer, et al., 2001). While there was considerable overlap between the interventions, the various combinations and number of intervention components in each study was also quite diverse. Training, audit & feedback, and reminders were the most commonly combined intervention components. In most cases, the training was characterised as academic detailing by the original authors. This involved outreach visits by the researchers and/or other trained experts to impart new knowledge, skills, and competencies to HCPs and to introduce and implement the additional tools and/or resources that comprised the multi-component intervention. Multi-component: results from eight studies Multi-component, evidence level II. In eight paediatric clinics located throughout central Massachusetts, Pbert et al. (2006) evaluated the effectiveness of a smoking prevention and cessation intervention that was delivered by providers to adolescents (aged 13-17 years) during routine medical care, based on patient exit interviews. The clinics were randomly allocated to either provide a special intervention (33 clinicians) or to the usual care (28 clinicians). The intervention which was provided by clinicians (physician, nurse practitioner, physician assistant, or other) was tailored so that the provider was trained to engage the adolescent in a non-threatening discussion about smoking and encouraged him/her to consider ones beliefs and intentions regarding smoking. It allowed the adolescent to take in the provider’s advice and to determine what course of action he or she would like to take. The intervention was based on the 5A model recommended by the US Public Health Service clinical practice guideline. An algorithm was used to help in assessing the adolescent’s smoking status and tailored the intervention on the basis of the individual’s current smoking status and responses to key questions asked. Outcomes were assessed in both groups using adolescent reports in patient exit interviews (PEIs). Ninety-eight percent of the PEIs were completed in person and 2% were completed by telephone after the visit. Providers were aged 42 years on average; both groups involved 24 physicians, whereas the rest were other clinicians (four in the usual care and nine in the intervention group). Data showed that intervention group clinicians were more likely to engage in the steps recommended with all adolescents, regardless of smoking status. For current smokers, all except two steps (asking how much the teen smokes and sharing additional reasons for not smoking) were performed more frequently by the intervention providers than by the usual care group. For non-smokers and former smokers, all steps were performed more frequently by the intervention providers than the usual care providers. When tested for predictors of PEI score and individual intervention steps, it was found that being in the intervention group was associated with increased PEI average for all individual intervention

62


steps. In terms of time spent by the provider in discussing smoking during the visit, patients in the intervention group reported that their providers spent 4.3 minutes (SD: 3.13 minutes), compared with 2.9 minutes (SD: 2.85 minutes) in the usual care condition (p < .0001). These findings suggest that the brief smoking prevention and cessation intervention was effectively implemented by providers for adolescents in a busy paediatric primary care clinic. One possible limitation of the study is that the outcomes were based on adolescent recall of provider intervention, thus the data may have been subject to recall bias. The generalisability of the study results to other adult patient populations is also not known. Multi-component, evidence level II. Katz et al. (2004) conducted a large randomised controlled trial to validate the results of an earlier pilot investigation of a guideline-based intervention in which intake clinicians assessed smoking status in all patients and provided brief cessation counselling for smokers. The study assessed the effectiveness of implementing the Agency for Healthcare Research and Quality (AHRQ) Smoking Cessation Clinical Practice Guidelines at eight community-based primary care clinics (four as test clinics and four as control clinics, with 75 clinicians) in southern Wisconsin among 2163 adult smokers. The Guideline recommends that all clinicians (including nurses) strongly advise their patients who use tobacco to quit. Staff at the intervention groups implemented the intervention for two months. The intervention was a five-component intervention involving tutorial for intake clinicians (nurses or medical assistants who document the reasons for the office visit and check patients’ vital signs), group and individual performance feedback for intake clinicians, use of a modified vital signs stamp, an offer of free nicotine replacement therapy, and proactive telephone counselling. Staff at control sites received only general information about the AHRQ Guideline. The percentage of patients who were asked about smoking status at test sites was higher than at control sites (87% vs. 67%; adjusted OR = 3.1, 95% CI 1.2 – 8.2; p <.001). This finding was similar for patients asked about their willingness to quit smoking (73% vs. 30%; adjusted OR = 6.4, 95% CI 3.7 – 10.8; p <.001), or those who were given literature about quitting C (38% vs. 3%; adjusted OR = 21, 95% CI 8.8 – 49; p < .001), were assisted with setting a quit date C (27% vs. 1%; adjusted OR = 33, 95% CI 11 – 100; p < .001) or were engaged about in a discussion about pharmacotherapy, which included nicotine replacement or bupropion therapy C (39% vs. 14%; adjusted OR = 3.9, 95% CI 2.5 – 6.3; p < .001). Although the overall percentage of patients who were advised to quit smoking ‘B’ by any clinic staff was not statistically significant, the percentage of patients who were advised to quit smoking by intake clinicians was higher at test sites than at control sites, and this difference was statistically significant (31% vs. 10%; adjusted OR = 5.1, 95% CI 2.7 – 9.7; p < .001). Multi-component, evidence level II. Sixty family physicians from 39 practices in Sydney, Australia participated in a cluster randomised controlled trial conducted by Young et al. (2002) to assess a multifaceted, practice-based intervention to improve smoking cessation advice. Each family practice (representing a cluster) was randomly assigned to receive either the intervention in smoking cessation advice (intervention group) or an intervention of identical format and intensity but about cervical screening (control group). The intervention, which was delivered over a four-month period, was a combination of academic detailing, audit and feedback, resources for family physicians, resources for practices, and resources for patients. Change in family physician behaviour was measured through patient recall of smoking cessation advice during their most recent consultation and medical record audit. Family physician self-report were also used to self-assess competence in smoking cessation advice. Overall 1,241 patients completed follow-up questionnaires (70% response rate among those consented to participate). Smokers were less likely than non-smokers to return the follow-up

63


questionnaire. At post-test, recall of a question about smoking status was higher among patients attending family physicians in the intervention group than in the control group. However, the difference in the size of the change in each group was not significant. Between baseline and post-test, there was a significant increase in smokers’ recall of family physicians’ brief advice to stop smoking in the intervention group. However, the difference in the size of this between group difference did not reach significance, although at post-test, more than half of smokers attending the FPs in the intervention group recalled being given advice to quit smoking (54%, adjusted 95% CI: 39 – 68%). Significant improvements between baseline and post-test were noted from intervention group smokers’ recall of specific components of smoking advice. All outcomes except discussing passive smoking and referral to a smoking cessation clinic showed significant improvement. However, only for the prescription of nicotine replacement patches and gum were these improvements significantly greater than in the control group. The confidence intervals for change in each group did not overlap and interaction terms were highly significant (p = .0056 and p = .0002, respectively). While smokers’ recall of four other behaviours (discussing the health risks of smoking, setting a quit date, providing practical advice, and providing written materials) increased significantly in the intervention group between baseline and post-test and remained significantly higher than in the control group at post-test, these between-group differences did not reach significance. Multi-component, evidence level II. A randomised controlled trial by Lemelin et al. (2001) evaluated the effectiveness of a multifaceted intervention delivered by nurse facilitators and designed to improve preventive care in capitation-based family practices in Canada. The intervention used the recommendations of the Canadian Task Force on Preventive Health Care and included audit and ongoing feedback, consensus building, opinion leaders and networking, academic detailing and education materials, reminder systems, patient-mediated activities, and patient education materials. The unit of randomisation and analysis was the primary care practice (one to six doctors). One of three nurse facilitators delivered the intervention (23 practices), whereas the control group had no intervention (23 practices). The primary study comparison was between the intervention and control practices with respect to changes from baseline (May 1997) to the end of the intervention period (Dec 31, 1998) in the performance of eight recommended and five inappropriate preventive manoeuvres. The data specific to smoking cessation counselling were not statistically significant. Follow-up results for general prevention outcomes showed that the index of preventive performance for both the control and intervention groups at baseline (before the intervention) were similar (32.1% (95% CI 27.2 – 37.0%) and 31.9% (95% CI 27.3 – 36.5%)). The intervention practices performed significantly better than the control practices in preventive manoeuvres as indicated by an 11.5% absolute improvement of the preventive performance index. This was further highlighted by intervention group clinicians performing fewer inappropriate manoeuvres at follow-up than the controls (an absolute improvement on the inappropriateness index of 4.4%). Multi-component, evidence level II. In a large randomised four group trial involving 234 residents from three primary care specialties, Strecher et al. (1991) evaluated the effectiveness of training and prompting under realistic conditions. A simple, feasible teaching programme, the tutorial was trialled alone or in combination with a second intervention which used a chart-based reminder, the prompt. The pragmatic study design used existing staff for training (not smoking cessation experts) and existing nursing staff to administer the prompt with little opportunity for physicians to reinforce counselling. The tutorial comprised two sessions in smoking cessation counselling including a slide show/lecture, video, handouts/flowchart, group discussion, and was delivered by the clinic director or another educator. The prompt was a non-computerised chart-based reminder to assist physicians to provide counselling: a flow-

64


chart placed on patients’ records by clinic nurses at patent check-in. The mean percentage of return patient smokers advised to quit rose by 16% in the tutorial + prompt group and 14% in the tutorial-only groups, but by only 6% and 2% in the prompt-only and control groups respectively. Thus, follow-up rates of brief advice ‘B’ reported by the tutorial + prompt group increased significantly in comparison to the control group (p < .05). Counselling content followed a similar pattern. However, the largest changes in the delivery of cessation support occurred with setting a quit date and prescribing NRT, but little change was observed in actual counselling rates per se. The authors concluded that whether used alone or in combination with the tutorial, the prompt, a second intervention using a chart based reminder, had less effect on residents’ counselling practices. Multi-component, evidence level III-3. Flocke et al. (2006) evaluated a ‘multi-component intervention’ using a relatively small community level pre-test/post-test study design, in the context of a general prevention programme. Approximately 50 primary care physicians from seven primary care practices participated in researcher-facilitated group discussions and practice-feedback and planning (teamwork). The participants were also provided with several office tools, including a web-based health behaviour change resource including a database of community programs and patient education materials, and a health behaviour prescription pad. Typically, ideas for implementing the intervention tools and decisions concerning the intervention plan were made during practice-wide meetings. Following the development of these preliminary ideas, a smaller team of practice members met with the facilitator to refine the plan and finalise decisions about tailored materials and implementation. While the intervention had a significant effect on increasing the rates of discussion of diet, exercise, and weight management, this was not so for the smoking outcomes. The post-intervention cohort did report higher rates of discussion of smoking cessation, but the increase was not statistically significant. Smokers’ reports of their physician having discussed smoking cessation increased from a relatively low 12.4% at baseline to a not statistically significant 14.0% post-intervention (p = .25; OR 1.09, 95% CI: 0.94 – 1.26). Flocke et al. (2006) reported that subsequent evaluation of the qualitative field notes suggested that for six of seven practices, the intervention was not appreciably implemented. In conclusion, Flocke et al. (2006) proposed that the intervention tools did show promise for increasing clinician discussion of health behaviours, information referral, and referral to community resources, but additional work to increase practice adoption of office tools is required. Multi-component, evidence level III-3. Ozer et al. (2001) used a before-and-after design to assess an intervention at three outpatient paediatrics clinics within a large managed care system throughout Northern California. The trial was an attempt to test an approach to implement national guidelines that recommend screening and counselling adolescents on a range of risky behaviours including smoking (general prevention). Eighty-nine clinicians in the three sites agreed to participate in the study and were trained in the intervention (86% were physicians and 14% were nurse practitioners). The goal of the intervention was to increase clinicians’ screening and brief counselling of adolescents in the targeted health risk areas of tobacco, alcohol, sexual behaviour, and safety (seatbelt and helmet use). As part of the intervention, resources of a health educator were provided to the clinics. All clinicians received training in adolescent preventive services, and clinicians used screening and charting forms customised for this study. Study adolescents met with their primary care providers for a ‘well visit’ that lasted 20 to 30 minutes. The overall intervention focused on adolescents taking responsibility for their health and learning ways to keep themselves healthy. Independent adolescent self-reports of clinician screening and counselling practices were obtained from three separate groups of adolescents (mean age of 14 years) who attended ‘well visits’ with their primary care providers. Following a visit, adolescents completed surveys

65


reporting on clinician screening and counselling for each of the target risk areas. Pre-implementation (three months), 104 adolescents completed surveys. Post-implementation of the training, tools, and health educator intervention, 211 adolescents completed surveys (five months). For 18 months post-implementation clinicians delivered services and 998 adolescents completed surveys. The evaluation of the implementation included analysing differences in screening and counselling rates at different periods. Results for screening showed adolescents reported being screened a greater percentage of the time at follow-up than at baseline for tobacco use (61% vs. 95%; p < .0001). Similar results were reported for brief counselling (advice) in tobacco use (55 % vs. 96 %; p < .0001). Multi-component, evidence level IV. Tremblay et al. (2001) reported on two cross-sectional mail surveys conducted as part of the evaluation of a five-year intervention programme to improve smoking cessation and prevention counselling practices of general practitioners (GPs) in Montreal, Quebec. Programme development was guided by Green and Kreuter’s ‘precede-proceed model’ (Green & Kreuter, 1991). The multi-component intervention was complex and designed to act at a number of different levels, from the individual clinician level to the public domain. Briefly, intervention components included training workshops for HCPs, publication of 11 articles about smoking cessation counselling in journals for Quebec physicians, publication of clinical guidelines for smoking cessation, production and distribution of two-sided plastic-coated algorithm card, publication and distribution of a smoking cessation guide presented in a cigarette pack format, publication and distribution of a list of smoking cessation providers in Montreal, public awareness campaigns and publication of two newspaper articles, and two articles in a consumer magazine to reinforce the importance of physicians in smoking cessation and to encourage smokers to ask for help from their physicians. The first survey was conducted between April 1, 1998, and July 31, 1998, in a simple random sample of 670 physicians selected from the 1997 Quebec College of Physicians database for Montreal. Eligible physicians included those holding a GP license from the Quebec College of Physicians and working in any clinical setting in Montreal. Three mailings of a self-administered French or English questionnaire with follow-up telephone calls to all non-respondents were undertaken to maximise participation. The follow-up survey was conducted between April and September 2000. This paper describes the theoretical model underlying the Physicians Taking Action Against Smoking program, as well as selected components of the program implemented during its first three years, as well as reporting on the results of the two cross-sectional surveys of independent random samples of GPs in 1998 and 2000, to monitor cessation counselling practices. Data indicated positive changes in the smoking cessation counselling practices of GPs in terms of using a system to identify smokers ‘Ask’ and cessation counselling. This includes the following GP activities: offers written educational materials (female GPs), discusses patients’ perception of pros and cons of smoking (male GPs), devotes two minutes to cessation counselling and discusses setting a quit date (male and female GPs), and refers patients to community resources (male GPs), all statistically significant differences between 1998 and 2000 (.001 < p < .047) and on a number of outcomes in 2000, between male and female physicians (p < .005). There were more positive changes in psychosocial aspects of counselling among female GPs over time than among male GPs. A subsequent paper by O’Loughlin, Makni, Tremblay and Karp (2007) (excluded from this review) reports data from a comprehensive re-analysis of the survey data for gender differences. The reader is directed to O’Loughlin and colleagues’ (2007) original paper for an in-depth discussion on gender differences and possible implications for the design of interventions aimed at changing GPs’ counselling practices. Results from the study demonstrated that a comprehensive strategy including a program to improve physicians’ counselling practices could be effective in reducing tobacco use.

66


However, the absence of a control group precludes attribution of the observed changes to the programme. There were some social/environmental changes that took place during the study period that may have influenced HCPs and/or patient behaviours, including the initiation of funding for NRT, mass media campaigns (possibly leading to patients prompting HCPs), and a law change to stop smoking in public places. It is difficult to know how these might have influenced the results. Also, HCPs’ self-perceptions could have been subject to recall bias and social desirability bias. The inclusion of a chart audit for the ‘Ask’ outcome rather than self-reports would have strengthened these data. However, the authors indicated that even though the implementation of the Physicians Taking Action Against Smoking program was still in progress, their experience so far suggested that it is feasible to implement a multi-dimensional program aimed at improving the smoking cessation counselling practices of HCPs, with limited resources. Multi-component: summary Table 18 summarises the evidence pertaining to multi-component interventions that is directly relevant to the current review, in accordance with the NHMRC dimensions of evidence. Broadly, the eight studies all assessed the impact of combining multiple approaches within one programme, with the overall aim being to encourage primary care clinicians to deliver smoking cessation interventions to their patients. Five of eight studies (62.5%) (Katz, et al., 2004; Pbert, et al., 2006; Strecher, et al., 1991; Tremblay, et al., 2001; Young, et al., 2002) were conducted within the context of smoking specific programmes and the remaining three studies (37.5%) (Flocke, et al., 2006; Lemelin, et al., 2001; Ozer, et al., 2001) reported smoking outcomes within the context of general prevention programmes. In addition to training health care professionals to deliver smoking cessation interventions (common to all identified multi-component interventions), the following intervention components were identified: reminders, algorithms, a health behaviour prescription pad, screening and charting forms, audit & feedback, education materials, resources for physicians, resources for practices, resources for patients, offer of free nicotine replacement therapy, availability of proactive telephone counselling, availability of a patient health educator, consensus building, opinion leaders and networking, patient-mediated activities, teamwork, a web-based database of community programs, publication of journal articles, publication of clinical guidelines for smoking cessation, public awareness campaigns, newspaper and magazine articles. Smoking specific studies In all cases, the studies of multi-component interventions as implemented in smoking specific contexts demonstrated positive effects on primary care clinicians’ delivery of smoking cessation interventions. Four of the five smoking-specific studies were randomised, and all demonstrated statistically significant increases in the assessment of smoking status and the delivery of brief advice. Additionally, four of five studies demonstrated statistically significant increases in the delivery of more intensive cessation counselling and/or cessation support. The one non-randomised trial (Tremblay, et al., 2001) (evidence level IV) evaluated a five-year intervention programme aimed at improving the smoking cessation and prevention counselling practices of GPs in Montreal. The intervention included a wide range of components, some directed at the clinician level, but also included patient-mediated interventions and other health promotion initiatives. Tremblay et al. (2001) reported statistically significant increases on all the smoking process outcomes, although the authors acknowledge the limitations of the study design and that the data were subject to several possible biases. Nevertheless, this very comprehensive strategy did demonstrate promise and more studies of this type may be warranted.

67


General prevention studies In contrast to the five studies conducted in smoking specific contexts, the three general prevention multi-component programmes were largely ineffective in increasing primary care clinicians’ delivery of smoking cessation interventions. However, all three studies reported positive results for the other outcomes measured (including screening, brief advice and counselling for diet, exercise, alcohol, sexual behaviour, mammography, Papanicolaou smear, influenza vaccination, hypertension treatment and seatbelt and helmet use). For example, Lemelin, et al. (2001) adopted a guidelines-based approach with seven intervention components that targeted eight preventive manoeuvres and five inappropriate manoeuvres. The authors reported a statistically significant 11% increase in the general prevention index (based on mean scores for the eight preventive and five inappropriate manoeuvres) but no significant increase for any smoking outcome. Similarly, Flocke, et al. (2006) trialled a six-component intervention including office systems. While poor implementation may have attenuated the overall results, the diet, exercise and weight management counselling outcomes all showed statistically significant improvements. No improvements were observed for any smoking outcomes– despite only 12.4% of patients initially reporting any discussion about smoking at baseline. Finally, Ozer, et al. (2001) combined training with office tools and a health educator. While the general prevention outcomes of screening, brief advice and counselling for alcohol use, risky sexual behaviour, and safety all showed statistically significant increases, for smoking, only the rates of screening and brief advice increased significantly. Smoking cessation counselling did not increase despite a dedicated health educator being available to support patient referrals. Multi-component: conclusions Taken together, these studies suggest that multi-component interventions, based on training, are effective in increasing the frequency with which primary care clinicians intervene with their patients who smoke. However, no clear patterns emerged pertaining to which combinations of intervention components were more or less effective or what relative contributions the individual components might play. Broadly, the more complex multi-component interventions did not appear notably more effective than the less complex two or three component interventions. Training, audit & feedback and reminders were the most commonly combined intervention components – these appeared to be effective. No multi-component intervention included financial incentives. General prevention programmes performed rather less well than smoking specific programmes. It appears that in general prevention programmes, the intervention-effect, on any one behaviour, may be diluted to a much greater extent with respect to smoking outcomes.

68


Table 18: Original study results: The effectiveness of Multi-component interventions to increase the delivery of effective smoking cessation treatments in primary care settings

MULTI-COMPONENT Multi-component: Level II evidence Pbert et al. (2006) – Training and reminder, smoking specific.

Proportion % ∆Outcomes UC SI

Chi square X2

p value

Everyone: Asked if ever smoked or tried smoking ‘A’ Discussed smoking ‘C’

77.0 62.4

95.8 91.4

181.87 319.46

p < .0005 p < .0005

Current smokers: Asked what you like about smoking ‘A’ Asked what you do not like about smoking ‘A’ Supported your concerns about smoking ‘C’ Stated single not smoking is single best thing you can do to protect your health ‘B’ Asked if your were interested in quitting ‘C’

24.1 22.9 57.8 72.0 58.5

74.2 68.8 79.6 84.4 86.7

44.03 37.12 9.74 3.97 17.33

p < .0005 p < .0005 p = .002 p < .05 p < .0005

Non-smokers / former smokers: Told you it is great you are not smoking ‘B’ Stated not smoking is single best thing you can do to protect your health ‘B’ Asked you reasons for not smoking ‘A’ Shared additional reasons not to smoke ‘C’ Talked about nicotine addiction ‘C’ Asked if you have friends who smoke ‘A’ Talked about addiction in your friends who smoke ‘C’

92.0 59.5 20.9 35.6 13.0 58.9 19.3

97.3 93.0 76.7 71.6 60.6 86.6 65.6

28.93 320.19 580.56 244.31 430.60 55.96 122.98

p < .0005 p < .0005 p < .0005 p < .0005 p < .0005 p < .0005 p < .0005

∆ From patient exit interviews, only statistically significant results are displayed. UC = Usual care SI = Special intervention

69


Table 18: Original study results: The effectiveness of Multi-component interventions to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

MULTI-COMPONENT Multi-component: Level II evidence Katz et al. (2004) – Multi-component, five-components including tutorials, feedback, a modified vital signs stamp, free NRT, smoking specific. Outcomes Control Intervention

p value Odds ratio (95% CI)

Ask “A” patients asked about smoking status

67% 87% p < .001 3.1, (1.2 to 8.2)

Advice “B” % of patients who were advised to quit smoking during the intervention by intake clinicians

10%

31%

p < .001

5.1, (2.7 to 9.7)

Assisted “C” (any) willingness to quit smoking given literature about quitting assisted with setting a quit date engaged about in a discussion about pharmacotherapy

30% 3% 1% 14%

73% 38% 27% 39%

p < .001 p < .001 p < .001 p < .001

6.4, (3.7 to 10.8) 21, (8.8 to 49) 33, (11 to 100) 3.9, (2.5 to 6.3)

Abstinence rates: At 2-month follow-up assessment At 6-month follow-up assessment Continuous abstinence at both 2 and 6 months

5.8% 9.8% 3.8%;

16.4% 15.4% 10.9%

p < .001 p = .009 p < .001

3.3, (1.9 to 5.6) 1.7, (1.2 to 2.6) 3.4, (1.8 to 6.3)

CI (confidence interval)

70



MULTI-COMPONENT Multi-component: Level II evidence Young et al. (2002) – Multi-component, audit and feedback plus resources for family physicians, resources for practices, and resources for patients, smoking specific.

Change in smoking cessation advice: post-test vs. baseline ∆Outcomes Intervention Control ICCF

Component of smoking cessation advice OR (95% CI) OR (95% CI) All patients Ask smoking status ‘A’a Smokers Advise smoker to quit ‘B’b

Discuss health risks of smoking ‘C’ Provide practical advice ‘C’ Set a “quit date” ‘C’ Provide written materials ‘C’ Recommend nicotine gum ‘C’c

Recommend nicotine patches ‘C’d

Arrange a follow-up appointment ‘C’ Document patient’s smoking status ‘C’e Document smoking cessation advice C

1.74 (1.31-2.31) 1.92 (1.06-3.49) 2.60 (1.43-4.74) 2.81 (1.46-5.41) 4.96 (1.85-13.32) 6.49 (2.51-16.76) 5.31 (2.68-10.51) 2.70 (1.44-4.40) 4.85 (1.36-17.3) 2.47 (1.75-3.50) 3.33 (1.39-7.97)

1.67 (1.60-1.75) 1.76 (0.78-3.98) 1.73 (0.80-3.72) 1.19 (0.55-2.58) 2.28 (0.94-5.52) 1.72 (0.65-4.52) 0.52 (0.22-1.24) 0.66 (0.32-1.38) 3.44 (0.74-2.77) 2.88 (1.18-7.04) 1.40 (0.41-4.79)

0.04 0.009 0.03 -0.0003 0.04 0.01 0.009 -0.01 -0.004 0.2 -0.006

∆Multivariate assessment of changes in smoking cessation advice, adjusted for clustering of patients within practices. aAdjusted for patient’s sex, smoking status, and new patient status. bAdjusted for patient’s age and stage of change for smoking cessation. cp for time x group interaction term = 0.0002. dp for time x group interaction term = 0.0056. eAdjusted for patient’s smoking status. FICC = intra-cluster correlation coefficients.

71



MULTI-COMPONENT Multi-component: Level II evidence Young et al. (2002) – Multi-component, audit and feedback plus resources for family physicians, resources for practices, and resources for patients, smoking specific.

Intervention Control p value

Outcomes Unadjusted (crude) rates of smoking cessation advice at baseline and post-test by group n/N (%) n/N (%) Patients’ recall of a question about their smoking status ‘A’a All patients Baseline Post-test Smokers Baseline Post-test Smokers’ recall of specific smoking cessation advice: Discuss health risks of smoking ‘C’ Baseline Post-test Provide practical advice ‘C’ Baseline Post-test Provide written materials ‘C’ Baseline Post-test Recommend nicotine gum ‘C’ Baseline Post-test

Recommend nicotine patches ‘C’ Baseline Post-test

Medical record audit Document of smoking status (all patients) Baseline Post-test

136/359 (24%) 149/462 (32%) 47/117 (40%) 38/71 (54%) 37/119 (31%) 39/72 (54%) 18/119 (15%) 24/72 (33%) 9/119 (8%) 25/72 (35%) 8/119 (7%) 20/72 (28%) 14/119 (12%) 18/72 (25%) 159/544 (29%) 211/440 (48%)

97/665 (14%) 162/669 (24%) 26/100 (26%) 53/120 (44%) 21/102 (21%) 39/124 (31%) 12/102 (12%) 17/124(12%) 7/102 (7%) 14/124 (11%) 9/102 (9%) 6/124 (5%) 13/102 (13%) 11/120 (9%) 67/579 (12%) 182/583 (31%)

p < .001 p = .003 p = .03 p = .08 p = .002 p = .5 p = .001 p = .8 p < .001 p = .6 p < .001 p = .8 p = .002 p < .001 p < .001

a n, Patients (n =5 at baseline and n =6 at post-test) attending specifically for smoking cessation advice were excluded from analyses to assess rates of family physician assessment of smoking status.

72



MULTI-COMPONENT Multi-component: Level II evidence Lemelin et al. (2001) – Multi-component, included audit and feedback, consensus building, opinion leaders and networking, academic detailing and education materials, reminder systems, patient-mediated activities, and patient education materials, general prevention study measuring the performance of eight recommended and five inappropriate preventive manoeuvres. *Outcomes: Smoking Via chart review

Control (%, 95% CI)

Intervention (%, 95% CI)

†Absolute improvement

p value

Assisted ‘C’ (any) Before

After Change Before

After Change

Preventive manoeuvres*

40.5

38.7 -18 37.6

41.2

36

7.2%

–

Intervention *Outcomes: Smoking Via exit survey and chart review Before (%) After (%) Change (%) From Patient survey From Chart reviews Absolute Difference

64.0 37.6 26.4

73.0 41.2 31.8

9.0 3.6 5.4

*Specific to smoking cessation counselling as determined from chart review (Grades A and B) including smoking cessation counselling (performance of smoking cessation counselling), and preventive manoeuvres performed in the intervention and control practices before and after the intervention, as determined by patient survey and/or chat review. Grade A, there is excellent evidence from repeated RCT trials to support the manoeuvre, Grade B, there is good evidence from cohort and case-control studies to support the manoeuvres. †Difference in change between intervention and control practices §Outcomes General prevention

Control (%, 95% CI)

Intervention (%, 95% CI)

†Absolute improvement

p value

Before After Change Before After Change Preventive performance index

32.1 (27.2-37.0)

31.9 (26.8-37.0)

-0.2 31.9 (27.3-36.5)

43.2 (38.4-48.0)

11.3 11.5% p < .001

Up-to-datedness index

54.6 (51.0-58.2)%

57.4 (54.1-60.7)

2.8 52.3 (48.6-56)%

62.3 (58.2-66.4)

10.0 7.2% p = .008

Inappropriate manoeuvres (inappropriateness index)

22.5 (NR)

25.5 (20.0-31.0)

3.0 20.5 n.r.

19.1 (15.6-22.6)

-1.4 4.4% p = .019

§General prevention: mean proportion of eligible patients who received the recommended manoeuvres (including “but not specific to” smoking cessation counselling), †Difference in change between intervention and control practices.

73



MULTI-COMPONENT Multi-component: Level II evidence Strecher et al. (1991) – Multi-component, training/prompt, smoking specific. *Outcomes Control Tutorial Prompt Tutorial & prompt Ask ‘A’ Assumed 100%– clinic nurses identified patients who

smoked Not measured

Pre-test 67% 64% 70% 63% †Advice ‘B’ Post-test 69% § 78% 76% § 79%

Pre-test 0.7 0.6 0.5 0.5 ‡Cessation ‘C’ Post-test 0.9 ∆1.5 0.9 ∆1.9 Quit rates ‘Q’ 1.7% 3.4% 5.7% 5.6% All non-significant *Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. †The percentage of smokers provided with brief advice, as measured by researcher administered patient exit interview. ‡The mean number of techniques used. The occurrence of five specific cessation support techniques used by physicians: setting a quit date, prescribing a quit date, prescribing NRT, giving self-help material, providing follow-up, as measured by blinded, researcher administered telephone interview. ‘Q’ Patient reported smoking status six months after exit interview (telephone follow-up, n = 937). §Statistically significant difference, compared to the control group, at the p < .05 level. ∆Statistically significant difference, all Tutorial compared to non-Tutorial, at the p < .001 level.

74



MULTI-COMPONENT Multi-component: Level III-3 evidence Flocke et al. (2006) – Multi-component, group discussion and practice feedback and planning, office tools, including a web-based health behaviour change resource including a database of community programs and patient education materials and a health behaviour prescription pad, general prevention. *†‡Outcomes Baseline Follow-up p value Odds ratio (95% CI) Ask ‘A’ – – Not reported Advice ‘B’ 12.4% 14% p = .25 1.09 (0.94-1.26) Cessation ‘C’ (any) – – Not reported *Further details of patient outcomes are available in the original authors’ publication. †Rates of discussion of smoking cessation as reported at the baseline pre-visit patient survey, compared to the patient follow-up phone survey (n = 784) conducted approximately eight weeks after the index visit. ‡Smoking outcomes reported within the context of a ‘general prevention’ trial of preventive services: advice to stop smoking, diet, exercise, weight management. Ozer et al. (2001) – Multi-component, training, customized screening and charting forms, general prevention, four preventive counselling procedures. *Outcomes Pre-implementation

(T1) Post-implementation (T2)

Follow-up (T3)

p value (T1-T2) and (T1-T3)

Chi square (T1-T2) and (T1-T3)

Ask ‘A’ Screening rates

61%

95%

90%

p < .001

61.1 75.6

Advice ‘B’ Counselling Rates Clinicians’ brief counselling

55%

96%

94%

p < .001

70.8 134.8

Assisted ‘C’ (any)

Not reported

*An example of a screening question: “Did your doctor ask if you smoke or chew tobacco?” ‘A’. An example of a counselling question for adolescents who were not engaging: “Did your doctor encourage you to remain a non-smoker or non-tobacco user?” ‘B’

75



MULTI-COMPONENT Multi-component: Level IV evidence Tremblay et al. (2001) – Multi-component, smoking cessation counselling practices of Montreal GPs, as determined by surveys in 1998 and 2000.

Year; % (and no.) of male physicians

Year, % (and no.) of female physicians Outcomes 1998

n = 204 2000 n = 184

p value†

1998 n = 133

2000 n = 132

† p value

‡§Ask ‘A’ Usually uses a system to identify smokers.

42.8 (86)

37.6 (67)

0.31

56.1 (73)

51.9 (68)

p = .49

Advice ‘B’ Not specifically reported for brief advice only. Assisted ‘C’ (any) §Offers cessation counselling to more than half of patients who smoke.

55.9 (95) 61.5 (112) 0.28 65.5 (74) 74.6 (97) p = .12

Devotes 2 min or more to cessation counselling when counselling is offered.

77.4 (106) 87.6 (148) 0.018 87.2 (83) 90.2 (110) p = .49

*‡Cessation support ‘C’ Discusses patients’ perception of pros and cons of smoking.

68.7 (92) 52.7 (87) 0.005 55.9 (57) 54.6 (66) p = .84

*§Offers written educational material. 35.6 (47) 28.1 (46) 0.16 30.7 (31) 47.2 (58) p = .012 Contemplation stage *§Offers written educational material.

44.0 (70) 46.7 (78) 0.63 55.0 (61) 73.2 (90) p = .004

Preparation stage *§Discusses strategies for quitting.

85.0 (136) 78.9 (127) 0.15 86.7 (98) 90.9 (110) p = .31

*§Discusses setting a quit date.

58.2 (92) 69.4 (111) 0.039 67.0 (75) 85.8 (103) p = .001

*§Offers a follow-up visit

45.0 (72) 41.9 (67) 0.57 56.8 (63) 55.8 (67) p = .89

*‡§Offers written educational material

36.3 (58) 36.3 (58) 1.00 51.3 (58) 64.2 (77) p = .047

*§Refers patient to community resources

18.1 (29) 9.5 (15) 0.02 22.7 (25) 22.3 (27) p = .94

*Offers support to more than half of patients at various stages of smoking cessation. †For difference in results between 1998 and 2000, calculated using a χ2 test. ‡Statistically significant difference in 1998 between male and female physicians (p < .05). §Statistically significant difference in 2000 between male and female physicians (p < .05).

76


77


Reminders Introduction Reminder interventions were those that introduced manual or computerised systems to automatically prompt health care providers to perform specified smoking cessation activities. These include concurrent or inter-visit reminders to professionals about desired smoking cessation actions, such as providing brief advice and offering cessation support. Five of the 37 included studies (13.5%) evaluated reminder systems: four randomised trials and one non-randomised before-and-after design. In three studies (Fiore, et al., 1995; Piper, et al., 2003; Rothemich, et al., 2008) the intervention was a reminder system that used some form of expanded vital sign stamp of expanded patient progress note-pages to prompt clinicians at the point of care. In the other two studies (Krist, et al., 2008; McPhee, et al., 1991), prompts were delivered to clinicians automatically via computer. One of these computer-based interventions (McPhee, et al., 1991) was evaluated within a general cancer prevention programme and the prompts were generated in accord with pre-determined schedules. In the other study of a computer-based intervention, and all of the non-electronic reminder interventions, the prompts were generated by nurses, medical assistants, and/or administrative staff who were instructed to assess the tobacco use status of every patient and record it, thus initiating the prompt (via ‘paper’ or electronic means). The underlying goal of all of the reminder interventions was to increase the delivery of smoking cessation counselling in primary care practices. Reminders: results from five studies Non-electronic, evidence level II. Rothemich et al. (2008) conducted a moderately sized cluster-randomised, controlled trial to assess the effect of a vital sign intervention on patient-reported frequency and intensity of tobacco cessation counselling. Eighteen primary care clinics with 73 clinicians were divided into five matched groups and randomised within the matched groups to the intervention or control condition. The practices included 14 family medicine practices, three internal medicine practices, and one practice featuring both specialties. Two practices were urban, two were rural, and 14 were suburban. Only one practice used electronic medical records. Within the intervention group, nurses and medical assistants were instructed to assess the tobacco use status of every adult patient and record it with the traditional vital signs. The intervention group received a one-hour training session three-week prior to the launch of the intervention. This involved all nurses and medical assistants responsible for processing arriving patients, escorting them to examination rooms (rooming staff), and most office managers and some practice medical directors and physicians. Vital signs rubber stamps were produced as well as a description of the study and instructions to modify the procedures by (1) stamping the encounter/progress notes where the usual vital signs were normally recorded, (2) asking every adult patient at every visit whether they used tobacco, and (3) recording the answer within the stamped imprint. The intervention did not address the counselling practices of staff or clinicians. The control practices did not use any systematic tobacco screening or identification system. The comparison period lasted six months (February through August 2004) and targeted a new cross-sectional sample of patients. The proportion of smokers reporting clinician counselling activity of any kind and the frequency of simple quit advice ‘B’ and more intensive discussion ‘C’ were all assessed from the 18 primary care practices via patient exit questionnaires, during the six-month comparison period. Results from the study showed that for all patients there was an improvement in ‘Ask’ rates from 26.3% to 66%, and for smokers from 49% to 79% (control vs. intervention, p < .001).

78


The vital sign intervention improved the frequency of smoking counselling in primary care, this is shown from answers to the following two questions: “If you smoke, did your doctor advise you today to stop smoking?” (brief advice) “If you smoke, did your doctor talk with you today to help you quit smoking?”(cessation support) Among 561 smokers attending the intervention practices, 61.9% reported receiving counselling, compared with 53.4% of the 588 smokers at control practices, for a difference of 8.6% (p = .04). The effect was largely restricted to simple advice, which was reported by 59.9% of intervention patients and 51.5% of control patients (p = .04). There was no significant increase in more extensive discussion ‘C’, with 32.5% and 29.3% of patients at intervention and control practices, respectively, reporting this type of counselling (p = .18). The authors commented that the vital sign intervention improves the frequency of smoking cessation counselling in primary care practices. And they observed an 8.6% proportional increase in counselling that consisted mainly of simple advice to quit but with little additional discussion of how to do so. Six months after the study ended, four of the nine intervention sites reported continuing the vital sign assessment. Electronic, evidence level II. McPhee et al. (1991) conducted a randomised controlled trial to test the efficacy of a computerised reminder system, supplemented by cancer education materials, in promoting 11 cancer prevention activities, including screening of smoking status and smoking cessation counselling. Forty full-time primary care physicians in small community-based practices in five northern California counties were randomised to either the intervention or the control group (one physician per practice). The intervention consisted of computer-generated cancer prevention reminders for overdue screening tests and smoking and dietary assessment and counselling, supplemented by cancer education materials. The prompts were generated in accord with pre-determined prevention schedules. The control group received no intervention elements. The mean performance scores of the intervention group physicians were significantly higher than controls for smoking assessment (45.0% vs. 32.4%, p = .014) and counselling (58.8% vs. 41.8%, p = .027) respectively. The authors concluded that the intervention appeared to address the physician barriers of forgetfulness, time constraints, and logistical difficulties. Further, the effectiveness of the reminders suggests that physician forgetfulness may account for a considerable part of the gap between cancer prevention recommendations and actual clinical performance. Non-electronic, evidence level III-2. Piper et al. (2003) in a randomised controlled trial examined the ability of an expanded vital sign stamp to increase rates of smoker identification, physician advice to quit smoking, physician assistance in quitting, and abstinence rates. In this study, five primary health care clinics were randomly assigned to either the intervention group (received the vital sign stamp) or the control group. Exit interviews were used to survey 9439 patients at the clinics both before and after the vital sign intervention was implemented. Patients who were identified as smokers were then contacted one year later for follow-up, the study began in February 1995, and all follow-up visits were completed by December 1998. There was an increase in the proportion of all patients who were asked by any clinic staff member (physician, nurse, medical assistant) about smoking ‘A’ both before and after experimental manipulation of the vital sign stamp. Results showed an overall increase of 9.6% and 30.9% in the control and intervention clinics respectively. The mean increase in asking behaviour of physicians at the vital signs clinics was statistically significantly greater than that of physicians at control clinics (mean 13.76; SD 7.5 vs. mean 2.89; SD 12.9; p = .002). The findings for rates of advice to quit, assistance with quitting, and abstinence rates were either unchanged or decreased. Although the vital sign stamp changed the physicians’ behaviour this

79


change resulted in rates of smoker identification of no higher than 66%, suggesting that even with this assessment tool physicians were identifying only two thirds of all smokers. Electronic, evidence level III-3. One very recent before-and-after study (evidence level III) used a cohort analysis to assess the impact of introducing an electronic system prompt (Krist, et al., 2008). The study looked at an electronic linkage system which prompts clinicians to offer behaviour counselling and referral to community services to help improve preventable health risks including smoking cessation. The study was conducted in nine primary care practices in the Virginia Commonwealth University Department of Family Medicine. The practices incorporated an Electronic Linkage System (eLinkS) to prompt clinicians at the point of care to offer behaviour counselling and then electronically refer patients to the appropriate community services that can help them improve preventable health risks. These risks were specifically related to diet, exercise, smoking, and alcohol consumption. In brief, it is an Ask, Advise, and electronic referral/e-mail approach in which a member of the health care team could send an e-mail to the patient with links to internet-based smoking cessation programmes or offer to provide an ongoing e-mail support, or refer the patient directly to an external speciality service provider. Using an existing electronic medical record system (EMR) as a platform, eLinkS was designed to help clinicians systematically perform elements of the 5As that are feasible in busy practice settings, to make it fast and easy to refer patients to intensive counselling outside the office, and to establish two-way communication between practices and community counsellors. Underpinning the programme is the concept of proactive rather than reactive counselling. Rooming staff take the patient’s vital signs, measure height and weight, and inquire about smoking status (‘Ask’) and the staff member enters this information into the EMR. When the information in the EMR reveals the presence of an unhealthy behaviour, the EMR automatically prompts the clinician to consider providing counselling and referral services. When telephone or group counselling is selected through eLinkS, the EMR automatically e-mailed contact information to the community counselling program staff, who then would contact the patient (‘proactively’, rather than having the patient contact the counselling staff , ‘reactive counselling’). When clinicians selected computer care, the EMR forwarded an e-mail to the patient with a link to the educational website and instructions for e-counselling. Training sessions for eLinkS of one-hour duration were held at all practices prior to launch. The intervention went live on April 16, 2006, and was discontinued five weeks later (May 22, 2006) when an overwhelming surge in referrals for intensive counselling exceeded available funds. Patients referred to intensive counselling during the exposure period were eligible to receive services for up to nine months (through February 2007). Weekly feedback reports notified participating clinicians of the number of counselling referrals they had placed. Findings from the pre- and post-implementation analysis indicated that the automated prompts and referrals successfully generated patient referrals for counselling services, leading to improved behaviours related to diet and exercise (which in turn led to weight loss), and enhanced quit rates among smokers. Depending on the type of smoking cessation intervention, 50 to 75% of patients reported no longer smoking at four months after referral, compared to 25 percent of patients receiving usual care. However, the authors stated that the participating clinicians were not necessarily representative of primary care clinicians generally, because a large percentage (44%) were family medicine residents, a group that is younger and comprises more women than non-residents. Intensive counselling was free to patients, eliminating cost as a barrier, and this may have enhanced the referral rate. The authors concluded that the intervention increased the rate at which patients were referred for intensive behavioral counseling compared to current practice norms (8% post intervention vs. 1%-5%). Given the evidence that intensive counseling is more effective in promoting behavior change than less intensive counseling, implementing eLinkS could have substantial public health benefits.

80


Non-electronic, evidence level III-3. By implementing a similar intervention to that of Piper et al. (2003) Fiore et al. (1995), examined the ability of an expanded vital sign stamp to increase rates of smoker identification, physician advice to quit smoking, and physician assistance in quitting. Using a before-and-after study design, the intervention was evaluated with 47 physicians and five nurse practitioners and 28,000 patients made 50,000 visits per year to the General Internal Medicine Clinics (GIMC) at the University of Wisconsin Hospital during the 16 month data collection period. Compared to baseline, the rate of clinicians asking about smoking doubled from 25.5% to 52.6% (p < .0001) at follow-up. There was also a significant increase in the percentage of smokers who reported that their clinician advised them that day to quit smoking ‘B’ (from approximately 49% at baseline to 70% at follow-up, p < .01) and in the percentage who reported that their clinician gave them specific advice that day on how to stop smoking ‘C’ (24% at baseline vs. 43% at follow-up, p < .01). In this study, expansion of the vital sign stamp to include smoking status was associated with significantly increased rates of identification of patients who smoke and significantly increased rates of intervention to inquire about and discourage cigarette smoking. However, the study used a before and after design, was conducted within a single clinic with a moderately small sample, and used patient self-report as the main data collection source. As such, the data are likely to be subject to confounding and probably limited generalisability. Reminders: summary Table 19 summarises the evidence pertaining to reminders that is directly relevant to the current review, in accordance with the NHMRC dimensions of evidence. Broadly, the five studies of reminder systems all aimed to assess the impact of prompting primary care clinicians to deliver smoking cessation interventions. Taken together, these studies suggest that reminder systems are effective in increasing the frequency with which primary care clinicians intervene with their patients who smoke. All of the studies demonstrated statistically significant increases in one or more of the ‘ABC’ outcomes, with three of five studies (60%) (Fiore, et al., 1995; Krist, et al., 2008; McPhee, et al., 1991) demonstrating statistically significant increases in the incidence of smoking cessation counselling. The other two studies reported either increased or unchanged counselling activities. All of the programmes involved sharing the responsibility for office-based smoking cessation activities among all staff. In these studies (with the possible exception of Krist, using scheduled computer prompts), the effectiveness of the programmes in increasing clinicians’ advice-giving rates, counselling, and ultimately patients’ quit rates, appears to be contingent on the level of compliance by ‘other’ staff with regards to the screening and prompting tasks. That is, the subsequent notation of patients’ smoking status and generation of the ‘prompt’. It was not clear what degree of compliance with these screening and prompting tasks might actually be achieved over time. Electronic reminder systems appear particularly promising. Reminders: conclusions In all cases, reminders increased the frequency of health professional intervention. Programmes were found to be generally effective in changing health care professionals’ advice-giving rates and counselling activities. Sharing the responsibility for office-based smoking cessation activities among all staff appears to be a promising strategy, and addressing the barrier of clinician ‘forgetfulness’ is one probable core constituent. Note that, in part, the relatively small number of included studies of reminder systems reflects the exclusion criteria used for this review. Several other studies of reminder systems have been undertaken prior to 1990 and/or report patient rather than provider outcomes. A number of studies evaluating the effectiveness of reminders in various settings and contexts are reviewed in Grimshaw et al. (2004) and Dicky et al. (1999), and several other original studies, for

81


example (Chang, Zimmerman, & Beck, 1995; Cohen, Stookey, Katz, Drook, & Smith, 1989; Cummings, Coates et al., 1989; Cummings, Richard et al., 1989) also provide additional information. The reader may wish to review these publications independently. Table 19: Original study results: The effectiveness of Reminder systems

to increase the delivery of effective smoking cessation treatments in primary care settings

REMINDERS Reminders: Level II evidence Rothemich et al. (2008) – Reminders (Vital sign), smoking specific. *Outcomes Control Intervention p value Odds ratio (95% CI) Ask ‘A’ Did a nurse or doctor ask you today if you smoke Ask ‘A’ all patients Ask ‘A’ smokers only

26, 3% 49.4%

66.0% 79.5%

p < .001 p < .001

OR not reported but 95% CI 35.5 to α OR not reported but 95% CI 24.0 to α

Advice ‘B’ “If you smoke, did your doctor advise you today to stop smoking?”

53.4%

61.9%

p = .04

OR not reported but 95% CI -2.9 to α Not significant

Assisted ‘C’ (any) 29.3% 32.5% p = .18

OR not reported but 95% CI 0.8 to α

*Patients response to exit questionnaires during a 6-month comparison period McPhee et al. (1991) – Reminders, computerised, general prevention, 11 cancer prevention activities. *Outcomes Control Intervention p value †Ask ‘A’ 32.4% (SD 13.9) 45.0% (SD 16.6) p = .014 ‡Advice ‘B/C’ 41.8% (SD 22.2) 58.8% (SD 23.0) p = .027 *Only smoking specific outcomes are reported, details of the other outcomes are available in the original publication. †Performance scores as percentage compliance with clinical guidelines. ‘A’, percentage of patients each physician assessed for smoking status during each 12 month period (pre/post intervention). ‡‘B/C’, the percentage of smokers (as assessed by physician) who had been counselled during one year. The exact level of counselling, beyond providing brief advice, was not clearly defined. Data extracted from a random sample of 60 medical records per physician, for 12-month pre-intervention and post-intervention.

82


Table 19: Original study results: The effectiveness of Reminder systems to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

REMINDERS Reminders: Level II evidence Piper et al. (2003) – Reminders, expanded vital sign stamp, smoking specific. *∆Outcomes

Control clinics Intervention clinics p value Odds ratio (95% CI)

Ask ‘A’. Overall increase in proportion of all participants asked Mean increase in asking behaviour of Physicians

9.6% 2.89% (SD 12.9)

30.9% 13.76% (SD 7.5)

Not reported p = .002

t44=-3.61

Advice ‘B’. Physicians advising identified smokers to quit (overall) Baseline Follow-up

56.9% 52.1%

60% 37.1%

Not reported

Assisted ‘C’ (any). Current smokers given help by a physician to set a quit date (overall) Baseline Follow-up Current smokers prescribed NRT by a physician Baseline Follow-up Current smokers whose physician arranged follow-up care Baseline Follow-up

6.8% 2.4% 9.3% 3.3% 8.9% 2.4%

4.4% 1.5% 8.5% 1.9% 4.9% 0.6%

Not reported

†Abstinence rates Baseline Follow-up

52.6% 54.1%

54.1% 50.4%

Not reported

∆ Proportion of all participants asked by any clinic staff person (physician, nurse, medical assistant) about smoking both before and after experimental manipulation of the vital sign stamp ‘A’. Changes in physician advice ‘B’, assistance and arranging follow-up visit ‘C’ (values are percentages). *Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. †Percentage of identified smokers who made a serious quit attempt NRT– nicotine replacement therapy

83


Table 19: Original study results: The effectiveness of Reminder systems to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

REMINDERS Reminders: Level II evidence Krist et al. (2008) – Reminders, Electronic Linkage System (eLinkS) prompt and referral, general prevention, diet, exercise, smoking, and alcohol consumption. *Outcomes Before After p value Odds ratio (95% CI) Current smoking a 922 (16.2%) ‘Ask’ Both BMI and tobacco use assessed

2117 (37.2%) b

‘Advice’ Clinicians advised patients

20%

Assisted ‘C’ (any) Referred for intensive counselling

0%**

8%

Not reported

Not reported

*Before and after launch of the eLINKS (number of patients 5679). Delivery of 5As measured by clinicians’ entries in the electronic medical record (EMR) dialogue box. The 5As are: Assess, Advise, Agree, Assist, and Arrange. aThe number refers to the risk factors as documented in the EMR. bRooming staff collected both BMI and tobacco-use information for 2117 patients who visited the practice. **Post-test, the rate of use of group classes and telephone counselling for smoking cessation as coordinated by the provider was essentially 0%, as these resources were not available for patients for free prior to the study. Fiore et al. (1995) – Reminders, Vital sign, smoking specific. *Outcomes Baseline Post Intervention p value

Chi square

Ask ‘A’ Rate of clinicians asking about smoking Rate of non-clinician asking about smoking Rate of clinician or non-clinician asking about smoking

25.5% 45.5% 57.5%

52.6% 65.7% 80.8%

p < .001 p < .001 p < .001

141.9 76.8 120.4

Advice ‘B’ (Advised that day by clinician) General recommendation to stop smoking Specific advice or suggestions to stop smoking

48.8% 23.8%

69.8% 42.6%

p < .01 p < .01

8.9 7.4

Assisted ‘C’ (any) Not reported *Using patient exit interviews, smokers were asked whether the clinician had advised them that day to stop smoking ‘B’ and/or provided specific advice or suggestions that day on how to stop smoking ‘B’.

84


85


Financial Incentives Introduction Financial incentives include any ‘pay-for-performance’ scheme or financial incentives or bonus payments offered or paid to individual practices or medical groups based on the performance of specified smoking cessation activities reaching or exceeding preset clinical performance targets. Note that all studies of financial incentives include ‘audit’ and some degree of ‘feedback’. These audit and feedback components are inherent to all incentive schemes because it is necessary to measure HCPs’ performance and make a comparison to some performance/payment threshold, and further, receipt of a payment (or not) is a form of feedback. Therefore, the following studies have been grouped together as financial incentives, not multi-component interventions. The principal component of all the interventions is ‘pay for performance’. Four of the 37 studies (11%) evaluated financial incentives. One moderately large 12 month cluster randomised controlled trial was conducted in the United States (Roski, et al., 2003) and three large population-based interrupted time series studies were also included: two recent studies in the United Kingdom (Coleman, et al., 2007; Millett, et al., 2007) and one similar study in the United States (Amundson, et al., 2003). Two of the studies evaluated financial incentives within the context of smoking specific trials (Millet; Roski) and the other two within the context of general prevention trials, including smoking outcomes. Financial incentives: results from four studies Financial incentives, evidence level II. In a trial of financial incentives and access to a centralised smoker registry, Roski et al. (2003) randomised GPs in 40 clinics to one of three groups. The ‘incentives’ group received financial incentives for reaching preset clinical performance targets (threshold set at 15 percentage points above the average baseline performance to trigger a $10,000 [US] bonus per month). The ‘registry’ group received financial incentives and access to a centralised smoker registry and audit and weekly feedback on the individual clinic and the overall group’s referral activity. The control group received printed versions of the smoking cessation guidelines only. Clinical practice patterns including tobacco user identification, providing advice to quit to current smokers, and referral activity was measured in two rounds of brief, one-page cross-sectional patient pen-and pencil exit surveys (4813 baseline and 4734 follow-up). Patient outcomes included accessing counselling, medication, and quit rates. Referrals to a free smoking cessation counselling programme were made available when clinic staff or providers asked patients to fill out a referral postcard or provided identified smokers with a toll-free telephone number. The authors reported that they were able to statistically significantly increase the use (or acceptance) of smoking cessation counselling programs for patients seen in the incentive and registry conditions. However, by the end of the trial, referral rates had dropped back to baseline levels. The study was able to detect a modest positive impact on clinicians’ identification (only) of tobacco users. The authors concluded that it was not clear whether significantly higher incentive payments would have been able to focus clinics’ attention more strongly on the need to provide smokers with advice and assistance to quit. The registry plus incentives intervention offered no apparent advantage over incentives alone, and patient quit rates remained essentially unchanged. Financial incentives, evidence level III-2. Millett et al. (2007) evaluated pay-for-performance in a longitudinal population-based design. The study examined two facets of pay-for-performance incentives: whether financial incentives improve process-of-care measures and whether financial incentives improve patient outcomes. Using a population of patients with

86


diabetes in 32 participating primary care practices in southwest London, the authors examined the impact of pay-for-performance incentives introduced in 2004 as part of the new general practitioner contract to improve support for smoking cessation and to reduce the prevalence of smoking among people with chronic diseases such as diabetes. The study particularly focused on the impact of the documentation of smoking status and on the delivery of smoking cessation advice. Findings from the study showed that the proportion of patients who had their smoking status recorded increased from 90.0% in 2003 to 98.8% in 2005, and that the proportion of patients who received smoking cessation advice increased from 48.0% to 83.5% during this period. Furthermore, there was a decrease in the prevalence of smoking in the study population from 20.0% in 2003 to 16.2% in 2005. These improvements were noticed more in the groups with the poorest performance in the 2003 study period and among ethnic minorities (populations that traditionally receive lower quality care). In summary, Millett et al. (2007) proposed that pay-for-performance incentives appear to be effective in increasing the delivery of cessation advice given by primary care physicians and in reducing the prevalence of smoking among patients with diabetes. However, it is not clear whether these incentives result in improved patient outcomes or whether they simply lead to improved documentation and use of processes of care. Furthermore, it is difficult to relate the observed effect to the intervention alone, as the design used did not control for secular trends. It is suggested that the effect of financial incentives on smoking prevalence may be greater when combined with other quality improvement initiatives within a comprehensive tobacco control strategy. Financial incentives, evidence level III-3. In the context of a network model health plan in the USA, Amundson et al. (2003) evaluated the effect of combining measurement, feedback, and financial incentives (paid for ‘ask’ and ‘advise’ rates ≥ 80%) on the rates at which network physicians identified tobacco users and documented provision of advice quit (but not cessation support/counselling/referral). In a four year time series evaluation, the researchers analysed data from 20 medical groups (6-500 individual HCPs per group comprising many clinics and practices). Baseline data (1996) was provided for the 12 months prior to the implementation of the intervention (14 489 patient charts) that illustrated a large variation in smoking documentation at the last visit: from 0% to 78% (ask rates 49%, range 24% - 78%; advise rates 32%, range 0% -71%). The intervention included feedback (also available publicly on a ‘Consumer Choice’ Website), recognition (awards dinner), and financial rewards paid to medical groups that achieved specific high levels of performance on selected measures of care quality (including smoking ‘ask’ and ‘advise’). The intervention was applied at the medical group level, and there were no additional practice-based or individual level interventions other than providing manuals describing the incentive system. Compared to baseline, mean ‘ask’ rates increased 44% (relative increase) by 1999 and mean ‘advise’ rates had increased 66% relative to baseline. The authors acknowledge that the trial did not utilise a control group, and it is possible physicians’ levels of smoking cessation behaviours were already increasing at the time of the trial, however, national rates were reported to have remained relatively static during the period of study. The authors concluded that performance feedback, combined with a financial incentive, appears to be an effective way for a health plan to improve physician compliance with tobacco treatment guidelines. However, the authors note, that as in their previous (failed) study (Solberg, et al., 2000) this level of sustained change within medical groups requires significant leadership engagement and commitment. Financial incentives, evidence level III-3. In a study not dissimilar to that conducted in the USA by Amundson et al. (2003), Coleman et al. (2007) evaluated the effect of new financial incentives introduced as part of a comprehensive new contract for UK general practitioners by

87


the National Health Service (NHS), in April 2004. The contract included substantial financial rewards for compliance with targets across the spectrum of clinical activity, and approximately 8% of new quality payments could be earned by complying with targets related to smoking cessation management (ascertainment of patient’s smoking status and recording the delivery of GPs’ brief smoking cessation advice but not cessation support/referral). Coleman et al. (2007) analysed data from 1 607 782 patients’ medical records that were held within a computerised, longitudinal, general practice database (The Health Improvement Network database of UK primary care medical records): 14 years pre-intervention and for 18 months post-intervention (for each year between 1990 and 2005 and for each quarter-year from 2003-2005). Compared to the first quarter of 2003, there was an 88% increase in recording of smoking status (rate ratio 1.88, 95% CI 1.87–1.89) and in brief advice to smokers (rate ratio 3.03, 95% CI 2.98–3.09), up to the first quarter of 2004, which was sustained until the first quarter of 2005. The authors concluded that closely-monitored targets coupled with financial incentives do appear to change GPs’ clinical behaviours. However, the authors also caution that some of the observed increase in advice-giving may not reflect more advice being given, but rather, more complete recording of advice that GPs already give, and in addition, that it is also unclear whether or not the observed changes will be sustained over time. Financial incentives: summary Table 20 summarises the evidence pertaining to financial incentives that is directly relevant to the current review, in accordance with the NHMRC dimensions of evidence. Broadly, the four studies of financial incentives all aimed to assess the impact of pay-for-performance schemes on the delivery of primary care smoking cessation interventions. The one randomised trial (Roski, et al., 2003) measured the assessment of patients’ smoking status, the delivery of brief advice, and referral to a smoking cessation counselling program via one-page cross-sectional patient pen-and pencil exit surveys (4813 baseline and 4734 follow-up). The remaining three population-based studies assessed the delivery of services via data extracted from large numbers of electronic patient records. For example, Amundson et al. (2003) analysed data from 20 medical groups and 14 489 patient charts and Coleman et al. (2007) analysed data from 1,607,782 patients' medical records that were held within a computerised, longitudinal, general practice database. In these three studies the outcome was similar: the rates at which physicians documented that tobacco users were identified A and advised to quit B, but not cessation support (generally, due to no targets having been implemented). In their USA-based study, Amundson et al. (2003) reported large between-practice variations in baseline documentation of smoking status and brief advice (from 0% to 78%), and the authors of both the London based study (Millett) and the UK National Health Service study (Coleman) noted that it was not clear whether the incentives resulted in improved patient outcomes or simply improved documentation. Notwithstanding, Roski et al. (2003) reported a modest positive impact on clinicians’ identification (only) of tobacco users, and the remaining studies (Amundson, et al., 2003; Coleman, et al., 2007; Millett, et al., 2007) demonstrated statistically significant increases in the documentation of smoking status and the documented delivery of brief advice Financial incentives: conclusion Taken together, the studies of financial incentives suggest that performance targets coupled with financial incentives do appear to change HCPs’ clinical behaviours, however there is insufficient evidence to determine what the optimal performance threshold levels and/or the optimum level of incentives might be. In addition, the possibility of unintended consequences has not been fully explored and it is also unclear whether or not the apparent positive influence of financial incentives on clinical practice could be sustained over time.

88


If one accepts that the ultimate success of any incentive scheme is whether or not the incentive offered actually enables or motivates a particular course of action (in this case the actual delivery, not merely documentation, of smoking cessation interventions in primary care) – then such pay-for-performance schemes still need to be tested with regard to the actual delivery of the more complex cessation support activities, such as the prescription of NRT and actually counselling patients to stop smoking. These are potentially important studies as they are recent, population-based, and test the merging of contractual obligations with incentivising performance and as such could go some way towards establishing new and accepted standards of care in the area of smoking cessation treatments in primary care settings. Thus, they step away from the traditional study of clinic-level interventions and move towards government-level policy initiatives. Incentive schemes may prove to be worthwhile initiatives within a comprehensive tobacco control strategy.

89


Table 20: Original study results: The effectiveness of Financial Incentives to increase the delivery of effective smoking cessation treatments in primary care settings

FINANCIAL INCENTIVES

Financial incentives: Level II evidence Roski et al. (2003) – Financial incentives, smoking specific. *Outcomes (% increase)

Control Incentive Registry plus Overall rate at follow-up

†Ask ‘A’ p value

6.2% –

14.1% ‡ p < .01

8.1% ‡ p < .01

49.8%

†Advice ‘B’ 18.3% –

24.2% p = .426

17.1% –

51.1%

†Cessation (offered assistance) ‘C’

19.7% –

24.2% p = .451

14.6% –

33.5%

§Accessed counselling ‘C’ 1% –

1.3% –

3.3% ‡ p < .001

1.6%

§Quit rate ‘Q’ 19.2% 22.4% –

21.7% p = .2

*Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. †At last visit, identification of tobacco use status, smokers receiving advice to quit, and smokers offered assistance to quit, based on 4734 patient exit surveys. Percentages are adjusted for age and gender differences between clinics. ‡Statistically significantly improved over improvements in the control group. §Patient smoking cessation behaviour outcomes: Quit rates and ‘accessed’ not ‘referral to’ counselling. Cohort of respondents to a telephone six month follow-up survey of patients initially identified as smokers via a baseline mail survey. Financial incentives: Level III-2 evidence Millett et al. (2007) – Incentives, pay for performance, smoking specific. *Outcomes 2003 2005

p value Percentage change

Ask ‘A’ Smoking status recorded Proportion of patients whose smoking status was recorded

90.0% 67.6%

98.8% 86.7%

p < .001 p < .001

8.8 †19.1

Advice ‘B’ Proportion of patients with documented smoking cessation advice

48.0%

83.5%

p < .001

35.5

Assisted ‘C’ (any) Not reported Prevalence of smoking: 20.0% 16.2% p < .001 –3.8

*Data represents % of patients and % change with adjusted OR for age, sex, ethnic background, deprivation group and practice-level clustering with (95% CI) †Change in percentage significant (p < .001) using McNemar test.

90


Table 20: Original study results: The effectiveness of Financial Incentives to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

FINANCIAL INCENTIVES

Financial incentives: Level III-2 evidence Amundson et al. (2003) – Financial incentives, general prevention. †‡Outcomes Pre-test

1996 Post-test 1999

p value % Change 1996-1999

Ask ‘A’ 49.7% ± 7% 73% ± 7% p < .001 24% (relative increase of 44% compared with baseline)

Advice ‘B’ 32% ±9% 53% ± 12% p < .005 21% (relative increase of 66% compared with baseline)

Assisted ‘C’ (any) – – – Not reported †Mean rates (and standard deviations) of the documentation of Ask and Advise in the patient medical records at the last visit. ‡Note: smoking outcomes reported within the context of a ‘general prevention’ trial. Other targets/incentives were in place at the time of the trial; however, these were not reported in this paper. Coleman et al. (2007) – Financial incentives, general prevention. *Outcomes Pre-test

† Incidence per 10 000 per quarter

Post-test † Incidence per 10 000 per quarter

p value Rate ratio RR and 95% confidence interval (95% CI)

Ask ‘A’ 1100 1350 1.88, 95% CI 1.87–1.89

Advice ‘B’ 1000 1300 3.03, 95% CI 2.98–3.09

Cessation ‘C’ – – § n.s. *Further details of patient outcomes are available in the original authors’ publication. §There was no consistent change in the incidence of receiving nicotine replacement therapy, however, there were no targets for NRT prescriptions. †Approximate, incidence abstracted from histograms (rate ratio reported by study authors).

91


Audit and Feedback Introduction Audit & Feedback interventions were those in which a summary of clinical performance (and/or comparisons with nominal or target performance) was fed back to clinicians, either on an individual or group basis, and whereby the information provided was intended to modify clinical practice. Three studies of audit & feedback were included and all were cluster randomised controlled trials (evidence level II). Collectively, the studies involved 677 health care professionals comprising mainly GPs and nurses. In two of the studies (Bentz, et al., 2007; Wadland, et al., 2007) the context was smoking-specific. The other study reported smoking outcomes within the context of a general prevention programme (Moher, et al., 2001). The outcomes measured varied, Bentz et al. (2007) reported ‘A’, ‘B’ and ‘C’ via an audit of the electronic medical record (EMR), Wadland et al. (2007) tracked HCPs referrals to a state-level tobacco Quitline ‘C’, and Moher et al. (2001) reported the ‘adequate assessment of risk factors’ (i.e. the documentation of smoking status in patient notes). Audit & Feedback: results from three studies Audit & Feedback, evidence level II. In a moderately sized cluster randomised trial, Bentz et al. (2007) evaluated the impact of an electronic health record (EHR) generated audit and feedback system on rates of referral to a state-level tobacco Quitline. Nineteen primary care clinics involving 279 HCPs (150 medical doctors, 14 registered nurses, eight nurse practitioners, and 107 medical assistants) were randomised into intervention clinics (n = 10) and control clinics (n = 9). Both groups received 30 minute training sessions for providers and staff to review evidence-based cessation strategies, and the EHR was modified to include an electronic Quitline fax-referral capability. In addition, the intervention group received feedback reports that rated individual HCP’s performance in asking, advising, assessing, and assisting with tobacco cessation, compared with a clinic average, and an achievable benchmark of care (defined as the average performance of the top 10% of providers being measured). During 12 months of follow-up, documented rates of assessing smoking status, advising, and assisting smokers were significantly improved in the intervention clinics compared with the control clinics (p < .001). The authors reported that it was simple to modify the EHR, redesign the clinic workflow, gain buy-in from staff, and incorporate the intervention into routine clinic practice. Multivariate analysis showed that the presence of a staff member who publicly advocated the use of the fax-referral process in each clinic (a ‘clinic champion’, usually a nurse) was an important covariate in predicting referrals. The authors recommended that future studies should focus on ways to recruit and reinforce the behaviour of clinic champions. The trial was contingent on the availability of an EMR that was suitably adaptable and had the capacity to produce feedback reports to individual HCPs based on individual and clinic-wide performance data. Thus, generalisability is limited to fundamentally similar contexts. Audit & Feedback, evidence level II. Wadland et al. (2007) conducted a cluster-randomised clinical trial in Michigan State University to assess the impact of an intervention for primary care clinicians which adopts quarterly comparative feed-back to generate referrals to a telephone quit-line service over an 18-month period (January 2003 and June 2004). The intervention was compared with general reminders to primary care clinicians about the performance of preventive services. Clinicians in the intervention group were given quarterly comparative feedback on their individual and group referrals and how their referral activity compared with the mean referrals based on the overall top 10% of clinicians. The comparative feedback report was a one-page report showing two graphs. The first graph displayed four sets

92


of data: the quarter benchmark for referrals, the individual clinician’s referrals, his or her practice group’s referrals, and the relative performance of the study group in numbers of referrals (percentage relative to the benchmark). The second graph showed the actual number of referrals made by the individual clinician per quarter. Three hundred and eight clinicians within 87 practices completed baseline survey forms. Baseline exit surveys of patients were conducted in each practice for two half-days in which more clinicians were in the practices (1,992 patients completed the questionnaire). The change from baseline to post-intervention in clinician referrals in both intervention and control groups, the estimated quit responses of referred smokers, as well as the incremental costs of the feedback intervention were all measured over the 18-month period of the study. The study found that specific feedback significantly influenced greater referrals in the intervention (feedback) than control (no feedback) group. However, across groups, some clinicians referred at a high rate and some had no referrals Results showed significant differences in the Absolute Quit Line referrals between the two groups over time. Both groups increased during the fifth quarter, but overall, referrals were very modest. There were greater referrals by fax than by telephone; telephone referrals compared with fax referrals were significantly more likely than fax referrals to actually result in Quit Line enrolment (77.1% vs. 42.1%, p < .001). More estimated quits occurred in the intervention practices than in the control practices (66 vs. 36). The study found that specific feedback significantly influenced greater referrals in the intervention (feedback) than control (no feedback) group. However, across groups, some clinicians referred at a high rate and some had no referrals. The results indicated a significant difference in the intervention group having an overall greater number of referrals than the control group, however, the majority of referrals were due to specific high-referring clinicians. One of the major limitations of this study is the incomplete analyses of individual clinician and practice-level data because of the clustering effect evident and a lack of adequate power to detect significant differences. Furthermore, results may not accurately reflect longitudinal changes as the exit surveys assessed only two ½-days per practice. However, the authors concluded that providing simple, clear, and comparable feedback on referrals to a quit-line service demonstrated a modest impact on the referral performance of clinicians but may require a major integration of practice and system supports to become effective. Audit & Feedback, evidence level II. In a cluster randomised controlled trial, Moher et al. (2001) compared the effectiveness of three interventions for improving the secondary preventive care of patients with coronary heart disease delivered at 21 general practices in Warwickshire (involving more than 62 GPs and more than 28 nurse practitioners). During the study, the Warwickshire Medical Advisory Audit Group conducted an audit of the care of patients with coronary heart disease, practices had to report audit results to qualify for health promotion payments. Patients were aged 55-75 years and were diagnosed with coronary heart disease. Practices were randomly allocated to either audit & feedback (audit group), or recall to general practitioner (GP recall group), or recall to nurse clinic (nurse recall group). Within the first group, practices were given summary practice-level audit results at a practice meeting (number of patients with established coronary heart disease, the prevalence of identified coronary heart disease in the practice, and the proportions of patients with adequate assessment, and pharmacological treatment). Practices were also given anonymous comparative summary data from the other participating practices. Practices were asked to provide usual care and were given no further support during the trial. The second group (GP recall) were given the same patient information as the audit group but were also given the names of patients identified as having coronary heart disease. There was ongoing support in setting up a register and recall for system for regular review of patients with coronary heart

93


disease by their general practitioner. The nurse recall group were given the same patient information as the GP recall group, a nurse facilitator gave ongoing support to the practices in setting up a register and recall system for systematic review of patients with coronary heart disease in a nurse-led clinic. The interventions started from May 1998 and continued until October 1999. The primary outcome was adequate assessment at 18 months follow-up from an audit of patients’ notes. One of the outcomes assessed was a record of smoking habit and, for smokers, a record of review in the past two years. Results showed that at baseline, about 70% of patients were adequately assessed for smoking status. At follow-up, the assessment of smoking status increased, especially in the nurse recall groups. At follow-up the groups differed largely in the proportions of patients being adequately assessed (95% in the nurse recall, 72% in the GP recall group, and 78% in the audit group). Compared with the audit group, the absolute increase in the proportion of patients adequately assessed for smoking status was significantly higher among nurse recall group (after adjustment for baseline levels) than the GP recall group. The study showed that setting up a patient register for a general practice would increase the probability of planned follow-up taking place no matter whether the responsibility lies with general practitioner or with the practice nurse, with the latter being more effective. However, the study authors concluded that increased follow-up alone does not necessarily improve clinical outcomes. The authors also acknowledged important limitations: non-participating practices tended to be smaller than participating practices and have minimal nursing support (so they might have had difficulties in implementing recall strategies); baseline levels of smoking status assessment were already relatively high in all groups (approximately 70%); and that the observed intervention effect may have been greater if the trial had not taken place in the context of a health authority audit initiative relating to patients with coronary heart disease, backed by a financial incentive. Audit & Feedback: summary Table 21 summarises the evidence pertaining to audit & feedback interventions that is directly relevant to the current review, in accordance with the NHMRC dimensions of evidence. Common to all three studies (Bentz, et al., 2007; Moher, et al., 2001; Wadland, et al., 2007) was the provision of comparative summary data describing individual health care professional’s performance (and/or clinic performance) against some reference standard (for example, the average performance of the top 10% of providers being measured). Also, all studies measured the outcomes objectively from patient chart audit and/or the electronic medical records. However, the studies differed in the number of outcomes measured, Bentz et al. (2007) measured HCPs’ performance in all the ‘ABC’s: documenting patients’ smoking status, providing brief advice and providing cessation support including electronic referral to counselling services. Wadland et al. (2007) only measured the differences in the absolute Quit Line referrals between the two groups over time ‘C’. Moher et al. (2001), within the context of a general prevention study, measured the ‘adequate assessment’ of smoking status at 18 months follow-up from an audit of patients’ notes and/or a record of review in the past two years. Notwithstanding, all three studies reported statistically significant between group differences on the outcomes measured. Audit & Feedback appears to have potential in promoting improved process outcomes, however, as Moher et al. (2001) concluded, increased assessment and follow-up alone does not necessarily improve clinical outcomes. Other considerations include the necessary widespread availability of electronic medical record systems capable of quickly and easily generating comparative feedback reports and whether such reports should be limited to smoking outcomes only or whether they should include general preventive indices as well. A further consideration is whether the reports should be able to differentiate by category, for example, to represent any

94


particular patient sub-groups such as diabetes patients, coronary heart disease patients, and/or the socio-economically disadvantaged. Audit & Feedback: conclusion Audit & Feedback appears to be an effective means of changing HCPs’ behaviour. However, additional research is required to establish the optimal methods of conducting audits and of delivering meaningful feedback, in the context of typical New Zealand primary care settings.

95


Table 21: Original study results: The effectiveness of Audit & Feedback to

increase the delivery of effective smoking cessation treatments in primary care settings

AUDIT & FEEDBACK Audit & Feedback: Level II evidence Bentz et al. (2007) – Audit & Feedback, smoking specific. *Outcomes Control Intervention p value Odds ratio (95% CI) **Ask ‘A’ 88.1% 94.5% p = .05 † 1.53 (1.05–2.23) Advice ‘B’ 52.7% 71.6% p = .001 †1.36 (1.33–1.39) Assisted ‘C’ (any) 10.5% 20.1% p = .001 †1.92 (1.78–2.08) Referral or receipt of calls at the Quitline ‘C’

4% §‡ 1.53 (1.05–2.23)

*Patient outcomes are outside the scope of this review. However, for completeness, quit rates and/or abstinence rates have been added here as reported by the study authors. Aspects of the study that relate specifically to these patient outcomes have not necessarily been appraised or synthesised further, and more complete details of patient outcomes are available in the original authors’ publications. **Percentage of all patient records with documented smoking status at follow-up. †Crude Odds ratio (95% CI). §After adjusting for case mix and the presence of a clinic champion, baseline ask rates, insurance mix, clinic size, and clinic volume. ‡Compared with population-based historical data. Wadland et al. (2007) – Audit & Feedback, smoking specific. *Outcomes Control (general reminders) Intervention

(feedback and reminder)

p value

Assisted ‘C’ (any) Absolute Quit Line referrals (by clinicians) Fax referrals Telephone referrals

220 168 (76.4%) 52 (23.6%)

484 427 (88.2%) 57 (11.8)%

p < .001

∆Estimated quits occurred 35.90 65.65 Total cost per quit Incremental cost per quit

$145.28 $108.93

$285.72 $170.04

*Number of referrals to Quit Line ∆Actual referrals x estimated quit rates = number of quits

96


Table 21: Original study results: The effectiveness of Audit & Feedback to increase the delivery of effective smoking cessation treatments in primary care settings (continued)

AUDIT & FEEDBACK Audit & Feedback: Level II evidence Moher et al. (2001) – Audit & Feedback, general prevention. †‡Outcomes

Baseline Mean number (range)

Follow up Mean number (range)

p value for difference between groups*

Audit group 73 (50-91) 78 (56-92)

p < .001

GP recall group 71 (47-96)

92 (77-100)

p = .001

Nurse recall group 71 (46-85)

95 (88-98)

p = .001

*Based on means adjusted for baseline. †Smoking results were presented within the context of a general prevention trial. Patients with coronary heart disease per practice at follow-up and percentage of these patients with adequate assessment at baseline and follow-up, by trial group. Values are mean (range) percentages unless stated otherwise, outcomes were the record of smoking habit and, for smokers, a record of review in the previous two years. ‡Includes the adequate assessment of all three risk factors (blood pressure, cholesterol and smoking status): nurse recall group vs. audit group, difference 33% (95% confidence interval 19% to 46%; P<0.001); GP recall group vs. audit group, difference 23% (10% to 36%; p = .002); nurse recall group vs. GP recall group, difference 9% (-3% to 22%; p = .13).

97


Other Interventions Introduction Of the 37 studies, two studies were not easily classified within the five main intervention types as previously described. One randomised trial (Richmond, et al., 1998) evaluated the effectiveness of reinforcement contact from the medical practitioner trainer as compared to training alone and another randomised controlled trial evaluated an algorithm as a stand-alone intervention (the provision of intervention algorithm at each patient visit compared to a no-algorithm control) (Ockene, et al., 1994). Other interventions: results from two studies Other interventions, evidence level II. The effect on GPs’ delivery of smoking cessation interventions of a training workshop and reinforcement contact from the medical practitioner trainer was compared to training only in a randomised controlled trial in Australian general practices. Richmond, Mendelsohn and Kehoe (1998) randomised 198 family physicians to the intervention group (training and reinforcement) or the control group (training but no reinforcement). The training consisted of a brief two hour workshop and the reinforcement component consisted of three brief telephone calls at two weeks, two months, and four months after the initial training by a peer doctor. Doctors completed a self-report questionnaire, mailed at six months after the intervention, which inquired about their current and future use of cessation interventions. Also, the number of smoking cessation information booklets given to smokers over the six-month study period was analysed. Three very brief phone calls in four months among full-time physicians were shown to significantly increase the distribution of smoking cessation information booklets given to patients. Follow-up phone calls after training increased GPs’ use of smoking cessation interventions but between-group differences only approached statistical significance (p = .06). Other interventions, evidence level II. Ockene et al. (1994) evaluated the effects of training physicians to use a patient-centred counselling approach to deliver smoking cessation counselling using an algorithm during the patient visit. Specifically, if the intervention had an impact on the type and number of smoking intervention steps taken by the physician. They randomised 50 physicians (40 internal medicine residents and 10 attending physicians) to an algorithm condition (provision of intervention algorithm at each patient visit) or a no-algorithm condition. Patient exit interviews were used to measure the steps used by physicians with patients. Five-hundred twenty-seven adult smokers were seen in an ambulatory clinic in an academic medical centre between June 1990 and April 1992. Results show no difference in the overall PEI scores or in individual PEI steps taken between the algorithm and no-algorithm conditions. The authors concluded that the provision of an intervention algorithm at each patient visit did not increase the likelihood that trained physicians who are cued to intervene will perform more of the intervention steps taught. Other interventions: summary Table 22 summarises the evidence pertaining to other intervention types directly relevant to the current review, in accordance with the NHMRC dimensions of evidence. Other interventions were interventions in which the principal component was not training, multi-component, a reminder system, an incentive, or an audit and feedback intervention, but the aim of the intervention was nevertheless to enable HCPs to provide better care for their smoking patients, by improving the delivery of smoking cessation treatments and activities within the clinic. The two interventions differed in that the algorithm was intended to act at the point of care (i.e. actually during the provider-patient interaction). In Richmond et al. (1998)

98


study, the brief telephone calls at two weeks, two months, and four months after the initial training was intended to reinforce the clinicians’ preventive behaviour generally rather than on a patient-by-patient basis. In both studies, the results showed no statistically significant between group differences on the process outcome of interest asking all patients smoking status, providing brief advice and various fundamentals of cessation support (except handing out self-help booklets; p < .05, Richmond et al. 1998). In both cases, the HCPs were trained, and baseline levels of providing brief advice were relatively high, between 70-85%. Other interventions: conclusion There is insufficient evidence to draw robust conclusions from only two relatively small studies. However, the results are consistent with other studies of low intensity single component interventions, in that it appears much greater and sustained efforts are required to significantly change health care professionals’ delivery of smoking cessation services.

99


Table 22: Original study results: The effectiveness of ‘other’ interventions to increase the delivery of effective smoking cessation treatments in primary care settings

OTHER Other interventions: Level II evidence Richmond et al. (1998) – Reinforcement, smoking specific. Outcomes Control (95% CI) Intervention p value †‘A/B/C’ 84% 93% p = .06 n.s ‡Cessation ‘C’ (self-help booklets) 32.8 (27.6 – 38.0) 40.1 (35.4 – 44.8) p < .05 †GP self-report of smoking cessation intervention/s delivered, at six months follow-up, including appropriate delivery of one or more of: brief advice, brief motivational intervention, encouragement to set a quit date, and offered nicotine replacement therapy. ‡The mean number of self-help booklets distributed to smokers by full-time physicians. Ockene et al. (1994) – Other, algorithm, smoking specific. Outcomes

Control Intervention (algorithm)

p-value Odds ratio (95% CI)

*PEI score 4.94 4.88 n.s. Not significant Ask ‘A’ – – Not reported Advice ‘B’ Advised to stop smoking

71.5% Not reported

Not reported

∆Assisted ‘C’ (any) Asked about past experiences

46.5%

Not reported

Not reported

Discussed smoking 84% Discussed reasons 62.4% Asked about problems 36.6% Discussed specific strategies 34.7%

Asked about dependency 25.6% Patient agreed to stop smoking or if not, patient agreed to cut down

(29.0%) or (21.1%)

Set date to stop/cut down 22.4% Put plan in writing 5.1% Discussed other behavioural changes

27.3%

Given written materials 16.9% Set future contact or if not planned to discuss later

12.1% or 43.6%

*Patient Exit Interview (PEI) scores. ∆There was no difference in overall PEI scores or in individual PEI steps taken between the algorithm and no-algorithm conditions.

100


101


Body of Evidence In an effort to illustrate the entire body of evidence directly relevant to the current review, Table 23 summarises the included evidence in accordance with the NHMRC guidelines. Only data for the primary outcomes ‘Ask’, ‘Brief advice’ and ‘Cessation support’ are summarised. This summary table is intended to provide a quick reference to the included studies and, in particular, the specific outcomes reported and the results. The summary characteristics also include the author, country, intervention type, the level and quality of evidence, and clinical relevance. In some cases, the clinically relevant effect is noted as mixed and this indicates that of the ABC outcomes measured, some demonstrated statistically significant beneficial effects while others did not. The evidence presented in this systematic review is derived from a range of studies representing predominantly medium to high levels of evidence. Of the 42 studies identified as eligible, five were systematic reviews/meta-analyses. Three of these systematic reviews contained both randomised trials and other non-randomised designs and were considered level I / III-3, the latter reflecting the same level as the non-randomised trials that were included within the review. Thirty-seven of 42 studies were original primary research including 14 cluster randomised controlled trials (C-RCTs), 11 randomised controlled trials (evidence level II), two non-randomised controlled clinical trials (evidence level III-2), six controlled before-and-after trials (evidence level III-3), three time-series designs (evidence level III-2) and one study analysed comparative data from two temporally different cross-sectional surveys (evidence level IV). Table 23 overviews the studies listed by level of evidence and alphabetically within each level. Despite the large volume of evidence, synthesising the body of evidence as a whole was to some extent problematic for several reasons: (1) the research covers a broad range of intervention types and for some interventions, there were a limited number of studies identified; (2) the research relates to both smoking specific and general prevention trials; (3) the outcomes reported in each study differ (i.e. not all of the studies attempted to measure all of the ‘ABC’ outcomes) and (4) participation rates were variable. Therefore, it is not appropriate to statistically meta-analyse the results. Nevertheless, the results of the 42 included studies indicate that there is evidence to support the routine use of a range of different intervention types aimed at increasing the delivery of effective smoking cessation treatments in primary care settings. The key findings are examined further in the following Discussion.

102


Table 23: Body of evidence: Effectiveness of interventions Strength of evidence

Study Country

Intervention Comparison Level of

evidence Quality of evidence

Statistical precision a

Clinically relevant effect?

Lancaster et al. (2002)

U.K. Training Usual care I Good Yes �

Anderson et al. (2004

The Netherlands

Training; office systems; educational facilitator; multi-component.

Usual care I / III-3* Good Yes �

Dicky et al. (1999)

U.S.A. Office tools; Teamwork. Usual care I / III-3* Good Yes �

Grimshaw et al. (2004)

U.K. All intervention types G Usual care I / III-3* Good Yes �mixed

Fiore et al. (2008)

U.S.A. Training; reminders. Usual care I** Good Yes �

* Systematic reviews that contained both randomised trials and lower levels of evidence were annotated level I – III-3, the latter reflecting the same level as the non-randomised trials that were included within the review. ** Meta-analysis, five comparisons, reported within a clinical practice guideline. Bentz et al. (2007)

U.S.A. Audit & Feedback No Audit & Feedback II Good ‘A’, p = .05 ‘B’, p = .001 ‘C’, p = .001

�

Borgiel & Williams (1999)

Canada Training: outreach facilitation, continuing medical education (CME).

No intervention/ Practice assessment report only

II Fair ‘A’, n.s. ‘B’, n.s. ‘C’, n.s.

�

Cockburn et al. (1992)

Australia Training: educational facilitator. – Un-trained volunteer – Postal delivery of Guideline

II Good ‘A’&’B’ p = .07 ‘C’ p = .02

�mixed

Dietrich et al. (1992)

U.S.A. Training: academic detailing G Training: academic detailing in office system design/implementation G

Cessation training vs. system training vs. no intervention

II Fair ‘A’, n.a. ‘B’, p = .017 ‘C’, n.a.

�mixed

Hogg et al. (2008)

Canada Training: outreach facilitation G Usual care II Good ‘A’, n.s. ‘B’, n.s. ‘C’, n.s.

�

103


Table 23: Body of evidence: Effectiveness of interventions (continued) Strength of evidence

Study Country





Joseph et al. (2004)

U.S.A. Training: academic detailing Usual care II Good ‘A’ p = .0001 ‘B’, n.s. ‘C’, n.s.

�mixed

Katz et al. (2004)

U.S.A. Multi-component: guideline based intervention Usual care II Fair ‘ABC’, p < .01 �

Lemelin et al. (2001)

Canada Multi-component G Usual care II Fair n.s., for smoking outcomes

�

Lennox et al. (1998)

Scotland Training: off-site workshops Usual care II Fair ‘A’, n.s. ‘B’, n.s. ‘C’, n.a.

�

McPhee et al. (1991)

U.S.A. Reminders G Usual care II Good ‘A’, n.a. ‘B’, p = .014 ‘C’, p = .027

�

McRobbie et al. (2008)

U.K. Training: lectures Usual care II Good ‘A’, n.a. ‘B’, n.a. ‘C’, 95% CI: 2.5-24.9

�

Moher et al. (2001)

U.K. Audit & Feedback G GP recall vs. Nurse recall II Fair ‘A’ p < .0001 ‘B’, n.a. ‘C’, n.a.

�

Morgan et al. (1994)

U.S.A. Training: academic detailing Usual care II Fair ‘A’, n.a. ‘B’, p < .01 ‘C’, p < .01

�

Ockene et al. (1997)

U.S.A. Training: continuing medical education (CME). Usual care II Fair ‘B/C’ (printed materials) p = .026

�mixed

104



Study Country





Ockene et al. (1994)

U.S.A. Other: algorithm no algorithm II Fair ‘A’, n.a. ‘B’, n.s. ‘C’, n.s.

�

Pbert et al. (2006)

U.S.A. Multi-component: training and reminder Usual care II Fair ‘ABC’, p < .01 �

Piper et al. (2003)

U.S.A. Reminders: vital sign stamp (non-electronic) Standard vital signs not including smoking status

II Fair ‘A’, p = .002 ‘B’, n.a. ‘C’, n.a.

�

Richmond et al. (1998)

Australia Other: reinforcement following training No reinforcement following training

II Fair Any ‘ABC’, p = .06 n.s. ‘B/C’ (printed materials) p < .05

�mixed

Roski et al. (2003)

U.S.A. Financial incentives No intervention/distribution of printed smoking cessation guidelines only

II Fair ‘A’, p < .01 ‘B’, n.s. ‘C’, n.s.

�mixed

Rothemich et al. (2008)

U.S.A. Reminders: vital sign (non-electronic) No intervention II Good ‘A’, n.a. ‘B’, p = .04 ‘C’, n.s.

�mixed

Solberg et al. (2000)

U.S.A. Training: continuous quality improvement (CQI) G Usual care II Good ‘A’, n.s. ‘B’, n.a. ‘C’, n.a.

�

Strecher et al. (1991)

U.S.A. Multi-component: training & reminders Tutorial & prompt vs. tutorial only vs. usual care

II Good ‘A’, n.a. ‘B’, p < .05 ‘C’, p < .001

�

Unrod et al. (2007)

U.S.A. Training: academic detailing Usual care II Fair ‘A’, p < .001 ‘B’, p < .001 ‘C’, p < .001

�

105



Study Country





Wadland et al. (2007)

U.S.A. Audit & Feedback Simple postcards informing of the Quit Line service

II Fair ‘A’, n.a. ‘B’, n.a. ‘C’ (referrals), p < .001

�

Young et al. (2002)

Australia Multi-component: audit and feedback plus resources

Intervention of identical format and intensity but about cervical screening

II Good ‘A’, n.s. ‘B’, n.s. ‘C’ (NRT), p < .005

�mixed

None identified – – – III-1 – – – Amundson et al. (2003)

U.S.A. Financial incentives G Pre-intervention III-2 Good ‘A’, p < .001 ‘B’, p < .005 ‘C’, n.a.

�

Coleman et al. (2007)

U.K. Financial incentives Pre-intervention III-2 Good ‘A’, RR 1.88, 95%CI 1.87-1.89 ‘B’, RR 3.03, 95%CI 2.98-3.09 ‘C’, n.a.

�

Goldstein et al. (2003)

U.S.A. Training: academic detailing Usual care III-2 Fair ‘A’, n.s. ‘B’, n.s. ‘C’, p = .06, n.s.

� Borderline

Kottke et al. (1992)

U.S.A. Training: academic detailing Usual care III-2 Good ‘A’, p < .05 ‘B’, p < .05 ‘C’, n.s.

�mixed

106



Study Country





Millett et al. (2007)

U.K. Financial incentives: pay for performance Pre-intervention III-2 Fair ‘A’, p < .001 ‘B’, p < .001 ‘C’, n.a.

�

DePue et al. (2002)

U.S.A. Training: academic detailing Pre-intervention III-3 Good ‘A’, OR 3.00, 95% CI 1.19-7.55 ‘B’, n.s. ‘C’, n.s.

�mixed

Fiore et al. (1995)

U.S.A. Reminders: vital sign stamp (non-electronic) Pre-intervention III-3 Fair ‘A’, p < .001 ‘B’, p < .01 ‘C’, n.a.

�

Flocke et al. (2006)

U.S.A. Multi-component: office systems G Pre-intervention III-3 Good ‘A’, n.a. ‘B’, n.s. ‘C’, n.a.

�

Krist et al. (2008)

U.S.A. Reminders: electronic system prompt (eLinkS) Pre-intervention III-3 Fair ‘A’, n.a. ‘B’, n.a. ‘C’, –

�

Ozer, et al. (2001)

U.S.A. Multi-component: training, office tools, health educator G

Pre-intervention III-3 Fair ‘A’, p < .001 ‘B’, p < .001 ‘C’, n.a.

�

Secker-Walker et al. (1992)

U.S.A. Training: workshops Pre-intervention III-3 Fair ‘A’, p < .001 ‘B’, p < .001 ‘C’, n.s.

�mixed

Tremblay et al. (2001)

Canada Multi-component Pre-intervention IV Fair A’, p < .001 ‘B’, p < .001 ‘C’, p < .05

�mixed

Abbreviations: �– yes; � – no; n.a. – not applicable; a True effect rather than a chance finding; G General prevention study; n.s. – not statistically significant; NRT– use of or prescription of nicotine replacement therapy; RR – rate ratio; OR – odds ratio.

107


Discussion Many programmes that have aimed to increase the frequency with which health professionals provide smoking cessation interventions in primary care settings have been shown to be effective in terms of process outcomes. These process outcomes include recording patients’ smoking status in the medical records, increasing the number of patients advised to quit, and increasing the number of patients receiving some form of smoking cessation support and/or counselling. However, the effects observed are variable, and in many cases, it is difficult to know if these differences are related to the interventions, the settings, or to the study contexts. Unfortunately, there is insufficient evidence to be able to ‘rank’ interventions for effectiveness or predict how any one intervention or combination of intervention components might perform in any particular context or setting. New Zealand primary health care settings are likely to be idiosyncratic, and careful selection of interventions along with robust process and impact evaluations of programmes should be conducted to ensure the optimum utilisation of resources. On the whole, it appears that when using interventions such as training, reminders, audit and feedback, and multi-component interventions, absolute improvements in practice of between 5% and 10% might be realistic. In the context of behaviour change interventions generally, such improvements are meaningful and relevant. In interpreting the results, from a public health perspective, it is important to keep in mind that even small changes in HCPs’ behaviours might be potentially important when many hundreds of patients are affected. Differences across outcomes The influence on clinicians’ behaviours imparted by such programmes does not appear to be reflected equally across the ‘ABC’ outcomes. Generally, the incidence of recording patients’ smoking status in medical records ‘A’ appears to be the most easily influenced behaviour. The more complex and time-intensive activities of providing brief advice ‘B’ or providing cessation counselling or referral to other cessation services or prescribing NRT and/or other smoking cessation support activities ‘C’, all seem to be more resistant to change.

108


Figure 6 depicts an apparent ‘complexity vs. frequency’ relationship. Figure 6 is not based on a formal secondary data analysis or any modelling of data, but simply reproduces data from three good quality, moderately large, and relatively recent studies, as typical examples (Bentz, et al., 2007; DePue, et al., 2002; Goldstein, et al., 2003).

Ask Brief Advice Cessation

GoldsteinDePue

Bentz

0

10

20

30

40

50

60

70

80

90

100

Perce

ntag

e of p

atien

ts rec

eiving

the

appr

opria

te int

erven

tions

Smoking cessation process outcomes

Delivery of smoking cessation interventions by HCPs vs. the measured ABC outcomes- results from three trials

Figure 6 An illustration of the apparent decline in cessation activities by HCPs with the increasing complexity

of the activities. Data sourced from three good quality, moderately large, and relatively recent studies studies as examples: Bentz et al. (2007), Goldstein et al. (2003) and DePue et al. (2002).

For the three example studies, it can be seen that as the complexity of the brief intervention increases, so the likelihood of the intervention actually being delivered by health care professionals appears to decrease. Generally, the follow-up rates of ‘ABC’ followed a similar pattern to that of the baseline rates; with greater improvements being demonstrated for ‘Ask’ and ‘brief advice’ and typically lesser improvements being demonstrated for ‘cessation support’. At follow-up, ‘ask’ rates were typically moderate to high and increasing the assessment and recording of patients smoking status was commonly achieved. In many cases the recording of all patients’ smoking status was completed by non-medical, rooming, and/or administrative staff rather than the GPs. The delivery of ‘brief advice’ to identified smokers, generally by GPs, was typically moderate and the provision of cessation support less common. The time required for cessation support activities and the number of activities available to GPs (e.g. one-on-one counselling, prescription of NRT, referral to external specialist counselling services, setting a quit date, arranging follow-up) arguably makes this component of the ‘ABC’ approach the most complex and time intensive, and it appears, the least commonly implemented. In another example (reporting only cessation support as the outcome) McRobbie et al. (2008) reported on the success of their training intervention in which intervention group HCPs were five times more likely to make referrals to community specialist cessation services. While very encouraging, it is of note that the behaviour ‘referral’ is relatively more simple than HCPs delivering smoking cessation counselling themselves ‘on the spot’ or arranging a follow-up visit and subsequently delivering cessation treatments (whether counselling of pharmacological or both). Thus, the availability of such external specialist services appears to be an important

109


factor here. Others reported diminishing effects with increased complexity of the behaviour (for example, Kottke, et al., 1992; Ockene, et al., 1997; Ozer, et al., 2001); the authors typically reported statistically significant increases in recording patients’ smoking status and providing brief advice but non-significant changes in cessation support activities. Other studies that reported specific activities within cessation support (Ockene, et al., 1994; Secker-Walker, et al., 1992) demonstrated a drop-off in the performance of the more ‘difficult’ tasks– recommending/prescribing NRT, counselling, and/or setting a quit date versus simply handing out information pamphlets. Differences across contexts: smoking specific vs. general prevention For ‘ABC’, general prevention programmes performed relatively poorly. In the main, the smoking process outcomes were influenced less favourably in general prevention trials than in trials of a similar intervention type, but for smoking only. This was evidenced particularly in the cases of training interventions and multi-component interventions. It appears that within general prevention programmes, the intervention effect may become diluted for all outcomes– and to a much greater extent with respect to smoking outcomes. This dilution effect might be related to the fact that focusing on multiple preventive manoeuvres results in attention being divided or it might be related to time constraints or other limitations. One recent study in the U.S. conducted by Yarnall et al. (2003) estimated that 7.4 hours of every working day would be required for a clinician to deliver all of the commonly recommended preventive services to each of his of her patients. There is insufficient evidence to determine the reasons for these differences between other outcomes and smoking outcomes in particular– the latter appearing much more resistant to change. Low participation rates and poor implementation of the intervention components also appears to be common with general prevention studies, even under researcher facilitated trial conditions, with volunteer health care professionals. Additional work is required to understand how to increase HCPs’ adoption of the tools and system changes needed to influence smoking cessation activities within general prevention programmes. Policy makers will need to prioritise those preventive manoeuvres that might potentially make the most difference to overall population health. These findings need to be taken within the context of the available resources and the prevalence profile of preventable illness in general in New Zealand. If a large initiative is undertaken aimed at influencing HCPs’ smoking cessation activities, then policy makers must consider the possibility that the effectiveness of other preventive interventions might be attenuated – and vice versa.

110


111


Rapid Review: Results and Implications for Practice Training (15 studies): training healthcare professionals is more effective than no training in increasing the numbers of smokers who have their smoking status recorded in the medical records, are offered brief advice about stopping and who subsequently receive at least some cessation support. It appears that the most effective form of training is training HCPs in the skills of actually delivering smoking cessation treatments to their patients (for example, training in effective counselling techniques or methods). Training HCPs to design and implement systems and to initiate system-wide changes does not appear to be a good use of resources. Thus, training will need to be focused on skills, be practice-based, utilise existing opportunities, and be supplemented by other interventions or systematic processes (for example, reminder systems). It appears that training-based interventions perform best when they are implemented alongside other interventions. Multi-component (eight studies): multi-component interventions, based on training, have been shown to be effective in increasing the frequency with which primary care clinicians intervene with their patients who smoke. However, no clear patterns emerged regarding which combinations of intervention components were more or less effective or what relative contributions the individual components might play. Simply increasing the number of intervention components delivered within a programme does not necessarily increase its overall performance. Reminders (five studies): all the trials of reminders demonstrated an increase in the frequency of health professional intervention. Programmes were found to be generally effective in changing health care professionals’ advice-giving rates and counselling activities. Sharing the responsibility for office-based smoking cessation activities among all staff appears to be a promising strategy and addressing the barrier of ‘clinician forgetfulness’ is one probable core constituent. Financial Incentives (three studies): the studies of financial incentives suggest that performance targets coupled with financial incentives do appear to change HCPs’ clinical behaviours; however, there is insufficient evidence to determine what the optimal performance threshold levels and/or the optimum level of incentives might be. In addition, the possibility of unintended consequences has not been fully explored and it is also unclear whether or not the apparent positive influence of financial incentives on clinical practice could be sustained over time. Audit & Feedback (three studies): audit & feedback appears to be an effective means of changing health care professionals’ behaviour. Audit & Feedback requires the widespread availability of electronic medical record systems capable of quickly and easily generating comparative feedback reports. However, the optimal methods of delivering such feedback are not fully understood. Other (two studies): there is insufficient evidence to draw robust conclusions for other intervention types. However, the results from two studies suggest that other low-intensity single component interventions are likely to be of limited value.

112


Contextual limitations and policy implications It is difficult to compare study contexts and settings exactly. Studies were selected on the basis that they reflected the New Zealand context/setting as much as possible: at the level of the patient-provider interaction. An exact match is not possible and no studies were identified that were conducted in New Zealand, with few studies being conducted in Australia. A large proportion of the research was conducted in the U.S.A. and differences in system and organisational aspects of care may have shaped the findings. Generalisability of findings to the New Zealand context should therefore be measured on a case-by-case basis with consideration being given to the possible influences of contextual, environmental and high-level policy factors. The interventions considered in this review are predominantly ‘office-based’, with the primary care clinicians being the point of contact for the patients needing care. Therefore, the reader should judge generalisability with respect to the nature of the ‘physician-patient encounter’ as well as the wider New Zealand context. Several authors question the effectiveness of interventions aimed at increasing the delivery of effective smoking cessation treatments outside of controlled clinical trial settings. That is, their feasibility and sustainability, when established in typical primary care settings– removed from the external influences of researcher expertise, funding and other resources. Solberg et al. (2000) highlighted the significant leadership engagement and commitment required; Ozer et al. (2001) acknowledged that the results of their trial were contingent on providing a health educator to the clinics at no cost; the Krist et al. (2008) electronic referral system (eLINKs) depleted available funding for free cessation counselling in just five weeks; and Moher, et al. (2001) acknowledged the possibly limited capacity of the control group practices to even deliver ‘usual care’ smoking cessation treatments, as they simply may not have had enough nurses. While questions of generalisability are applicable to all clinical trials, arguably, this may be more problematic when the intervention is ‘operator dependant’, that is, dependent (at least in part) on the attitudes, enthusiasm, competence and motivation of HCPs, funders, and other stakeholders. Policy-makers need to consider the implications associated with any changes in clinical practice and assess their likely impact on recourse utilisation and/or any subsequent unintended consequences. As Krist et al. (2008) demonstrated, any system that increases the rates of referral of smokers to speciality counselling services is contingent on these services being available and adequately resourced. Obviously, all preventive activities must compete with the other pressing demands on general practitioners and providing patients with quality smoking cessation support is no exception. Intensive strategies that entail multiple follow-up visits may make unrealistic demands on limited clinical time in primary care settings. Team approaches, in which the different tasks of the ABC approach are shared between different health professionals and organisations offer one possible solution to this problem. Other issues: Patient outcomes As has been stated, patient outcomes were not within the scope of this review. However, having assessed the eligibility of 464 resources and appraised and extracted data from the 42 included-studies, one obvious trend could not be overlooked – that of improved process outcomes but little strong evidence to suggest that the interventions lead to changes in patients’ smoking behaviour. Only 11 of 37 primary studies included patient outcomes (studies reporting only patient outcomes being specifically excluded), but of the 11 primary studies that did report patient outcomes, eight (73%) failed to demonstrate a statistically significant improvement in patient smoking outcomes (Goldstein, et al., 2003; Lennox, et al., 1998; Piper, et al., 2003; Richmond, et al., 1998; Strecher, et al., 1991; Unrod, et al., 2007; Young, et al., 2002). Similarly, the systematic review by Lancaster and Fowler (2000) found that six of the

113


eight (75%) included studies reporting quit rates failed to demonstrate any statistically significant improvements (although these were mainly older trials of training workshops and tutorials). Nevertheless, this lack of evidence is not evidence of lack of efficacy. In most cases, patient outcomes (e.g. quit rates) were not the primary outcomes of interest and some studies were not sufficiently powered to detect what can be relatively small, yet important effects. Three studies included in this review did report statistically significant and clinically important improvements in quit rates: one trial of a training intervention (Morgan, et al., 1996), one trial of a multi-component intervention (Katz, et al., 2004), and one trial of audit and feedback (Wadland, et al., 2007). In summary, several study authors (DePue, et al., 2002; Kottke, et al., 1992; Lancaster & Fowler, 2000; McRobbie, et al., 2008) state largely similar conclusions, that changing patients’ smoking behaviours is much more complicated than changing the behaviour of individual clinicians, and that influencing patient outcomes may require more intensive and sustained efforts, with multiple strategies, which are reinforced at higher policy levels. DePue et al. (2002) also point out that unfortunately, the majority of smokers are in lower-income populations, and they are more likely to receive medical care in settings that are less than ideal, thus the potential of effective interventions may not be fully realised.

114


115


Economic Considerations There is insufficient information to inform an economic analysis, given the multiple intervention types and/or components identified and the range of ‘ABC’ outcomes reported. This is complicated further by inconsistent and/or partial reporting of the outcomes and the use of non-standardised measurement instruments. This review summarises the evidence pertaining to a wide range of interventions and intervention components, rather than one ‘standard’ intervention, and the costs of these interventions are not known. It is reported extensively in the literature that patient-level smoking cessation interventions are among the most cost-effective interventions in healthcare – if not the ‘Gold Standard’. However, the cost-effectiveness of programmes aimed at increasing Health Care Professionals’ delivery of these smoking cessation interventions is not well documented. Estimating resource utilisation and any possible cost off-sets and/or savings to the health care system remains beyond the scope of this report.

116


117


Summary and Conclusions It is clear that increasing the delivery of effective smoking cessation treatments in primary care settings is an important if not essential step in achieving better patient care, but much work is still needed to translate such achievements into improved patient health. More New Zealand specific studies are needed to assimilate and integrate relevant information that may assist in increasing the delivery of effective smoking cessation treatments across the range of settings in New Zealand. While the literature indicates that several interventions offer much promise, it is recommended that the limitations of the research base be given due consideration and that any issues of generalisability be carefully evaluated before the implementation of any programme that aims to enhance smoking cessation practices in primary care. Appropriate prioritisation of resources is necessary when considering preventive services in general. It is known that changes in health care outcomes tend to lag behind those in physician performance and this trend has been observed here. Although beyond the scope of this review, obvious barriers include patients not accepting health care professionals’ recommendations, the socioeconomic and educational status of patients, other environmental influences, and the effectiveness of the clinical interventions themselves. Notwithstanding, smoking remains an important focus, as stopping smoking is probably the best way an individual can improve his or her own health. The literature reviewed suggests that health care professionals who are properly trained, supported, resourced, prompted, given personalised feedback are not subject to disincentives, and for whom the task is made uncomplicated, are more likely to deliver smoking cessation interventions to all their patients who smoke when compared to health care professionals delivering usual care.

118


119


References Amundson, G., Solberg, L. I., Reed, M., Martini, E. M., & Carlson, R. (2003). Paying for

quality improvement: compliance with tobacco cessation guidelines. Joint Commisson Journal on Quality and Safety, 29(2), 59-65.

Anderson, P., & Jané-Llopis, E. (2004). How can we increase the involvement of primary health care in the treatment of tobacco dependence? A meta-analysis. Addiction, 99(3), 299-312.

Bentz, C. J., Bayley, K. B., Bonin, K. E., Fleming, L., Hollis, J. F., Hunt, J. S., et al. (2007). Provider feedback to improve 5A's tobacco cessation in primary care: A cluster randomized clinical trial. Nicotine and Tobacco Research, 9(3), 341-349.

Borgiel, A. E. M., & Williams, J. I. (1999). Evaluating the effectiveness of 2 educational interventions in family practice. CMAJ: Canadian Medical Association Journal, 161(8), 965.

Brown, D., & Schulte, A. C. (1987). A social learning model of consultation. Professional psychology, research and practice, 18(3), 283-287.

Centers for Disease Control and Prevention (1992). Public Health Focus: Effectiveness of Smoking-Control Strategies -- United States. MMWR Morbidity and Mortality Weekly Report, 41(35), 645-647,653.

Chang, H. C., Zimmerman, L. H., & Beck, J. M. (1995). Impact of chart reminders on smoking cessation practices of pulmonary physicians. American Journal of Respiratory and Critical Care Medicine, 152(3), 984-987.

Cockburn, J., Ruth, D., Silagy, C., Dobbin, M., Reid, Y., Scollo, M., et al. (1992). Randomised trial of three approaches for marketing smoking cessation programmes to Australian general practitioners. British Medical Journal, 304(6828), 691-694.

Cohen, S. J., Stookey, G. K., Katz, B. P., Drook, C. A., & Smith, D. M. (1989). Encouraging primary care physicians to help smokers quit. A randomized, controlled trial. Ann Intern Med, 110(8), 648-652.

Coleman, T., Lewis, S., Hubbard, R., & Smith, C. (2007). Impact of contractual financial incentives on the ascertainment and management of smoking in primary care. Addiction, 102(5), 803-808.

Cummings, S. R., Coates, T. J., Richard, R. J., Hansen, B., Zahnd, E. G., VanderMartin, R., et al. (1989). Training physicians in counseling about smoking cessation. A randomized trial of the "Quit for Life" program. Ann Intern Med, 110(8), 640-647.

Cummings, S. R., Richard, R. J., Duncan, C. L., Hansen, B., Vander, M. R., Gerbert, B., et al. (1989). Training physicians about smoking cessation: a controlled trial in private practice. Journal of general internal medicine : official journal of the Society for Research and Education in Primary Care Internal Medicine, 4, 482-489.

Cummings, S. R., Rubin, S. M., & Oster, G. (1989). The cost-effectiveness of counseling smokers to quit. Jama, 261(1), 75-79.

DePue, J. D., Goldstein, M. G., Schilling, A., Reiss, P., Papandonatos, G., Sciamanna, C., et al. (2002). Dissemination of the AHCPR clinical practice guideline in community health centres. (Research Paper). Tobacco Control, 11(4), 329(327).

Dickey, L. L., Gemson, D. H., & Carney, P. (1999). Office system interventions supporting primary care-based health behavior change counseling. American Journal of Preventive Medicine, 17(4), 299-308.

120


Dietrich, A. J., O'Connor, G. T., Keller, A., Carney, P. A., Levy, D., & Whaley, F. S. (1992). Cancer: improving early detection and prevention. A community practice randomised trial. British Medical Journal, 304(6828), 687-691.

Doll, R., Peto, R., Boreham, J., & Sutherland, I. (2004). Mortality in relation to smoking: 50 years observations on male British doctors. British Medical Journal(328), 1519.

Edwards, R., Wilson, N., & Thomson, G. (2008). Tobacco control in New Zealand: top marks for new smoking cessation guidelines, must try harder elsewhere. The New Zealand Medical Journal, 121(1276), 5-8.

Egger, M., Dickersin, K., & Davey Smith, G. (2001 ). Problems and limitations in conducting systematic reviews. In Systematic Reviews in Health Care: Meta-analysis in Context. London, UK: BMJ Publishing Group.

Fiore, M. C., Bailey, W. C., Cohen, S. J., & et al. (1996). Smoking Cessation: Clinical Practice Guideline No 18 (Vol. AHCPR Publ No 96-0692). Rockville, Md: US Department of Health and Human Services, Public Health Service, Agency for Health Care Policy and Research.

Fiore, M. C., Jaén, C. R., Baker, T. B., Bailey, W. C., Benowitz, N. L., Curry, S. J., et al. (2008). Treating tobacco use and dependence:2008 update. from U.S. Department of Health and Human Services. Public Health Service: http://www.surgeongeneral.gov/tobacco/treating_tobacco_use08.pdf

Fiore, M. C., Jorenby, D. E., Schensky, A. E., Smith, S. S., Bauer, R. R., & Baker, T. B. (1995). Smoking status as the new vital sign: Effect on assessment and intervention in patients who smoke. Mayo Clinic Proceedings, 70(3), 209-213.

Flocke, S. A., Gordon, L. E., & Pomiecko, G. L. (2006). Evaluation of a Community Health Promotion Resource for Primary Care Practices. American Journal of Preventive Medicine, 30(3), 243-251.

Glasgow, R., Lichtenstein, E., & Marcus, A. (2003). Why don’t we see more translation of health promotion research to practice? Rethinking the efficacy-to-effectiveness transition. American Journal of Public Health(93), 1261-1267.

Glynn, T. J., Manley, M. W., & Pechacek, T. F. (1990). Physician-initiated smoking cessation program: the National Cancer Institute trials. Prog Clin Biol Res, 339, 11-25.

Goldstein, M. G., Niaura, R., Willey, C., Kazura, A., Rakowski, W., DePue, J., et al. (2003). An academic detailing intervention to disseminate physician-delivered smoking cessation counseling: smoking cessation outcomes of the Physicians Counseling Smokers Project. Preventive Medicine: An International Journal Devoted to Practice and Theory, 36(2), 185-196.

Green, L. W., & Kreuter, M. W. (1991). Health Promotion Planning: An Educational and Environmental Approach (2 ed.). Palo Alto: Mayfield Publishing Co.

Grimshaw, J., Eccles, M., & Tetroe, M. A. (2004). Implementing Clinical Guidelines: Current Evidence and Future Implications. The Journal of Continuing Education in the Health Professions, 24, S31-S37.

Grimshaw, J., Thomas, R. E., MacLennan, G., Fraser, C., Ramsay, C. R., Vale, L., et al. (2004). Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technology Assessment, 8(6), iii-iv, 1-72.

Hogg, W., Lemelin, J., Graham, I. D., Grimshaw, J., Martin, C., Moore, L., et al. (2008). Improving prevention in primary care: Evaluating the effectiveness of outreach facilitation. Family Practice, 25(1), 40-48.

Joseph, A. M., Arikian, N. J., An, L. C., Nugent, S. M., Sloan, R. J., & Pieper, C. F. (2004). Results of a Randomized Controlled Trial of Intervention to Implement Smoking

121


Guidelines in Veterans Affairs Medical Centers: Increased Use of Medications Without Cessation Benefit. Medical Care, 42(11), 1100-1110.

Katz, D. A., Muehlenbruch, D. R., Brown, R. L., Fiore, M. C., & Baker, T. B. (2004). Effectiveness of Implementing the Agency for Healthcare Research and Quality Smoking Cessation Clinical Practice Guideline: A Randomized, Controlled Trial. Journal of the National Cancer Institute, 96(8), 594-603.

Kottke, T. E., Solberg, L. I., Brekke, M. L., Conn, S. A., Maxwell, P., & Brekke, M. J. (1992). A controlled trial to integrate smoking cessation advice into primary care practice: Doctors Helping Smokers, Round III. The Journal of Family Practice, 34(6), 701-708.

Krist, A. H., Woolf, S. H., Frazier, C. O., Johnson, R. E., Rothemich, S. F., Wilson, D. B., et al. (2008). An Electronic Linkage System for Health Behavior Counseling. Effect on Delivery of the 5A's. American Journal of Preventive Medicine, 35(5 SUPPL.), S350-S358.

Lancaster, T., & Fowler, G. (2000). Training health professionals in smoking cessation. Cochrane Database Systematic Reviews(3), Art. No.: CD000214. DOI: 000210.001002/14651858.

Lemelin, J., Hogg, W., & Baskerville, N. (2001). Evidence to action: a tailored multifaceted approach to changing family physician practice patterns and improving preventive care. CMAJ: Canadian Medical Association Journal, 164(6), 757.

Lennox, A. S., Bain, N., Taylor, R. J., McKie, L., Donnan, P. T., & Groves, J. (1998). Stages of Change training for opportunistic smoking intervention by the primary health care team. Part I : randomised controlled trial of the effect of training on patient smoking outcomes and health professional behaviour as recalled by patients. Health Education Journal, 57(2), 140-149.

Litt, J. (2005). Smoking and GPs: time to cough up: successful interventions in general practice. Australian family physician, 34(6), 425-429.

McPhee, S. J., Bird, J. A., Fordham, D., Rodnick, J. E., & Osborn, E. H. (1991). Promoting cancer prevention activities by primary care physicians. Results of a randomized, controlled trial. Journal of the American Medical Association, 266(4), 538-544.

McRobbie, H., Hajek, P., Feder, G., & Eldridge, S. (2008). A cluster-randomised controlled trial of a brief training session to facilitate general practitioner referral to smoking cessation treatment. Tobacco Control, 17(3), 173-176.

Millett, C., Gray, J., Saxena, S., Netuveli, G., & Majeed, A. (2007). Impact of a pay-for-performance incentive on support for smoking cessation and on smoking prevalence among people with diabetes. Canadian Medical Association Journal, 176(12), 1705-1710.

Ministry of Health (2004). Clearing the Smoke: A five-year plan for tobacco control in New Zealand (2004-2009). Wellington: Ministry of Health.

Ministry of Health (2006). Tobacco Trends 2006: Monitoring tobacco use in New Zealand. Wellington: Ministry of Health.

Ministry of Health (2007a). New Zealand Smoking Cessation Guidelines. Wellington: Ministry of Health.

Ministry of Health (2007b). New Zealand Tobacco Use Survey 2006. Wellington: Ministry of Health.

Ministry of Health (2008). Literature Review for the Revision of the New Zealand Smoking Cessation Guidelines. Wellington: Ministry of Health.

122


Ministry of Health (2009). Implementing the ABC approach for smoking cessation: framework and work programme. Wellington: Ministry of Health.

Moher, M., Yudkin, P., Wright, L., Turner, R., Fuller, A., Schofield, T., et al. (2001). Cluster randomised controlled trial to compare three methods of promoting secondary prevention of coronary heart disease in primary. BMJ, 322(7298), 1338-1342.

Morgan, G. D., Noll, E. L., Orleans, C. T., Rimer, B. K., Amfoh, K., & Bonney, G. (1996). Reaching midlife and older smokers: Tailored interventions for routine medical care. Preventive Medicine, 25(3), 346-354.

Mulrow, C. D., & Oxman, A. (1997). Cochrane Collaboration Handbook. Oxford: The Cochrane Collaboration.

National Health and Medical Research Council (1999). How to use the evidence: assessment and application of scientific evidence. . Canberra: NHMRC.

New Zealand Ministry of Health, & King, A. (2000). The New Zealand health strategy. Wellington, N.Z.: Ministry of Health.

NHMRC (2008). Additional levels of evidence and grades for recommendations for developers of guidelines. Canberra: NHMRC.

O’Dea, D. (2004). An economic evaluation of the Quitline nicotine replacement therapy (NRT) service. Wellington: Ministry of Health.

O’Loughlin, J., Makni, H., Tremblay, M., & Karp, I. (2007). Gender differences among general practitioners in smoking cessation counseling practices. Preventive Medicine 45, 208-214.

Ockene, J. K., Adams, A., Pbert, L., Luippold, R., Hebert, J. R., Quirk, M., et al. (1994). The Physician-Delivered Smoking Intervention Project: factors that determine how much the physician intervenes with smokers. Journal of General Internal Medicine, 9(7), 379-384.

Ockene, J. K., Lindsay, E. A., Hymowitz, N., Giffen, C., Purcell, T., Pomrehn, P., et al. (1997). Tobacco control activities of primary-care physicians in the Community Intervention Trial for Smoking Cessation. COMMIT Research Group. Tobacco Control, 6 Suppl 2, S49-S56.

Oxman, A. D., Thomson, M. A., Davis, D. A., & Haynes, R. B. (1995). No magic bullets: a systematic review of 102 trials of interventions to improve professional practice. Cmaj, 153(10), 1423-1431.

Ozer, E. M., Adams, S. H., Lustig, J. L., Millstein, S. G., Camfield, K., El Diwany, S., et al. (2001). Can it be done? Implementing adolescent clinical preventive services. Health services research, 36(6 Pt 2), 150-165.

Pbert, L., Fletcher, K. E., Flint, A. J., Young, M. H., Druker, S., & DiFranza, J. (2006). Smoking prevention and cessation intervention delivery by pediatric providers, as assessed with patient exit interviews. Pediatrics, 118(3), e810-824.

Peto, R., Darby, S., Deo, H., Silcocks, P., Whitley, E., & Doll, R. (2000). Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. British Medical Journal, 321, 323-329.

Piper, M. E., Fiore, M. C., Smith, S. S., Jorenby, D. E., Wilson, J. R., Zehner, M. E., et al. (2003). Use of the vital sign stamp as a systematic screening tool to promote smoking cessation. Mayo Clin Proc, 78(6), 716-722.

Public Health Intelligence (2004). Tobacco Facts. Occasional Report. Wellington: Ministry of Health.

123


Richmond, R., Mendelsohn, C., & Kehoe, L. (1998). Family Physicians' Utilization of a Brief Smoking Cessation Program Following Reinforcement Contact after Training: A Randomized Trial. Preventive Medicine, 27(1), 77-83.

Roski, J., Jeddeloh, R., An, L., Lando, H., Hannan, P., Hall, C., et al. (2003). The impact of financial incentives and a patient registry on preventive care quality: increasing provider adherence to evidence-based smoking cessation practice guidelines. Preventive Medicine, 36(3), 291-299.

Rothemich, S. F., Woolf, S. H., Johnson, R. E., Burgett, A. E., Flores, S. K., Marsland, D. W., et al. (2008). Effect on cessation counseling of documenting smoking status as a routine vital sign: An ACORN study. Annals of Family Medicine, 6(1), 60-68.

Secker-Walker, R. H., Solomon, L. J., Flynn, B. S., LePage, S. S., Crammond, J. E., Worden, J. K., et al. (1992). Training obstetric and family practice residents to give smoking cessation advice during prenatal care. American Journal of Obstetrics and Gynecology, 166(5), 1356-1363.

Solberg, L. I., Kottke, T. E., Brekke, M. L., Magnan, S., Davidson, G., Calomeni, C. A., et al. (2000). Failure of a continuous quality improvement intervention to increase the delivery of preventive services. A randomized trial. Effective Clinical Practice, 3(3), 105-115.

Soumerai, S. B., McLaughlin, T. J., & Avorn, J. (1989). Improving drug prescribing in primary care: a critical analysis of the experimental literature. The Milbank quarterly, 67(2), 268-317.

Strecher, V. J., O'Malley, M. S., Villagra, V. G., Campbell, E. E., Gonzalez, J. J., Irons, T. G., et al. (1991). Can residents be trained to counsel patients about quitting smoking? Results from a randomized trial. Journal of General Internal Medicine, 6(1), 9-17.

The AGREE Collaboration (2003). Appraisal of Guidelines for Research & Evaluation (AGREE) Agree Instrument training manual, the AGREE instrument: short form. Retrieved 22.04.2009: www.agreecollaboration.org

Tremblay, M., Gervais, A., Lacroix, C., O'Loughlin, J., Makni, H., & Paradis, G. (2001). Physicians taking action against smoking: An intervention program to optimize smoking cessation counselling by Montreal general practitioners. Canadian Medical Association Journal, 165(5), 601-607.

Tsevat, J. (1992). Impact and cost-effectiveness of smoking interventions. Am J Med, 93(1A), 43S-47S.

Unrod, M., Smith, M., Spring, B., DePue, J., Redd, W., & Winkel, G. (2007). Randomized controlled trial of a computer-based, tailored intervention to increase smoking cessation counseling by primary care physicians. Journal of General Internal Medicine, 22(4), 478-484.

Wadland, W. C., Holtrop, J. S., Weismantel, D., Pathak, P. K., Fadel, H., & Powell, J. (2007). Practice-based referrals to a tobacco cessation quit line: assessing the impact of comparative feedback vs general reminders. Annals of Family Medicine, 5, 135-142.

West, R. (2005). Time for a change: putting the Transtheoretical (Stages of Change) Model to rest. Addiction, 100(8), 1036–1039.

Wilson, N., Thomson, G., & Edwards, R. (2008). Use of four major tobacco control interventions in New Zealand: a review. The New Zealand Medical Journal, 121(1276).

Yarnall, K. S., Pollak, K. I., Østbye, T., Krause, K. M., & Lloyd Michener, J. (2003). Primary Care: Is There Enough Time for Prevention? American Journal of Public Health, 93(4).

124


Young, J. M., D'Este, C., & Ward, J. E. (2002). Improving family physicians' use of evidence-based smoking cessation strategies: A cluster randomization trial. Preventive Medicine, 35(6), 572-583.

125


Glossary Absolute risk reduction (ARR) The difference between adverse outcomes in the treated and placebo groups in a clinical trial. Absolute risk The observed or the calculated risk of an event in a population. Age standardisation A procedure for adjusting rates designed to minimise the effects of differences in age composition when comparing rates for different populations. Analysis of variance (ANOVA) A statistical analysis involving the comparison of variance reflecting different sources of variability. A priori Not based on prior study or examination. Before and after study A situation in which the investigator compares outcomes before and after the introduction of a new intervention. Bias Deviation of results or inferences from the truth, or processes leading to such deviation. Any trend in the collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are systematically different from the truth. Blinded study A study in which observers and/or subjects are kept ignorant of the group to which they are assigned. When both observers and subjects are kept ignorant, the study is referred to as double blind. Capitation A system of paying providers a defined price for each consumer who is registered with a provider. Case control study An epidemiological study involving the observation of cases (persons with the disease, such as cervical cancer) and a suitable control (comparison, reference) group of persons without the disease. The relationship of an attribute to the disease is examined by comparing retrospectively the past history of the people in the two groups with regard to how frequently the attribute is present. See also nested case control. Case series A descriptive study of a subset of a defined population (i.e. a single patient or group of patients) which aims to describe the association between factors or attributes which the sample are exposed to, and the probability of occurrence of a given disease or other outcome. Case series are collections of individual case reports, which may occur within a fairly short period of time. Cohen classification A system that categorises the pooled effect size into small, medium and large categories. A small effect accounts for ≤ 1% of the variance of the population, a medium effect accounts for 1% to ≤ 5.9% of the population variance, and a large effect accounts for between 5.9% and 13.8% of the variance. Cohort study The analytic method of epidemiologic study in which subsets of a defined population can be identified who are, have been, or in the future may be exposed or not exposed in different degrees, to a factor or factors hypothesised to influence the probability of occurrence of a given disease or other outcome. Studies usually involve the observation of a large population, for a prolonged period (years), or both. Community care Corresponds to ambulatory and domiciliary services provided other than through hospitals to patients who are resident at their home, hotel, prison, barrack, etc.

126


Confidence interval The computed interval with a given probability, e.g. 95%, that the true value of a variable such as a mean, proportion, or rate is contained within the interval. The 95% CI is the range of values in which it is 95% certain that the true value lies for the whole population. Confounder A third variable that indirectly distorts the relationship between two other variables, because it is independently associated with each of the variables. Confounding A situation in which the measure of the effect of an exposure on risk is distorted because of the association of exposure with other factor(s) that influence the outcome under study. Cost benefit analysis An economic analysis in which the costs of medical care and the benefits of reduced loss of net earnings due to preventing premature death or disability are considered. Cost effectiveness (CE) Involves the relationship between costs and effects, providing information on whether a technology is being delivered to those who would benefit from it with an optimal use of resources. It is expressed as a ratio of the effects (number of lives saved, number of disability days avoided) obtained for a specific cost (expressed in dollars). For example, the numerator may be the difference in lifetime costs between one intervention and another, while the denominator may be the difference in life expectancies associated with the two interventions. Low cost effectiveness ratios are desirable. Cost minimisation analysis A particular type of cost effectiveness analysis in which it is assumed the outcome is the same in all comparison groups. The focus is, therefore, on the comparative costs of different interventions. Coverage The number, percent, or proportion of eligible people reached by a programme. Cross-sectional study A study that examines the relationship between diseases (or other health related characteristics), and other variables of interest as they exist in a defined population at one particular time. Descriptive study A study concerned with, and designed only to describe the existing distribution of variables, without regard to causal or other hypotheses. Ecological study A study in which the units of analysis are populations or groups of people, rather than individuals. Effectiveness A measure of the extent to which a specific intervention, procedure, regimen, or service, when deployed in the field in routine circumstances, does what it is intended to do for a specified population. Efficiency The effects or end results achieved in relation to the effort expended in terms of money, resources and time. The extent to which the resources used to provide a specific intervention, procedure, regimen, or service of known efficacy and effectiveness are minimised. Evidence based Based on valid empirical information. Generalisability Applicability of the results to other populations. Grey literature That which is produced by all levels of government, academics, business and industry, in print and electronic formats, but which is not controlled by commercial publishers.

127


High risk groups Usually refers to groups of people who have been identified as having a higher than expected, or higher than average for the population as a whole, incidence of the disease in question. Hospitalisations A term used as an indicator of morbidity of diseases in a community. A hospitalisation in New Zealand health statistics includes inpatients who leave hospital to return home, transfer to another hospital or institution, or die in hospital after formal admission. That is a count of episodes of care rather than individuals. Incidence The number of new events (cases; e.g. of disease) occurring during a certain period, in a specified population. Indicator An item of quantitative or qualitative information reported to enable the monitoring of a condition or the performance of an organisation. Inpatient A person admitted to hospital for medical, surgical or psychiatric treatment, observation or care, who spends at least one night in the hospital. A healthy person accompanying a sick person is included if formally admitted as a boarder. Intention to treat A method for data analysis in a randomised controlled trial in which individual outcomes are analysed according to the group to which they were randomised, even if they never received the treatment to which they were assigned. Intention to treat analysis A method for data analysis in a randomised controlled trial in which individual outcomes are analysed according to the group to which they were randomised, even if they never received the treatment to which they were assigned. Managed care organisation An organisation or service provider which is given responsibility for ensuring that a defined population receives a defined set of services in a co-ordinated way. Matching The process of making a study group and a comparison group comparable with respect to extraneous factors. Mean Calculated by adding all the individual values in the group and dividing by the number of values in the group. Median Any value that divides the probability distribution of a random variable in half. For a finite population or sample the median is the middle value of an odd number of values (arranged in ascending order) or any value between the two middle values of an even number of values. Meta-analysis The process of using statistical methods to combine the results of different studies. The systematic and organised evaluation of a problem, using information from a number of independent studies of the problem. Misclassification The erroneous classification of an individual, a value, or an attribute into a category other than that to which it should be assigned. Morbidity Illness. Mortality The number of deaths from a specified disease which are diagnosed or reported during a defined period of time in a given population. Multiple regression Any analysis of data that takes into account a number of variables simultaneously. National Minimum Dataset Data on specific procedures required from service providers by the New Zealand Ministry of Health.

128


Natural history The course of a disease from onset to resolution. Number needed to treat (NNT) The number of patients who need to be treated to achieve one additional favourable outcome. Calculated as 1/ARR. If the intervention harmed people, the term would be the number needed to harm. Odds ratio (OR) A measure of the degree or strength of an association. In a case control or a cross-sectional study, it is measured as the ratio of the odds of exposure (or disease) among the cases to that among the controls. OECD Organisation for Economic Co-operation and Development. There are 24 countries in the OECD. Outpatient A person who goes to a health care facility for a consultation, and who leaves the facility within three hours of the start of the consultation. An outpatient is not formally admitted to the facility. Power The ability of a study to demonstrate an association if one exists. Prevalence The number of events in a given population at a designated time (point prevalence) or during a specified period (period prevalence). Primary care First contact, continuous, comprehensive and coordinated care provided to individuals and populations undifferentiated by age, gender, disease or organ system. Providers Organisations and health professionals providing health services. Random sample A sample that is arrived at by selecting sample units such that each possible unit has a fixed and determinate probability of selection. Randomised controlled trial An epidemiologic experiment in which subjects in a population are randomly allocated into groups to receive or not receive an experimental, preventive or therapeutic procedure, manoeuvre, or intervention. Randomised controlled trials are generally regarded as the most scientifically rigorous method of hypothesis testing available in epidemiology. Recall bias Systematic bias due to differences in accuracy or completeness of recall or memory of past events or experiences. Reference standard An independently applied test that is compared to a screening or diagnostic test being evaluated in order to verify the latter’s accuracy. A reference standard, therefore, provides an accurate or “truth” diagnosis for verification of positive and negative diagnoses. It is sometimes described as providing “final truth determination”. Relative risk (RR) The ratio of the risk of disease or death among the exposed to the risk among the unexposed. It is a measure of the strength or degree of association applicable to cohort studies and RCTs. Relative risk reduction (RRR) The proportional reduction in rates of bad events between experimental and control participants in a trial. If there was an increase in the rate of bad events, the term would then be relative risk increase. Risk factor An exposure or aspect of personal behaviour or lifestyle, which on the basis of epidemiologic evidence is associated with a health-related condition. Secondary care Surgical and medical services that are generally provided in a hospital setting. In many cases, access to these services is by referral from a primary care health professional such as a general practitioner.

129


Selection bias Error due to systematic differences in characteristics between those who are selected for inclusion in a study and those who are not (or between those compared within a study and those who are not). Sensitivity analysis A method to determine the robustness of an assessment by examining the extent to which results are affected by changes in methods, values of variables, or assumptions. Sensitivity Sensitivity is the proportion of truly diseased persons in a screened population who are identified as diseased by a screening test. Sensitivity is a measure of the probability of correctly diagnosing a case, or the probability that any given case will be identified by the test. Specificity The proportion of truly non-diseased persons who are so identified by a screening test. It is a measure of the probability of correctly identifying a non-diseased person with a screening test. Standardised mortality ratio (SMR) The ratio of the number of deaths observed in the study group or population to the number that would be expected if the study population had the same specific rates as the standard population, multiplied by 100. Systematic review Literature review reporting a systematic method to search for, identify and appraise a number of independent studies. Variance A measure of the variation shown by a set of observations, defined by the sum of the squares of deviation from the mean, divided by the number of degrees of freedom in the set of observations.

130


131


Appendix A: Working Party Membership Name and area of expertise Affiliation Karen Evison National Programme Manager Tobacco Policy & Implementation Team Sector Capability & Innovation Directorate

Ministry of Health

Dr Ashley Bloomfield Acting Deputy Director-General Sector Capability & Innovation Directorate

Ministry of Health

132


133


Appendix B: Search Strategy EMBASE.com ID. Query Results #1 'smoking'/exp OR 'smoking' 170,431 #2 'smoking cessation'/de 22,278 #3 'tobacco dependence'/de 6,109 #4 'nicotinism'/de 6,109 #5 #1 OR #2 OR #3 OR #4 171,900 #6 smoking:ti,ab OR smoker*:ti,ab OR 'tobacco addict':ti,ab OR 'tobacco addicts':ti,ab 124,323 #7 'nicotine abstination':ti,ab OR 'nicotine cessation':ti,ab 22 #8 'tobacco *1 cessation':ti,ab 699 #9 'nicotine abuse':ti,ab OR 'nicotine addiction':ti,ab OR 'nicotine dependence':ti,ab 2,806 #10 nicotinism:ti,ab OR tobaccoism:ti,ab 116 #11 'tobacco abuse':ti,ab OR 'tobacco addiction':ti,ab OR 'tobacco dependence':ti,ab 1,435 #12 'nicotine use disorder':ti,ab OR 'tobacco use disorder':ti,ab 17 #13 'quit *5 tobacco':ti,ab OR 'tobacco *5 quit':ti,ab 300 #14 #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 126,101 #15 #5 OR #14 178,542 #16 'health care personnel'/exp 523,305 #17 'physician'/de 80,451 #18 doctor*:ti,ab OR clinician*:ti,ab OR physician*:ti,ab OR practitioner*:ti,ab 434,312 #19 'health *1 practitioner':ti,ab OR 'health *1 professional':ti,ab 4,807 #20 'health *1 practitioners':ti,ab OR 'health *1 professionals':ti,ab 30,898 #21 'healthcare practitioner':ti,ab OR 'healthcare professional':ti,ab 491 #22 'healthcare practitioners':ti,ab OR 'healthcare professionals':ti,ab 4,322 #23 'health *1 provider':ti,ab OR 'healthcare provider':ti,ab 3,655 #24 'health *1 providers':ti,ab OR 'healthcare providers':ti,ab 18,016 #25 'health *1 personnel':ti,ab OR 'healthcare *1 personnel':ti,ab 4,848 #26 'health *1 workers':ti,ab OR 'healthcare *1 worker':ti,ab 12,788 #27 #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 876,681 #28 'primary health care'/exp 66,345 #29 'family medicine'/de 4,809 #30 'general practice'/de 56,134 #31 'medical practice'/de 53,339 #32 'medical care'/de 14,480 #33 'primary *1 care':ti,ab OR 'first line care':ti,ab 61,758 #34 'family medicine':ti,ab OR 'family practice':ti,ab 10,854 #35 'general *1 practice':ti,ab OR 'medical practice':ti,ab 39,973 #36 'general *1 medicine':ti,ab 4,971 #37 'medical assistance':ti,ab OR 'medical care':ti,ab 32,599 #38 #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 247,459 #39 #27 AND #38 113,171 #40 'general practitioner'/de 39,137 #41 gp:ti,ab OR 'general practitioner':ti,ab OR 'general physician':ti,ab 50,274 #42 gps:ti,ab OR 'general practitioners':ti,ab OR 'general physicians':ti,ab 28,792 #43 'family doctor':ti,ab OR 'family physician':ti,ab OR 'family practitioner':ti,ab 6,366 #44 'family doctors':ti,ab OR 'family physicians':ti,ab OR 'family practitioners':ti,ab 9,172 #45 #39 OR #40 OR #41 OR #42 OR #43 OR #44 182,690 #46 #15 AND #45 4,070 #47 abc:ti,ab OR 'ask brief *1 cessation':ti,ab 12,129 #48 5a:ti,ab OR '5 a':ti,ab OR '5 as':ti,ab OR 5as:ti,ab 37,016

134


#49 'ask advise assess assist *1 arrange':ti,ab 29 #50 'ask *3 smoking':ti,ab OR 'asking *3 smoking':ti,ab 73 #51 'asks *3 smoking':ti,ab OR 'asked *3 smoking':ti,ab 162 #52 'advise *3 quit':ti,ab OR 'advice *3 quit':ti,ab 205 #53 'advises *3 quit':ti,ab OR 'advised *3 quit':ti,ab OR 'advising *3 quit':ti,ab 142 #54 'brief advice':ti,ab OR 'brief *1 counseling':ti,ab OR 'brief *1 counselling':ti,ab 472 #55 'assess *3 quit':ti,ab OR 'assesses *3 quit':ti,ab OR 'assessed *5 quit':ti,ab 52 #56 'assessment *3 quit':ti,ab OR 'assessing *3 quit':ti,ab 13 #57 'assist *3 cessation':ti,ab OR 'assisted *3 cessation':ti,ab 91 #58 'assisting *3 cessation':ti,ab OR 'assists *3 cessation':ti,ab 24 #59 'assistance *3 cessation':ti,ab 31 #60 'arrange *3 follow up':ti,ab OR 'arranged *3 follow up':ti,ab 111 #61 'arranges *3 follow up':ti,ab OR 'arranging *3 follow up':ti,ab 39 #62 'arrange *3 followup':ti,ab OR 'arranged *3 followup':ti,ab 2 #63 'arranges *3 followup':ti,ab OR 'arranging *3 followup':ti,ab 0 #64 'arrangement *3 followup':ti,ab OR 'arrangement *3 followup':ti,ab 1 #65 'cessation support':ti,ab OR 'cessation assistance':ti,ab 147 #66 'induce *3 cessation':ti,ab OR 'induces *3 cessation':ti,ab 51 #67 'induced *3 cessation':ti,ab OR 'inducing *3 cessation':ti,ab 209 #68 'inducement *3 cessation':ti,ab OR 'induction *3 cessation':ti,ab 20 #69 'cessation counseling':ti,ab OR 'cessation counselling':ti,ab 533 #70 'smoking status':ti,ab 6,674

#71 #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 OR #57 OR #58 OR #59 OR #60 OR #61 OR #62 OR #63 OR #64 OR #65 OR #66 OR #67 OR #68 OR #69 OR #70

57,564

#72 #46 AND #71 507 #73 'intervention study'/de 4,399 #74 intervention*:ti,ab 354,616 #75 #73 OR #74 355,681 #76 #72 AND #75 218 #77 #46 AND #75 981 #78 'smoking'/exp/dm_di 1 #79 'smoking'/exp/dm_pc 6,002 #80 'tobacco dependence'/dm_di 283 #81 'tobacco dependence'/dm_pc 874 #82 #78 OR #79 OR #80 OR #81 6,912 #83 #77 AND #82 76 #84 'smoking'/exp/dm_th 214 #85 'smoking'/exp/dm_dt 141 #86 'tobacco dependence'/dm_th 1,057 #87 'tobacco dependence'/dm_dt 1,725 #88 #84 OR #85 OR #86 OR #87 2,501 #89 #77 AND #88 50 #90 ((smok*:ti OR tobacco:ti) AND intervention*:ti) 1,257 #91 #46 AND #90 132 #92 #76 OR #83 OR #89 OR #91 359 #93 #76 OR #83 OR #89 OR #91 AND [1980-2009]/py 359 #94 #76 OR #83 OR #89 OR #91 AND [english]/lim AND [1980-2009]/py 323

135


The Cochrane Library ID Search Results #1 MeSH descriptor Smoking explode all trees 3920 #2 MeSH descriptor Tobacco Use Cessation explode all trees 1866 #3 MeSH descriptor Tobacco Use Disorder explode all trees 417 #4 (smoking or smoker* or "tobacco addict" or "tobacco addicts"):ti,ab,kw 10436 #5 ("nicotine abstination" or "nicotine cessation"):ti,ab,kw 0 #6 (tobacco near cessation):ti,ab,kw 440 #7 ("nicotine abuse" or "nicotine addiction" or "nicotine dependence"):ti,ab,kw 219 #8 (nicotinism or tobaccoism):ti,ab,kw 2 #9 ("tobacco abuse" or "tobacco addiction" or "tobacco dependence"):ti,ab,kw 145 #10 ("nicotine use disorder" or "tobacco use disorder"):ti,ab,kw 421 #11 (tobacco near quit):ti,ab,kw 46 #12 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11) 10582 #13 MeSH descriptor Health Personnel explode all trees 3840 #14 (doctor* OR clinician* OR physician* OR practitioner*):ti,ab,kw 18338 #15 ((health near/1 practitioner) OR (health near/1 professional)):ti,ab,kw 689 #16 ((health near/1 practitioners) OR (health near/1 professionals)):ti,ab,kw 689 #17 ("healthcare practitioner" OR "healthcare professional"):ti,ab,kw 23 #18 ("healthcare practitioners" OR "healthcare professionals"):ti,ab,kw 127 #19 ((health near/1 provider) OR "healthcare provider"):ti,ab,kw 87 #20 ((health near/1 providers) OR "healthcare providers"):ti,ab,kw 146 #21 ((health near/1 personnel) OR (healthcare near/1 personnel)):ti,ab,kw 1494 #22 ((health near/1 workers) OR (healthcare near/1 worker)):ti,ab,kw 335 #23 (#13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22) 21864 #24 MeSH descriptor Primary Health Care explode all trees 2311 #25 MeSH descriptor Family Practice explode all trees 1994 #26 ((primary near/1 care) OR "first line care"):ti,ab,kw 4303 #27 ("family medicine" OR "family practice"):ti,ab,kw 2479 #28 ((general near/1 practice) OR "Medical practice"):ti,ab,kw 3070 #29 (general near/1 medicine):ti,ab,kw 190 #30 ("medical assistance" OR "medical care"):ti,ab,kw 1287 #31 (#24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30) 9789 #32 (#23 AND #31) 4106 #33 MeSH descriptor Physicians, Family explode all trees 375 #34 (GP OR "general practitioner" OR "general physician"):ti,ab,kw 1810 #35 (GPs OR "general practitioners" OR "general physicians"):ti,ab,kw 1705

136


#36 ("family doctor" OR "family physician" OR "family practitioner"):ti,ab,kw 153 #37 ("family doctors" OR "family physicians" OR "family practitioners"):ti,ab,kw 365 #38 (#32 OR #33 OR #34 OR #35 OR #36 OR #37) 5901 #39 (#12 AND #38) 325 #40 (abc OR "ask brief near/1 cessation"):ti,ab,kw 266 #41 (5A OR "5 A" OR "5 As" OR 5As):ti,ab,kw 571 #42 (Ask Advise Assess Assist near/1 Arrange):ti,ab,kw 1 #43 ((Ask near/3 smoking) OR (Asking near/3 smoking)):ti,ab,kw 27 #44 ((Asks near/3 smoking) OR (Asked near/3 smoking)):ti,ab,kw 27 #45 ((Advise near/3 quit) OR (Advice near/3 quit)):ti,ab,kw 63 #46 ((Advises near/3 quit) OR (Advised near/3 quit) OR (Advising near/3 quit)):ti,ab,kw 20 #47 ("brief advice" OR (brief near/1 counseling) OR (brief near/1 counselling)):ti,ab,kw 159 #48 ((Assess near/3 quit) OR (Assesses near/3 quit) OR (Assessed near quit)):ti,ab,kw 28 #49 ((Assessment near/3 quit) OR (Assessing near/3 quit)):ti,ab,kw 18 #50 ((Assist near/3 cessation) OR (Assisted near/3 cessation)):ti,ab,kw 21 #51 ((Assisting near/3 cessation) OR (Assists near/3 cessation)):ti,ab,kw 21 #52 (Assistance near/3 cessation):ti,ab,kw 17 #53 ((Arrange near/3 "follow up") OR (Arranged near/3 "follow up")):ti,ab,kw 10 #54 ((Arranges near/3 "follow up") OR (Arranging near/3 "follow up")):ti,ab,kw 10 #55 ((Arrange near/3 followup) OR (Arranged near/3 followup)):ti,ab,kw 0 #56 ((Arranges near/3 followup) OR (Arranging near/3 followup)):ti,ab,kw 0 #57 ((Arrangement near/3 followup) OR (Arrangement near/3 followup)):ti,ab,kw 0 #58 ("cessation support" OR "cessation assistance"):ti,ab,kw 18 #59 ((induce near/3 cessation) OR (induces near/3 cessation)):ti,ab,kw 19 #60 ((induced near/3 cessation) OR (inducing near/3 cessation)):ti,ab,kw 19 #61 ((inducement near/3 cessation) OR (induction near/3 cessation)):ti,ab,kw 2 #62 ("cessation counseling" OR "cessation counselling"):ti,ab,kw 118 #63 "smoking status":ti,ab,kw 561

#64 (#40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 OR #55 OR #56 OR #57 OR #58 OR #59 OR #60 OR #61 OR #62 OR #63)

1781

#65 (#39 AND #64) 83 #66 MeSH descriptor Intervention Studies explode all trees 1218 #67 intervention*:ti,ab,kw 47675 #68 (#66 OR #67) 47675 #69 (#65 AND #68) 68 #70 (#39 AND #68) 219 #71 MeSH descriptor Smoking explode all trees with qualifier: DI 0

137


#72 MeSH descriptor Smoking explode all trees with qualifier: PC 910 #73 MeSH descriptor Tobacco Use Cessation explode all trees with qualifier: DI 0 #74 MeSH descriptor Tobacco Use Cessation explode all trees with qualifier: PC 0 #75 MeSH descriptor Tobacco Use Disorder explode all trees with qualifier: DI 22 #76 MeSH descriptor Tobacco Use Disorder explode all trees with qualifier: PC 55 #77 (#71 OR #72 OR #73 OR #74 OR #75 OR #76) 960 #78 (#70 AND #77) 46 #79 MeSH descriptor Smoking explode all trees with qualifier: TH 141 #80 MeSH descriptor Smoking explode all trees with qualifier: DT 142 #81 MeSH descriptor Tobacco Use Cessation explode all trees with qualifier: TH 0 #82 MeSH descriptor Tobacco Use Cessation explode all trees with qualifier: DT 3 #83 MeSH descriptor Tobacco Use Disorder explode all trees with qualifier: TH 126 #84 MeSH descriptor Tobacco Use Disorder explode all trees with qualifier: DT 108 #85 (#79 OR #80 OR #81 OR #82 OR #83 OR #84) 483 #86 (#70 AND #85) 6 #87 ((smok* or tobacco) AND intervention*):ti 712 #88 (#39 AND #87) 45 #89 (#69 OR #78 OR #86 OR #88) 119 #90 (#89), from 1980 to 2008 119

138


139


Appendix C: Included Studies Amundson, G., Solberg, L. I., Reed, M., Martini, E. M., & Carlson, R. (2003). Paying for quality improvement: compliance with tobacco

cessation guidelines. Joint Commisson Journal on Quality and Safety, 29(2), 59-65. Anderson, P., & Jané-Llopis, E. (2004). How can we increase the involvement of primary health care in the treatment of tobacco dependence?

A meta-analysis. Addiction, 99(3), 299-312. Bentz, C. J., Bayley, K. B., Bonin, K. E., Fleming, L., Hollis, J. F., Hunt, J. S., et al. (2007). Provider feedback to improve 5A's tobacco

cessation in primary care: A cluster randomized clinical trial. Nicotine and Tobacco Research, 9(3), 341-349. Borgiel, A. E. M., & Williams, J. I. (1999). Evaluating the effectiveness of 2 educational interventions in family practice. CMAJ: Canadian

Medical Association Journal, 161(8), 965. Cockburn, J., Ruth, D., Silagy, C., Dobbin, M., Reid, Y., Scollo, M., et al. (1992). Randomised trial of three approaches for marketing

smoking cessation programmes to Australian general practitioners. British Medical Journal, 304(6828), 691-694. Coleman, T., Lewis, S., Hubbard, R., & Smith, C. (2007). Impact of contractual financial incentives on the ascertainment and management of

smoking in primary care. Addiction, 102(5), 803-808. DePue, J. D., Goldstein, M. G., Schilling, A., Reiss, P., Papandonatos, G., Sciamanna, C., et al. (2002). Dissemination of the AHCPR clinical

practice guideline in community health centres. (Research Paper). Tobacco Control, 11(4), 329(327). Dickey, L. L., Gemson, D. H., & Carney, P. (1999). Office system interventions supporting primary care-based health behavior change

counseling. American Journal of Preventive Medicine, 17(4), 299-308. Dietrich, A. J., O'Connor, G. T., Keller, A., Carney, P. A., Levy, D., & Whaley, F. S. (1992). Cancer: improving early detection and

prevention. A community practice randomised trial. British Medical Journal, 304(6828), 687-691. Fiore, M., Jaén, C., Baker, T., Bailey, W. C., Benowitz, N. L., Curry, S. J., et al. (2008). Treating tobacco use and dependence:2008 update.

from U.S. Department of Health and Human Services. Public Health Service: http://www.surgeongeneral.gov/tobacco/treating_tobacco_use08.pdf

Fiore, M. C., Jorenby, D. E., Schensky, A. E., Smith, S. S., Bauer, R. R., & Baker, T. B. (1995). Smoking status as the new vital sign: Effect on assessment and intervention in patients who smoke. Mayo Clinic Proceedings, 70(3), 209-213.

Flocke, S. A., Gordon, L. E., & Pomiecko, G. L. (2006). Evaluation of a Community Health Promotion Resource for Primary Care Practices. American Journal of Preventive Medicine, 30(3), 243-251.

Goldstein, M. G., Niaura, R., Willey, C., Kazura, A., Rakowski, W., DePue, J., et al. (2003). An academic detailing intervention to disseminate physician-delivered smoking cessation counseling: smoking cessation outcomes of the Physicians Counseling Smokers Project. Preventive Medicine: An International Journal Devoted to Practice and Theory, 36(2), 185-196.

Grimshaw, J. M., Thomas, R. E., MacLennan, G., Fraser, C., Ramsay, C. R., Vale, L., et al. (2004). Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technology Assessment, 8(6), iii-iv, 1-72.

Hogg, W., Lemelin, J., Graham, I. D., Grimshaw, J., Martin, C., Moore, L., et al. (2008). Improving prevention in primary care: Evaluating the effectiveness of outreach facilitation. Family Practice, 25(1), 40-48.

Joseph, A. M., Arikian, N. J., An, L. C., Nugent, S. M., Sloan, R. J., & Pieper, C. F. (2004). Results of a Randomized Controlled Trial of Intervention to Implement Smoking Guidelines in Veterans Affairs Medical Centers: Increased Use of Medications Without Cessation Benefit. Medical Care, 42(11), 1100-1110.

Katz, D. A., Muehlenbruch, D. R., Brown, R. L., Fiore, M. C., & Baker, T. B. (2004). Effectiveness of Implementing the Agency for Healthcare Research and Quality Smoking Cessation Clinical Practice Guideline: A Randomized, Controlled Trial. Journal of the National Cancer Institute, 96(8), 594-603.

Kottke, T. E., Solberg, L. I., Brekke, M. L., Conn, S. A., Maxwell, P., & Brekke, M. J. (1992). A controlled trial to integrate smoking cessation advice into primary care practice: Doctors Helping Smokers, Round III. The Journal of Family Practice, 34(6), 701-708.

Krist, A. H., Woolf, S. H., Frazier, C. O., Johnson, R. E., Rothemich, S. F., Wilson, D. B., et al. (2008). An Electronic Linkage System for Health Behavior Counseling. Effect on Delivery of the 5A's. American Journal of Preventive Medicine, 35(5 SUPPL.), S350-S358.

Lancaster, T., & Fowler, G. (2000). Training health professionals in smoking cessation. Cochrane Database Systematic Reviews(3), Art. No.: CD000214. DOI: 000210.001002/14651858.

Lemelin, J., Hogg, W., & Baskerville, N. (2001). Evidence to action: a tailored multifaceted approach to changing family physician practice patterns and improving preventive care. CMAJ: Canadian Medical Association Journal, 164(6), 757.

Lennox, A. S., Bain, N., Taylor, R. J., McKie, L., Donnan, P. T., & Groves, J. (1998). Stages of Change training for opportunistic smoking intervention by the primary health care team. Part I : randomised controlled trial of the effect of training on patient smoking outcomes and health professional behaviour as recalled by patients. Health Education Journal, 57(2), 140-149.

McPhee, S. J., Bird, J. A., Fordham, D., Rodnick, J. E., & Osborn, E. H. (1991). Promoting cancer prevention activities by primary care physicians. Results of a randomized, controlled trial. Journal of the American Medical Association, 266(4), 538-544.

McRobbie, H., Hajek, P., Feder, G., & Eldridge, S. (2008). A cluster-randomised controlled trial of a brief training session to facilitate general practitioner referral to smoking cessation treatment. Tobacco Control, 17(3), 173-176.

Millett, C., Gray, J., Saxena, S., Netuveli, G., & Majeed, A. (2007). Impact of a pay-for-performance incentive on support for smoking cessation and on smoking prevalence among people with diabetes. Canadian Medical Association Journal, 176(12), 1705-1710.

Moher, M., Yudkin, P., Wright, L., Turner, R., Fuller, A., Schofield, T., et al. (2001). Cluster randomised controlled trial to compare three methods of promoting secondary prevention of coronary heart disease in primary. BMJ, 322(7298), 1338-1342.

Morgan, G. D., Noll, E. L., Orleans, C. T., Rimer, B. K., Amfoh, K., & Bonney, G. (1996). Reaching midlife and older smokers: Tailored interventions for routine medical care. Preventive Medicine, 25(3), 346-354.

140


Ockene, J. K., Adams, A., Pbert, L., Luippold, R., Hebert, J. R., Quirk, M., et al. (1994). The Physician-Delivered Smoking Intervention Project: factors that determine how much the physician intervenes with smokers. Journal of General Internal Medicine, 9(7), 379-384.

Ockene, J. K., Lindsay, E. A., Hymowitz, N., Giffen, C., Purcell, T., Pomrehn, P., et al. (1997). Tobacco control activities of primary-care physicians in the Community Intervention Trial for Smoking Cessation. COMMIT Research Group. Tobacco Control, 6 Suppl 2, S49-S56.

Ozer, E. M., Adams, S. H., Lustig, J. L., Millstein, S. G., Camfield, K., El Diwany, S., et al. (2001). Can it be done? Implementing adolescent clinical preventive services. Health services research, 36(6 Pt 2), 150-165.

Pbert, L., Fletcher, K. E., Flint, A. J., Young, M. H., Druker, S., & DiFranza, J. (2006). Smoking prevention and cessation intervention delivery by pediatric providers, as assessed with patient exit interviews. Pediatrics, 118(3), e810-824.

Piper, M. E., Fiore, M. C., Smith, S. S., Jorenby, D. E., Wilson, J. R., Zehner, M. E., et al. (2003). Use of the vital sign stamp as a systematic screening tool to promote smoking cessation. Mayo Clin Proc, 78(6), 716-722.

Richmond, R., Mendelsohn, C., & Kehoe, L. (1998). Family Physicians' Utilization of a Brief Smoking Cessation Program Following Reinforcement Contact after Training: A Randomized Trial. Preventive Medicine, 27(1), 77-83.

Roski, J., Jeddeloh, R., An, L., Lando, H., Hannan, P., Hall, C., et al. (2003). The impact of financial incentives and a patient registry on preventive care quality: increasing provider adherence to evidence-based smoking cessation practice guidelines. Preventive Medicine, 36(3), 291-299.

Rothemich, S. F., Woolf, S. H., Johnson, R. E., Burgett, A. E., Flores, S. K., Marsland, D. W., et al. (2008). Effect on cessation counseling of documenting smoking status as a routine vital sign: An ACORN study. Annals of Family Medicine, 6(1), 60-68.

Secker-Walker, R. H., Solomon, L. J., Flynn, B. S., LePage, S. S., Crammond, J. E., Worden, J. K., et al. (1992). Training obstetric and family practice residents to give smoking cessation advice during prenatal care. American Journal of Obstetrics and Gynecology, 166(5), 1356-1363.

Solberg, L. I., Kottke, T. E., Brekke, M. L., Magnan, S., Davidson, G., Calomeni, C. A., et al. (2000). Failure of a continuous quality improvement intervention to increase the delivery of preventive services. A randomized trial. Effective Clinical Practice, 3(3), 105-115.

Strecher, V. J., O'Malley, M. S., Villagra, V. G., Campbell, E. E., Gonzalez, J. J., Irons, T. G., et al. (1991). Can residents be trained to counsel patients about quitting smoking? Results from a randomized trial. Journal of General Internal Medicine, 6(1), 9-17.

Tremblay, M., Gervais, A., Lacroix, C., O'Loughlin, J., Makni, H., & Paradis, G. (2001). Physicians taking action against smoking: An intervention program to optimize smoking cessation counselling by Montreal general practitioners. Canadian Medical Association Journal, 165(5), 601-607.

Unrod, M., Smith, M., Spring, B., DePue, J., Redd, W., & Winkel, G. (2007). Randomized controlled trial of a computer-based, tailored intervention to increase smoking cessation counseling by primary care physicians. Journal of General Internal Medicine, 22(4), 478-484.

Wadland, W. C., Holtrop, J. S., Weismantel, D., Pathak, P. K., Fadel, H., & Powell, J. (2007). Practice-based referrals to a tobacco cessation quit line: assessing the impact of comparative feedback vs general reminders. Annals of Family Medicine, 5, 135-142.

Young, J. M., D'Este, C., & Ward, J. E. (2002). Improving family physicians' use of evidence-based smoking cessation strategies: A cluster randomization trial. Preventive Medicine, 35(6), 572-583.

141


Appendix D: Study Characteristics: All Primary Studies Table 24: Study characteristics: all primary studies

TRAINING Level II evidence Citation Study Population Intervention Comparator Outcomes Hogg et al. (2008) Canada

Cluster randomised controlled trial 12 months Quality: Good General prevention

Fee-for-service primary care practices in Ontario n ≈ 160 physicians n = 54 primary care practices comprising no more than six physicians n = 3049 patients

Training: Outreach Facilitation Two nurses visited each practice on average once a month for a year for an average of 46 minutes/visit. Visits include: audit & feedback, consensus building, including integrating preventive care into routine/episodic care visits, tools for improving the quality of preventive care delivery (such as prevention flow sheets, chart flags, sticker reminders, electronic reminders, patient care records).

No intervention Asking about smoking status ‘A’ and providing smoking cessation counselling ‘B’/ ‘C’ as determined from patient surveys and prescription of nicotine replacement therapy ‘C’ as detected through a chart review. Note: smoking outcomes reported within the context of a ‘general prevention’ trial comprising 14 preventive manoeuvres in total.

McRobbie et al. (2008) U.K.

Cluster randomised controlled trial 3 months Quality: Good

General practices from three east London primary care trusts n = 91 practices n ≈ 250 GPs

Training A 40-minute interactive training session by ‘expert’ educators addressing the rationale and skills for referral of smokers for treatment.

No-intervention/referral, guidance by post only

The number of referrals recorded by the local specialist smoking cessation services over three months after the intervention ‘C’

142


Table 24: Study characteristics: all primary studies (continued) TRAINING Level II evidence Citation Study Population Intervention Comparator Outcomes Unrod et al. (2007) U.S.A.

Randomised controlled trial 6 months Quality: Fair

Internal or family medicine specialty, in the four largest metropolitan boroughs, Bronx, Brooklyn, Manhattan, and Queens who plans to continue practicing in current location for at least one year with at least 75 patient visits per week. n = 70 physicians (35 intervention, 35 control) Baseline: 518 adult smokers (n= 270 Intervention, n=248 control) Analysed (6m follow-up: Intervention n=237, control 228)

Training: Academic Detailing A 40-minute visit by a health educator to the physician’s office to provide individual physician training in brief smoking cessation counselling based on the 5As* Clinical Practice Guideline following an “academic detailing approach” * Ask, Assess, Advise, Assist (1-4), and Arrange

Usual smoking cessation practices

Using PEI (Patient Exit Interviews): Physician performance of the 5As Quit rates and smoking behaviours

Joseph et al. (2004) U.S.A.

Cluster randomised controlled trial 12 months Quality: Good

Veterans Affairs medical centres n = 20 n ≈ 200 HCPs n = 5678 patients

Training: Academic Detailing Two day off-site training of a facilitator and one key advocate per site. Two-to- three day on-site visit per site to identify barriers and prioritise goals specific to each site and to design strategies to increase delivery of smoking intervention to all smokers. Recommended strategies included: (1) the use of non-physician personnel for documentation, (2) use of electronic clinical reminder systems, (3) adaptation of note templates to include smoking status, (4) increasing the availability of smoking cessation medications, (5) improving the identification of smoking status in the medical records.

No-intervention/printed smoking cessation guidelines only.

(1) Recorded smoking status ‘A’. (2) If the record indicated that cessation was discussed or the provision of advice or medication ‘B’/‘C’ (3) Provision of counselling or referral to a smoking cessation clinic ‘C’.

143


Table 24: Study characteristics: all primary studies (continued) TRAINING Level II evidence Citation Study Population Intervention Comparator Outcomes Solberg et al. (2000) U.S.A.

Cluster randomised controlled trial 24 months Quality: Good General prevention

All HCPs from primary care clinics in greater Minneapolis n = 44 clinics n ≈ 500 primary care Doctors

Training: Continuous Quality Improvement (CQI) Leadership support and training in CQI methods for implementing improvement processes to build prevention systems.

No intervention/usual care

Smoking status– ‘A’ The proportion of patients who were up-to-date for each service within three months of an index primary care visit on the basis of chart audit.

Borgiel & Williams (1999) Canada

Cluster randomised controlled trial 12 months Quality: Fair

Family Physicians n = 2 clinics n = 57 HCPs

Training: Outreach Facilitation Continuing Medical Education (CME), structured information for identifying relative strengths and weakness in the practice, educational workshop, follow-up phone call to physicians and review.

No intervention/ Practice Assessment Report only.

Documentation of patient data including smoking status based on three sources: patient questionnaires, chart audits and physician questionnaires.

Lennox et al. (1998) Scotland

Cluster randomised controlled trial 14 months Quality: Fair

GPs, nurses, and other staff. n = 16 practices n ≈ 140 HCPs

Training A one-day ‘Stages of Change’ training (including motivational interviewing) workshops for primary health care teams.

No intervention/ ‘untrained’ controls

GPs Ask/Brief advice ‘A’/‘B’ Nurses Ask/Brief advice ‘A’/‘B’ Other HCPs Ask/Brief advice ‘A’/‘B’, as assessed by postal questionnaire of smoking patients at eight and 14 months follow-up. Note – also point prevalence and continuous abstinence.

Ockene et al. (1997) U.S.A.


Physicians in the intervention communities participating in the continuous medical education (CME). n = 22 communities n = 869 physicians

Training: Continuing Medical Education (CME) Three levels of training activities for physicians (basic, comprehensive, and “train the trainers”), and one type of training for staff in physicians’ offices.

Not clearly stated, probably usual care or no intervention

Reported smoking intervention practices of physicians, patient reports of physician intervention activities

144


Table 24: Study characteristics: all primary studies (continued) TRAINING Level II evidence Citation Study Population Intervention Comparator Outcomes Morgan et al. (1994) U.S.A.

Randomised controlled trial 6-months (a two year trial) Quality: Fair

Primary care practices (February 1991 through February 1992) in suburban Philadelphia and eastern Pennsylvania n = 39 primary care practices including all physicians and non-physician office staff. 659 patients aged 50-74 years

Training: Office-based Physician training On-site training to implement a modified National Cancer Institute (NCI) smoking cessation intervention (Ask about smoking at every opportunity, Advice all smokers to stop, Assist the patient to stop smoking, and Arrange for follow-up support).

Usual care Provider adherence, included eight areas, two of them correspond to ‘brief advice’ and ‘cessation support’

Cockburn et al. (1992) Australia

Randomised controlled trial 4 months Quality: Good

General practitioners n = 264

Training: Educational facilitator Personal delivery (by appointment) and presentation of a smoking cessation kit by an educational facilitator with a follow up visit six weeks later.

(1) Delivery to the receptionist by a friendly volunteer courier (unannounced visits) with a follow up phone call six weeks later. (2)Postal delivery with a follow up letter six weeks later.

Measured use of components of the kit at follow-up. Offering simple advice to quit ‘B’, giving the patient a self help booklet, a "help" card, cessation advice, "contract" card/quit date ‘C’. GPs self-report of delivery of smoking cessation interventions via questionnaire.

Training: Education Education: a one day, expert-led, group education meeting for physicians discussing smoking cessation advice for 120 minutes (and other services for 30 to 60 minutes each).

Dietrich et al. (1992) U.S.A.

Randomised controlled trial, 2x2 factorial trial of two interventions alone and in combination 12 months follow-up. Quality: Fair General prevention

Office based family physicians and general internists. n = 98 practices Training: Office systems

Assistance from a project facilitator in the design and implementation of office routines that support provision of cancer early detection and preventive services, including team work (division of responsibilities for providing services), preventive care flow sheets, external identifiers for smokers' charts, health education posters and brochures, and patient held health diaries.

Education vs. office system intervention vs. no intervention.

Incidence of recommendations to quit smoking as assessed by patient exit questionnaires ‘A’ and ‘B’. Note: smoking outcomes reported within the context of a ‘general prevention’ trial.

145


Table 24: Study characteristics: all primary studies (continued) TRAINING Level III-2 evidence Citation Study Population Intervention Comparator Outcomes Goldstein et al. (2003) U.S.A.

Non-randomised Clinical trial 24 months Quality: Fair

Community-based sample of primary care physicians. n = 259 physicians n = 4295 smokers

Training: Academic detailing. Multi-component office-based intervention incorporated assessment of physicians’ baseline knowledge and attitudes, identifying clear goals, use of supportive written materials, office-based outreach educational component.

No intervention Note: simultaneous delivery of a home-based smoking cessation intervention to both groups by a team of collaborating investigators.

The proportion of patients who reported that their physician talked to them about their smoking habit ‘A’, or advised them to quit smoking ‘B’, and/or arranged a follow-up visit to discuss smoking ‘C’, at 12, 18, or 24-month follow-up interview.

Kottke et al. (1992) U.S.A.

Non-randomised Clinical trial 18 months Quality: Good

Non-volunteer primary care clinics within an independent practice association-type managed health care plan. n = 19 clinics n = 142 primary care physicians and other support staff.

Training: Academic detailing. Organisational support/ outreach educational workshops. A clinic-wide intervention system to support physicians. Responsibility for smoking cessation activities was widely dispersed among all personnel who come in contact with patients–program is tailored to the unique attributes of each clinic. Intervention based on the “social learning model of consultation” (Brown & Schulte 1887). Intervention included on average 5.7 site visits, 24 telephone calls, and 5.6 mailings per practice site over the 18-month period and provided information about a feasible and effective systems, help with problem solving, assessment of program progress, and feedback and reinforcement.

No intervention at control clinics.

Rates of smoking cessation ‘asking’ ‘A’ and ‘advice’ ‘B’ as measured pre-intervention and post-intervention patient surveys.

146


Table 24: Study characteristics: all primary studies (continued) TRAINING Level III-3 evidence Citation Study Population Intervention Comparator Outcomes DePue et al. (2002) U.S.A.

Pre-test/post-test. 12 months follow-up. Quality: Good

HCPs across 14 sites in Rhode Island. n = 151 (27 physicians; 53 nurses; one social worker; 70 other staff (including medical assistants, dieticians, counsellors, receptionists, and dentists). n = 1591 patients

Training: Academic detailing. Three, one hour on-site provider training sessions, on review of effective tobacco interventions, use of office systems, and tobacco counselling skill building.

No intervention vs. follow-up; HCPs their own controls.

Documentation in the last visit (percentage recording): ‘A’ notation of smoking status. ‘B’ was considered any notation of advice or discussion about quitting smoking. ‘C’ (reported here as a composite of Assist & Arrange).

Secker-Walker et al. (1992) U.S.A.

Pre-test/post-test. 24 months Quality: Fair

19 Obstetric residents (14 female), 16 family physician residents (11 female) at the University of Vermont, in Maternal-Infant Care clinic, provides prenatal care to women with little or no insurance. Pregnant women (smokers). n = 35 Residents 250 pregnant women (40-45% smokers at the onset of pregnancy

Training: workshops. Single-component educational training for residents in obstetrics and family practices to provide smoking cessation advice during prenatal care. This is a one-hour workshop + refresher sessions and progress.

Before and after the training of residents providing both the intervention group and the control group (basic care). Basic care: basic care as outlined on a separate protocol prompt sheet and consists of acknowledging women’s smoking status, delivering a rationale for quitting, giving a strong recommend-ation to quit, and providing a quit-smoking booklet at the first ante-partum visit. Women were not encouraged to set a quit date, nor were they referred to a smoking cessation counsellor.

Video rating (comparisons of pre-training and post-training videotapes), adherence to protocol. Comparing and rating (through seven criteria) videotapes of the residents providing their usual advice to a simulated patient before training, with videotapes of each resident delivering the intervention protocol to the same simulated patient at the end of their refresher session. ‘A’ Ask, ‘B’ Brief advice, and Cessation ‘C’ (all included in protocol prompt sheet).

147


Table 24: Study characteristics: all primary studies (continued) MULTI-COMPONENT Multi-component: Level II evidence Citation Study Population Intervention Comparator Outcomes Pbert et al. (2006) U.S.A.


Paediatric clinics through Massachusetts (Paediatricians, nurse practitioners, physician assistants and paediatric residents). n = 8 clinics n = 61 HCPs 2710 adolescents 13-17 years

Multi-component: Training and reminder. Special Intervention (SI): One-hour group training and reminders to deliver a brief smoking prevention and cessation intervention to adolescents in the context of routine paediatric primary care practice.

Usual Care (UC)

Percentage of intervention steps completed by paediatric provider (reported in patient exit interviews). All patients asked if ever smoked or tried smoking ‘A’.

Katz et al. (2004) U.S.A.


Intake clinicians from community based clinics in southern Wisconsin. Intervention n = 8 Community-based clinics n = 75 intake clinicians 2163 consecutive adult patients

Multi-component: Guideline-based intervention. A tutorial for intake clinicians, group and individual performance feed-back for intake clinicians, use of a modified vital signs stamp, an offer of free nicotine replacement therapy, and proactive telephone counselling.

No intervention Performance of recommended smoking cessation activities by clinic staff. Abstinence at two and six months.

Young et al. (2002) Australia

Cluster randomised controlled trial. 6 months Quality: Good

All family physicians (FP) in a defined geographical area in Sydney. n = 39 practices n = 60 family physicians 1788 adult patients

Multi-component: audit and feedback plus resources. Combination of audit and feedback, resources for FPs, resources for practices, and resources for patients.

Intervention of identical format and intensity but about cervical screening.

Recall of patients about FP questioning about smoking status ‘A’, advice to quit ‘B’, provision of practical advice ‘C’.

Lamelin et al. (2001) Canada

Randomised controlled trial. 18 months Quality: Fair General prevention

Health services organisations in Ontario (community primary care practices that have a payment system based on capitation). n = 46 practices n = 561 physicians Mean number of patients seen per practice per day was 34.4 for intervention and 33.0 for comparison group.

Multi-component: seven components. Seven intervention strategies delivered by nurse facilitators: Audit and ongoing feedback, consensus building, opinion leaders and networking, academic detailing and education materials, reminder systems, patient-mediated activities, and patient education materials.

No intervention Proportion of eligible patients who received preventive manoeuvres. One of these manoeuvres involved Smoking cessation counselling ‘C’.

148


Table 24: Study characteristics: all primary studies (continued) MULTI-COMPONENT Multi-component: Level II evidence Citation Study Population Intervention Comparator Outcomes

Multi-component: Training Tutorial: Two sessions in smoking cessation counselling including slide show/lecture, video, handouts, flowchart, group discussion; delivered by the clinic director or another educator.

Strecher et al. (1991) U.S.A.

Randomised factorial four group design. 6 months Quality: Fair

All resident physicians from 11 primary care programmes in three university medical centres. n = 261 physicians

Multi-component: Reminders Chart-based reminders to assist physicians to provide counselling (flow-chart placed on patients’ records by clinic staff at patent check-in).

Tutorial and prompt vs. Tutorial only vs. Prompt only vs. No intervention.

The percentage of smokes provided with brief advice ‘B’ (by patient exit interview). The occurrence of five specific cessation support techniques used by physicians ‘C’: setting a quit date, prescribing a quit date, prescribing NRT, giving self-help material, providing follow-up (by blinded, researcher administered telephone interview). Also, patient reported smoking status at six months (quit rates) ‘Q’.

Multi-component: Level III-3 evidence Flocke et al. (2006) U.S.A.

Pre-test/post-test. 12 months follow-up. Quality: Good General prevention

Solo physician practices and single-specialty family practices ranging in size from two to 12 clinicians, in either urban or suburban settings. n = 7 primary care practices n ≈ 50 primary care physicians n = 784 patients

Multi-component: Office systems. A practice-tailored intervention using a web-based educational/referral resource, a Prescription for Health Pad and practice feedback and planning (teamwork).

No intervention/ pre-test baseline care vs. follow-up, HCPs their own controls.

Patient report of discussion of smoking ‘A’ and ‘B’ as assessed by patient exit interview. Note: smoking outcomes reported within the context of a ‘general prevention’ trial.

149


Table 24: Study characteristics: all primary studies (continued) MULTI-COMPONENT Multi-component: Level III-3 evidence Citation Study Population Intervention Comparator Outcomes Ozer et al. (2001) U.S.A.

Pre-test/post-test. 18 months duration Quality: Fair General prevention

Outpatient paediatrics clinics within a large managed care system throughout Northern California. n = 3 clinics n = 89 clinicians (86% physicians; 14% nurse practitioners). 1313 adolescents

Multi-component: Provider Training, Tools, Health Educator. Three components: (1) Clinician training to increase the delivery of preventive services; (2) Clinic tools: adolescent screening and provider charting forms; and (3) Health educator to the clinics.

Post-implementing the intervention.

Adolescent reports of clinician screening and counselling practices; an example of a screening question is: “Did your doctor ask if you smoke or chew tobacco?” ‘A’. An example of a counselling question for adolescents who were not engaging is: “Did your doctor encourage you to remain a non-smoker or non-tobacco user?” ‘B’. Examining changes in screening and counselling practices from the pre- to post-implementation period.

150


Table 24: Study characteristics: all primary studies (continued) MULTI-COMPONENT Multi-component: Level IV evidence Citation Study Population Intervention Comparator Outcomes Tremblay et al. (2001) Canada

Two cross-sectional surveys from 1998 and 2000. Quality: Fair

GPs practising in Montreal in 1997. 5679 patients 337 (GPs in 1998), 316 (GPs in 2000)

Multi-component: five-year intervention program to improve the smoking cessation counselling practices of general practitioners in Montreal. The interventions were: Thirty-two 60-90 minute workshops about smoking cessation counselling. Two 30-minute conferences on smoking cessation counselling. Publication of 11 articles about smoking cessation counselling in journals for Quebec physicians. Publication of clinical guidelines for smoking cessation. Production and distribution of 2-sided plastic-coated card presenting a summary, in algorithm format, of smoking cessation interventions. Publication and distribution of a smoking cessation guide, presented in a cigarette pack format. Publication and distribution of Breath Free for Life, a list of smoking cessation in Montreal. Support of involvement of the College des medicines du Quebec in awareness to campaigns by Quebec Council on Tobacco and Health to encourage patients to stop smoking and to ask for help from their physicians. Collaboration in the publication of two newspaper articles, and two articles in a consumer magazine to reinforce the importance of physicians in smoking cessation and to encourage smokers to ask for help from their physicians.

1998 vs. 2000

Trends in the smoking cessation counselling practices.

151


Table 24: Study characteristics: all primary studies (continued) REMINDERS Reminders: Level II evidence Citation Study Population Intervention Comparator Outcomes Rothemich et al. (2008) U.S.A.


Primary care clinicians (family physicians, general internists, nurse-practitioners, or physician’s assistants) from 18 practices caring for adult patients. n = 18 clinics (clusters) n = 73 clinicians (14 family medicine practices, three internal medicine practices, and one practice featuring both specialties. Two practices were urban, two were rural, and 14 were suburban). n = 73 clinicians 6,729 adult patients

Reminders: Vital sign Nurses and medical assistants were instructed to assess the tobacco use status of every adult patient and record it with the traditional vital signs.

No use of any systematic tobacco screening or identification system.

Proportion of smokers reporting clinician counselling of any kind and the frequency of two counselling subcomponents: simple quit advice ‘B’ and more intensive discussion ‘C’. Proxy for intervention: Ask ‘A’.

McPhee et al. (1991) U.S.A.

Cluster randomised controlled trial. 12 months Quality: Good General prevention

Primary care physicians n = 40 practices n = 40 physicians (one physician per practice).

Reminders: Computer-generated cancer prevention reminders of overdue screening tests, and smoking and dietary assessment and counselling, supplemented by cancer education materials.

No intervention Smoking status recorded. Smoking cessation counseling, extracted from the medical records. Note: smoking outcomes reported within the context of a ‘general prevention’ trial.

Piper et al. (2003) U.S.A.

Comparative controlled study of pre-post design. 47 months Quality: Fair

Primary health care clinics in Madison. n = 5 Primary health care clinics n = 130 Physicians 1611 current adult smokers

Reminders: Vital sign. Vital sign stamp to remind clinicians to ask about smoking status.

Standard vital signs that did not include smoking status, and staff did not receive training.

Baseline: if the physician has asked the patient about smoking status ‘A’; if currently smoked, if planned to quit in the next 12 months ‘C’. Follow-up questions: Mean number of cigarettes per day, whether made a serious quit attempt within the past year, and whether smoked in the past seven days ‘A’.

152


Table 24: Study characteristics: all primary studies (continued) REMINDERS Reminders: Level III-3 evidence Citation Study Population Intervention Comparator Outcomes Krist et al. (2008) U.S.A.

Pre-test/post-test. Two year-pre and one year-post launch of the intervention. Exposure period five weeks. Quality: Fair General prevention

Primary care practices in the Tidewater region of Virginia. n = 9 Primary health care clinics n = 48 clinicians 5679 patients

Reminders: Electronic system prompt (eLinkS). An early input from clinicians at study sites, on patient examination and obtaining vital signs to enter into medical record the EMR (electronic medical record) displays a screen prompt to remind the staff to enter tobacco-use status (as well as height and weight).

Pre launch of the intervention (Current practice).

Delivery of 5As measured by clinicians’ entries in the EMR dialogue box.

Fiore et al. (1995) U.S.A.

Pre-post design. 16 month duration Quality: Fair

General Internal Medicine Clinic (GIMC) staff at the University of Wisconsin Hospitals and Clinics. n = 1 clinic n = 52 clinicians (47 physicians; 5 nurse practitioners) 28,000 patients

Reminders: Vital sign stamps. Making smoking status part of the vital sign assessment.

After expanding the vital signs to include smoking status.

Rate of clinicians/other than clinician or both asking about smoking ‘A’. Percentage of smokers who reported that their clinician advised them that day to quit smoking ‘B’.

153


Table 24: Study characteristics: all primary studies (continued) FINANCIAL INCENTIVES Financial incentives: Level II evidence Citation Study Population Intervention Comparator Outcomes

Financial incentives Incentive group: financial incentives for reaching preset clinical performance targets.

Roski et al. (2003) U.S.A.

Cluster randomised controlled trial– 3 groups. 12 months Quality: Fair

HCPs providing family practice, internal medicine and obstetrics/ gynaecology services in a large Midwestern multi-specialty group practice. n = 40 clinics n = 4734 patients

Financial incentives and office support/audit & feedback Registry group: financial incentives and access to a centralised smoker registry and audit and weekly feedback on referral activity.

No-intervention/ distribution of printed versions of the smoking cessation guidelines only.

Clinical practice patterns: tobacco user identification ‘A’, providing advice to quit to current smokers ‘B’, referral activity ‘C’ measured by patient exit surveys. Note: patient outcomes also reported– counselling, medication, quit rates.

Financial incentives: Level III-2 evidence Millett et al. (2007) U.K.

Population-based longitudinal study (cohort) 15 months (over two study periods before 2003 and 2005). Quality: Fair

Study area has 36 general practices in Wandsworth Primary Care Trust in southwest London, England (a total registered population of 243 519) majority (74%) aged less than 45years. n = 32 Practices (agreed to participate) n = ? physicians

Financial incentives: pay for performance. Pay-for-performance incentive as a method to improve quality of care received by patients introduced in 2004.

Pre-(2003) and post-intervention (2005).

Changes in achievement between the two study periods. Smoking status recorded ‘A’, and smoking cessation advice ‘B’.

Financial incentives: Level III-3 evidence Amundson et al. (2003) U.S.A.

Interrupted time series without a parallel control group. 36 months (1996-1999) Quality: Good General prevention

Medical groups* within the Minnesota HealthPartners network-model health plan. n = 20 medical groups of 6-500 individual HCPs per group n = 5929 patients *Medical groups comprise many clinics/practices

Financial incentives: Financial reward to medical groups that achieve specific high levels of performance on selected measures of care quality (including smoking A&B) as defined by the Outcomes Recognition Program (ORP). Includes audit and feedback that is also available publicly on a ‘Consumer Choice’ Website.

No intervention prior to 1997 compared to 1997; 1998; 1999.

Documentation of Ask ‘A’ and Advise ‘B’ in the patient medical records at the last visit. Note: smoking outcomes reported within the context of a ‘general prevention’ trial. Other ORP targets were in place at the time of the trial, however these were not reported in this paper.

154


Table 24: Study characteristics: all primary studies (continued) FINANCIAL INCENTIVES Financial incentives: Level III-3 evidence Citation Study Population Intervention Comparator Outcomes Coleman et al. (2007) U.K.

Interrupted time series: data analysis of longitudinal general practice database, 15 years (14 years pre – 18 months post (1990 to 2005). Quality: Good

All Family physicians identified by the U.K. primary care medical records [The THIN network]). All identified GPs n = 1,607,782 patients.

Financial incentives: Introduction of substantial financial rewards for compliance with targets, specifically, smoking status (A) and recording the delivery of GPs’ brief smoking cessation advice (B) (but not Cessation support/referral).

No intervention/ pre-test (compared to the first quarter of 2003).

The incidence of recording of smoking status in medical records ‘A’ and, in smokers, the receipt of GPs' smoking cessation advice ‘B’, and prescriptions for nicotine addiction treatments and prescriptions for nicotine replacement therapy ‘C’.

155


Table 24: Study characteristics: all primary studies (continued) AUDIT & FEEDBACK Audit & Feedback: Level II evidence Citation Study Population Intervention Comparator Outcomes Bentz et al. (2007) U.S.A.


HCPs from 19 primary care clinics in Oregon caring for a diverse population of 175,000 patients. n = 19 clinics n = 279 HCPs

Audit & Feedback: Initial 30-min training sessions for staff in all clinics to review evidence-based cessation strategies and electronic medical record (EMR) modified to include an electronic Quitline fax-referral capability and audit-and-feedback reports to clinic staff.

Initial 30-minute training sessions for staff in all clinics to review evidence-based cessation strategies and computer system modified to include an electronic Quitline fax-referral capability but no audit-and-feedback reports to clinic staff.

EHR-documented tobacco-related performance of ask ‘A’, advise, assess ‘B’, and assist, and referral of tobacco users to Quit Line ‘C’.

Wadland et al. (2007) U.S.A.

Cluster Randomised controlled trial. 18 months Quality: Fair

Primary care clinicians within practices in Michigan. n = 87 primary care practices n = 308 clinicians 1992 patients

Audit & Feedback: Quarterly comparative feedback report with 2 components: – Performance of study group in numbers of referrals and percentage relative to the benchmark. – Actual number of referrals made by the individual clinician per quarter.

Simple postcards informing of the Quit Line services.

Changes in clinician referrals from baseline to post-intervention ‘C’; estimated quit responses of referred smokers, ‘Q’; Estimated incremental costs of the feedback intervention.

Moher et al et al. (2001) U.K.

Cluster RCT 18 months. Quality: Fair General prevention

All general practices in Warwickshire. n = 79 primary care practices eligible, but only 21 participated. n = 62 to ≥ 89 GPs, and a total of ≥ 28 practice nurses (calculated) 1906 patients

Audit & Feedback: Summary audit results at a practice meeting. Practices asked to provide usual care and were given no further support during the trial. Recall to GP: Practices given same patient information as audit group but also the names of patients identified as having coronary heart disease. Recall to nurse clinic: Practices given same patient information as GP recall group.

Audit vs. GP recall vs. Nurse recall.

‘A’ (at baseline and at follow-up at 18 months) A record of smoking habit and, for smokers, a record of review in the previous two years.

156


Table 24: Study characteristics: all primary studies (continued) OTHER Other: Level II evidence Citation Study Population Intervention Comparator Outcomes Richmond et al. (1998) Australia

Randomised controlled trial. 6 months Quality: Good

Family physicians who accepted invitations and participated in a workshop to be trained in smoking cessation techniques. n = 198

Other: Reinforcement following training. A two hour workshop of training in a brief smoking cessation interventions and reinforcement contact from the medical practitioner trainer, consisting of three brief telephone calls at two weeks, two months, and four months after training, by a peer doctor.

No-intervention/ two-hour workshop to be trained in a brief smoking cessation intervention but no reinforcement contact from the medical practitioner trainer.

Doctors self-report of their current use of cessation interventions ‘ABC’. The number of smoking cessation information booklets given to smokers ‘C’.

Ockene et al. (1994) U.S.A.

Randomised controlled trial. 12 months Quality: Fair

Internal medicine residents and general internal medicine attending physicians receiving training at the University of Massachusetts Medical School’s postgraduate year one and two residency program from July 1989 through June 1991. n = 50 internal medicine residents (40), and attending physicians (10). 540 smokers (17-75years)

Other: Algorithm Algorithm instructing the physician in the ten smoking intervention steps.

No algorithm. Physician performance measure through Patient Exit Interview (PEI). PEI measures: discussed smoking ‘C’, advised to stop smoking ‘B’, discussed reasons ‘C’, asked about past experiences, asked about problems ‘C’, discussed specific strategies ‘C’, asked about dependency ‘C’, patient agreed to stop smoking or if not, patient agreed to cut down, set date to stop/cut down ‘C’, put plan in writing ‘C’, discussed other behavioural changes ‘C’, given written materials ‘C’, set future contact or if not planned to discuss later ‘C’.

157


Appendix E: Excluded Studies Annotated by Reason for Exclusion

The citations provided were obtained and prepared independently as they appeared in EMBASE.com and the Cochrane Library. The citations were printed verbatim and may have original errors and misspellings. Annotated exclusions have been added. Treating tobacco use and dependence: a systems approach. A

guide for health care administrators, insurers, managed care organizations, and purchasers. (2000). Retrieved from http://www.ahrq.gov/clinic/tobacco/systems.htm Title/abstract excluded: Inappropriate study design.

Agency for Healthcare Research and Quality (2008). New evidence provides clinicians with better tools to help smokers quit. Rockville, MD. : Press Release Date: May 7, 2008 Retrieved from http://www.ahrq.gov/news/press/pr2008/tobupdatepr.htm Title/abstract excluded: Inappropriate study design.

Agency for Healthcare Research and Quality (2008). Treating tobacco use and dependence: 2008 update. Retrieved from http://www.ahrq.gov/path/tobacco.htm Title/abstract excluded: Inappropriate study design.

Aira, M., Kauhanen, J., Larivaara, P., & Rautio, P. (2004). Differences in brief interventions on excessive drinking and smoking by primary care physicians: Qualitative study. Preventive Medicine, 38(4), 473-478. Title/abstract excluded: Inappropriate study design.

Albrecht, S. A., Maloni, J. A., Thomas, K. K., Jones, R., Halleran, J., & Osborne, J. (2004). Smoking cessation counseling for pregnant women who smoke: scientific basis for practice for AWHONN's SUCCESS project. Journal of obstetric, gynecologic, and neonatal nursing : JOGNN / NAACOG, 33(3), 298-305. Title/abstract excluded: Inappropriate study design.

Almeida, O. P., & Pfaff, J. J. (2005). Depression and smoking amongst older general practice patients. Journal of Affective Disorders, 86(2-3), 317-321. Title/abstract excluded: Inappropriate study design.

An, L. C., Zhu, S. H., Nelson, D. B., Arikian, N. J., Nugent, S., Partin, M. R., et al. (2006). Benefits of telephone care over primary care for smoking cessation: A randomized trial. Archives of Internal Medicine, 166(5), 536-542. Title/abstract excluded: Incorrect population.

Anczak, J. D., & Nogler 2nd., R. A. (2003). Tobacco cessation in primary care: maximizing intervention strategies. Clin Med Res, 1(3), 201-216. Title/abstract excluded: Inappropriate study design.

Andrews, J. O., Tingen, M. S., & Harper, R. J. (1999). A model nurse practitioner-managed smoking cessation clinic. Oncology nursing forum, 26(10), 1603-1610. Title/abstract excluded: Inappropriate study design.

Andrews, J. O., Tingen, M. S., Waller, J. L., & Harper, R. J. (2001). Provider feedback improves adherence with AHCPR Smoking Cessation Guideline. Preventive Medicine, 33(5), 415-421. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Aveyard, P., Griffin, C., Lawrence, T., & Cheng, K. K. (2003). A controlled trial of an expert system and self-help manual intervention based on the stages of change versus standard self-help materials in smoking cessation. Addiction, 98(3), 345-354. Title/abstract excluded: Incorrect population.

Avila-Tang, E., Apelberg, B. A., Yamaguchi, N., Katanoda, K., Sobue, T., & Samet, J. M. (2008). Modeling the

Health Benefits of Smoking Cessation in Japan. Tobacco Control, tc.2007.024620. Title/abstract: Included. Full paper excluded: Inappropriate study design .

Bachmann, O. M., & Butler, C. (1998). Correction: Design and analysis of studies evaluating smoking cessation interventions where effects vary between practice or practitioner (Family Practice (1996) 13 (402-407) [1]. Family Practice, 15(3), 289. Title/abstract excluded: Inappropriate study design.

Bass, F. (1996). Mobilizing physicians to conduct clinical intervention in tobacco use through a medical-association program: 5 years' experience in British Columbia. Canadian Medical Association Journal, 154(2), 159-164. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Battista, R. N., Williams, J. I., Boucher, J., Rosenberg, E., Stachenko, S. J., Adam, J., et al. (1991). Testing various methods of introducing health charts into medical records in family medicine units. Cmaj, 144(11), 1469-1474. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Bauld, L., Coleman, T., Adams, C., Pound, E., & Ferguson, J. (2005). Delivering the English smoking treatment services. Addiction, 100(Suppl2), 19-27. Title/abstract excluded: Inappropriate study design.

Baumann, L. C., Chang, M. W., & Hoebeke, R. (2002). Clinical outcomes for low-income adults with hypertension and diabetes. Nursing research, 51(3), 191-198. Title/abstract excluded: Inappropriate study design.

Belcher, D. W. (1990). Implementing preventive services. Success and failure in an outpatient trial. Arch Intern Med, 150(12), 2533-2541. Title/abstract: Included. Full paper excluded: Incorrect population.

Benn, K. (2007). Audit of inpatient smoking-cessation support services at Harrogate and District NHS Foundation Trust (HDFT). Pharmacy in Practice, 17(2), 73-76. Title/abstract excluded: Incorrect population.

Bentz, C. J., Bayley, K. B., Bonin, K. E., Fleming, L., Hollis, J. F., & McAfee, T. (2006). The feasibility of connecting physician offices to a state-level tobacco quit line. American Journal of Preventive Medicine, 30(1), 31-37. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Bernstein, S. L., & Becker, B. M. (2002). Preventive Care in the Emergency Department: Diagnosis and management of smoking and smoking-related illness in the emergency department: A systematic review. Academic Emergency Medicine, 9(7), 720-729. Title/abstract excluded: Inappropriate study design.

Bernstein, S. L., Boudreaux, E. D., Cydulka, R. K., Rhodes, K. V., Lettman, N. A., Almeida, S. L., et al. (2006). Tobacco Control Interventions in the Emergency Department: A Joint Statement of Emergency Medicine Organizations. Journal of Emergency Nursing, 32(5), 370-381. Title/abstract excluded: Inappropriate study design.

Bertakis, K. D., & Azari, R. (2007). Determinants of physician discussion regarding tobacco and alcohol abuse. Journal of health communication, 12, 513-525. Title/abstract excluded: Inappropriate study design.

158


Block, D. E., Hutton, K. H., & Johnson, K. M. (2000). Differences in tobacco assessment and intervention practices: A regional snapshot. Preventive Medicine, 30(4), 282-287. Title/abstract excluded: Inappropriate study design.

Bonevski, B., Sanson-Fisher, R. W., Campbell, E., Carruthers, A., Reid, A. L. A., & Ireland, M. (1999). Randomized Controlled Trial of a Computer Strategy to Increase General Practitioner Preventive Care. Preventive Medicine, 29(6), 478-486. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Borum, M. L. (2000). A comparison of smoking cessation efforts in African Americans by resident physicians in a traditional and primary care internal medicine residency. Journal of the National Medical Association, 92(3), 131-135. Title/abstract excluded: Incorrect intervention.

Braun, B. L., Fowles, J. B., Solberg, L. I., Kind, E. A., Lando, H., & Pine, D. (2004). Smoking-Related Attitudes and Clinical Practices of Medical Personnel in Minnesota. American Journal of Preventive Medicine, 27(4), 316-322. Title/abstract excluded: Inappropriate study design.

Brightwell, R. E., Caruana, M. F., Mitchell, S., Cameron, A. E. P., & Osman, I. S. (2003). Implementing best medical therapy in patients with peripheral arterial occlusive disease: The cost to a specialist vascular clinic. Journal of Clinical Excellence, 4(4), 411-419. Title/abstract excluded: Inappropriate outcomes.

Bronson, D. L., Flynn, B. S., Solomon, L. J., Vacek, P., & Secker-Walker, R. H. (1989). Smoking cessation counseling during periodic health examinations. Archives of Internal Medicine, 149(7), 1653-1656. Title/abstract excluded: Inappropriate study design.

Brown, D., & Schulte, A. C. (1987). A social learning model of consultation. Professional psychology, research and practice, 18(3), 283-287. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Brull, R., Ghali, W. A., & Quan, H. (1999). Missed opportunities for prevention in general internal medicine. Canadian Medical Association Journal, 160(8), 1137-1140. Title/abstract excluded: Inappropriate study design.

Buffels, J., Degryse, J., Decramer, M., & Heyrman, J. (2006). Spirometry and smoking cessation advice in general practice: A randomised clinical trial. Respiratory Medicine, 100(11), 2012-2017. Title/abstract excluded: Incorrect population.

Bull, F. C., Jamrozik, K., & Blanksby, B. A. (1999). Tailored advice on exercise - Does it make a difference. American Journal of Preventive Medicine, 16(3), 230-239. Title/abstract excluded: Incorrect population.

Burton, L. C., Paglia, M. J., German, P. S., Shapiro, S., & Damiano, A. M. (1995). The effect among older persons of a general preventive visit on three health behaviors: smoking, excessive alcohol drinking, and sedentary lifestyle. The Medicare Preventive Services Research Team. Preventive Medicine, 24, 492-497. Title/abstract excluded: Incorrect population.

Butler, C., & Bachmann, M. (1996). Design and analysis of studies evaluating smoking cessation interventions where effects vary between practices or practitioners. Family Practice, 13(4), 402-407. Title/abstract excluded: Inappropriate study design. Background.

Butler, C. C., Pill, R., & Stott, N. C. H. (1998). Qualitative study of patients' perceptions of doctors' advice to quit smoking: Implications for opportunistic health promotion. British Medical Journal, 316(7148), 1878-1881. Title/abstract excluded: Inappropriate study design.

Butler, C. C., Rollnick, S., Cohen, D., Bachmann, M., Russell, I., & Stott, N. (1999). Motivational consulting versus brief advice for smokers in general practice: A

randomized trial. British Journal of General Practice, 49(445), 611-616. Title/abstract excluded: Incorrect intervention.

Campbell, E. E., Villagra, V. G., & Rogers, C. S. (1991). Teaching and promoting smoking cessation counseling in primary care residencies: Description of a method. Teaching and Learning in Medicine, 3(1), 20-27. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Campbell, N. C., Ritchie, L. D., Thain, J., Deans, H. G., Rawles, J. M., & Squair, J. L. (1998). Secondary prevention in coronary heart disease: A randomised trial of nurse led clinics in primary care. Heart, 80(5), 447-452. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Canga, N., De Irala, J., Vara, E., Duaso, M. J., Ferrer, A., & Martinez-Gonzalez, M. A. (2000). Intervention study for smoking cessation in diabetic patients: A randomized controlled trial in both clinical and primary care settings. Diabetes Care, 23(10), 1455-1460. Title/abstract excluded: Incorrect population.

Cene, C. W., Yanek, L. R., Moy, T. F., Levine, D. M., Becker, L. C., & Becker, D. M. (2008). Sustainability of a multiple risk factor intervention on cardiovascular disease in high-risk African American families. Ethnicity & disease, 18(2), 169-175. Title/abstract excluded: Incorrect population.

Centers for Disease Control and Prevention (1992). Public Health Focus: Effectiveness of Smoking-Control Strategies -- United States. MMWR Morbidity and Mortality Weekly Report, 41(35), 645-647,653. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Chang, H. C., Zimmerman, L. H., & Beck, J. M. (1995). Impact of chart reminders on smoking cessation practices of pulmonary physicians. American Journal of Respiratory and Critical Care Medicine, 152(3), 984-987. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Chapin, J., & Root, W. (2004). Improving obstetrician-gynecologist implementation of smoking cessation guidelines for pregnant women: An interim report of the Americal College of Obstetricians and Gynecologists. Nicotine and Tobacco Research, 6(SUPPL. 2), S253-S257. Title/abstract excluded: Inappropriate study design.

Chilmonczyk, B. A., Palomaki, G. E., Knight, G. J., Williams, J., & Haddow, J. E. (1992). An unsuccessful cotinine-assisted intervention strategy to reduce environmental tobacco smoke exposure during infancy. American Journal of Diseases of Children, 146(3), 357-360. Title/abstract excluded: Incorrect intervention.

Chinet, L. o., Broers, B., Humair, J.-P., & Narring, F. o. (2006). Youth Smoking Cessation: An Update for Primary Care Physicians. Substance Use & Misuse, 41(9), 1251-1261. Title/abstract excluded: Inappropriate study design.

Christakis, D. A., Garrison, M. M., Ebel, B. E., Wiehe, S. E., & Rivara, F. P. (2003). Pediatric smoking prevention interventions delivered by care providers: A systematic review. American Journal of Preventive Medicine, 25(4), 358-362. Title/abstract excluded: Incorrect population.

Clark, J. M., Haverty, S., & Kendall, S. (1990). Helping people to stop smoking: a study of the nurse's role. Journal of advanced nursing, 15(3), 357-363. Title/abstract excluded: Inappropriate study design.

Cleghorn, G. D., Nguyen, M., Roberts, B., Duran, G., Tellez, T., & Alecon, M. (2004). Practice-based interventions to improve health care for Latinos with diabetes. Ethnicity & disease., 14(3 Suppl 1), S117-S121. Title/abstract excluded: Incorrect population.

159


Cohen, D., DiCicco-Bloom, B., Ohman Strickland, P., Headley, A., Orzano, J., Levine, J., et al. (2004). Opportunistic approaches for delivering preventive care in illness visits. Preventive Medicine, 38(5), 565-573. Title/abstract excluded: Inappropriate study design.

Cohen, S. J., Stookey, G. K., Katz, B. P., Drook, C. A., & Smith, D. M. (1989). Encouraging primary care physicians to help smokers quit. A randomized, controlled trial. Ann Intern Med, 110(8), 648-652. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Coleman, T. (2004). ABC of smoking cessation: Cessation interventions in routine health care. British Medical Journal, 328(7440), 631-633. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Coleman, T., Kralikova, E., & Himmerova, V. (2004). [ABC of smoking cessation. Use of simple advice and behavioural support]. Cas Lek Cesk, 143(10), 713-715. Title/abstract excluded: Inappropriate study design.

Coleman, T., Pound, E., Adams, C., Bauld, L., Ferguson, J., & Cheater, F. (2005). Implementing a national treatment service for dependent smokers: initial challenges and solutions. Addiction, 100 Suppl 2, 12-18. Title/abstract excluded: Inappropriate study design.

Coleman, T., Wilson, A., Barrett, S., Wynne, A., & Lewis, S. (2007). Distributing questionnaires about smoking to patients: impact on general practitioners' recording of smoking advice. BMC Health Serv Res, 7, 153. Title/abstract excluded: Inappropriate study design.

Coleman, T., Wynn, A. T., Barrett, S., Wilson, A., & Adams, S. (2001). Intervention study to evaluate pilot health promotion payment aimed at increasing general practitioners' antismoking advice to smokers. Bmj, 323(7310), 435-436. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Collins, R. L., D'Angelo, S., Stearns, S. D., & Campbell, L. R. (2005). Training pediatric residents to provide smoking cessation counseling to parents. TheScientificWorldJournal, 5, 410-419. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Conger, B., Nelson, E. C., & Dietrich, A. J. (1987). Effectiveness of physician antismoking advice. American Journal of Preventive Medicine, 3(4), 223-226. Title/abstract excluded: Inadequate sample size.

Conroy, M. B., Majchrzak, N. E., Regan, S., Silverman, C. B., Schneider, L. I., & Rigotti, N. A. (2005). The association between patient-reported receipt of tobacco intervention at a primary care visit and smokers' satisfaction with their health care. Nicotine & Tobacco Research, 7(Suppl1), S29-S34. Title/abstract excluded: Inappropriate study design.

Copeland, L. (2003). An exploration of the problems faced by young women living in disadvantaged circumstances if they want to give up smoking: Can more be done at general practice level? Family Practice, 20(4), 393-400. Title/abstract excluded: Inappropriate study design.

Córdoba, G. R., Martín-Moros, J. M., Aznar, T. P., Die, M. S., Sanz-Andrés, M. C., & Forés, C., M.D. (1990). [Effect of follow-up offer versus counseling alone in a sample of dissonant smokers. Evaluation of the influence of the tobacco habit of health professionals in this intervention]. Atencion primaria / Sociedad Española.de Medicina de Familia.y Comunitaria., 7, 350-354. Title/abstract excluded: Incorrect population.

Cornuz, J. (2002). Smoking cessation intervention in clinical practice. Onkologie, 25(5), 413-418. Title/abstract excluded: Incorrect population.

Cornuz, J. (2006). Treating tobacco use and dependence in clinical practice. Expert Opinion on Pharmacotherapy, 7(6), 783-792. Title/abstract excluded: Incorrect population.

Cornuz, J., Gilbert, A., Pinget, C., McDonald, P., Slama, K., Salto, E., et al. (2006). Cost-effectiveness of pharmacotherapies for nicotine dependence in primary care settings: A multinational comparison. Tobacco Control, 15(3), 152-159. Title/abstract excluded: Inappropriate study design.

Costa, A. J. (1989). Smoking cessation. Ohio medicine : journal of the Ohio State Medical Association, 85(12), 982-983. Title/abstract excluded: Inappropriate study design.

Cox, L. S., Cupertino, A. P., Mussulman, L. M., Nazir, N., Greiner, K. A., Mahnken, J. D., et al. (2008). Design and baseline characteristics from the KAN-QUIT disease management intervention for rural smokers in primary care. Preventive Medicine, 47(2), 200-205. Title/abstract excluded: Inappropriate study design.

Crawford, M. A., Woodby, L. L., Russell, T. V., Turner, T. J., Hardin, J. M., & Harrington, T. M. (2005). Tobacco use assessment and counselling practices among Alabama primary care physicians. Quality in Primary Care, 13(3), 163-170. Title/abstract excluded: Inappropriate study design.

Cromwell, J., Bartosch, W. J., Fiore, M. C., Hasselblad, V., & Baker, T. (1997). Cost-effectiveness of the clinical practice recommendations in the AHCPR guideline for smoking cessation. Journal of the American Medical Association, 278(21), 1759-1766. Title/abstract excluded: Inappropriate study design.

Cummings, S. R., Coates, T. J., Richard, R. J., Hansen, B., Zahnd, E. G., VanderMartin, R., et al. (1989). Training physicians in counseling about smoking cessation. A randomized trial of the "Quit for Life" program. Ann Intern Med, 110(8), 640-647. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Cummings, S. R., Richard, R. J., Duncan, C. L., Hansen, B., Vander, M. R., Gerbert, B., et al. (1989). Training physicians about smoking cessation: a controlled trial in private practice. Journal of general internal medicine : official journal of the Society for Research and Education in Primary Care Internal Medicine, 4, 482-489. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Cummings, S. R., Rubin, S. M., & Oster, G. (1989). The cost-effectiveness of counseling smokers to quit. Jama, 261(1), 75-79. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Dalal, H. M., & Evans, P. H. (2003). Achieving national service framework standards for cardiac rehabilitation and secondary prevention. British Medical Journal, 326(7387), 481-484. Title/abstract excluded: Incorrect population.

Danish Institute for Health Services Research (DSI) (2004). Smoking cessation interventions in pharmacies. Demand from smokers and the ability of general practitioners to recruit participants (Brief record). Retrieved from http://www.dsi.dk Title/abstract excluded: Incorrect intervention.

Daskalakis, C., Goldberg, R. J., Ockene, J. K., Kalan, K., Hosmer, J., & Pbert, L. (1993). Comparison of patients' and their resident physicians' responses regarding smoking-cessation interventions. Academic Medicine, 68(2), 168-170. Title/abstract excluded: Inappropriate study design.

Davies, M. J., Heller, S., Skinner, T. C., Campbell, M. J., Carey, M. E., Cradock, S., et al. (2008). Effectiveness of the diabetes education and self management for ongoing and newly diagnosed (DESMOND) programme for people with newly diagnosed type 2 diabetes: Cluster

160


randomised controlled trial. BMJ, 336(7642), 491-495. Title/abstract excluded: Incorrect population.

Davis, D. A., Thomson, M. A., Oxman, A. D., & Haynes, R. B. (1995). Changing physician performance. A systematic review of the effect of continuing medical education strategies. Jama, 274(9), 700-705. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Demers, R. Y., Neale, A. V., Adams, R., Trembath, C., & Herman, S. C. (1990). The impact of physicians' brief smoking cessation counseling: A MIRNET study. Journal of Family Practice, 31(6), 625-629. Title/abstract excluded: Incorrect population.

Demmel, R., Rist, F., Hagen, J., Aulhorn, I., Scheuren, B., Scherbaum, N., et al. (2003). Secondary prevention beyond screening and brief advice. Suchtmedizin.in Forschung.und Praxis, 5, 33-36. Title/abstract excluded: Incorrect intervention.

Doherty, W. J., Becker, L. A., Burge, S. K., Whitehead, D., & Baird, I. (1988). Ecological analysis of an intervention study of smoking cessation in medical practices. The Family practice research journal, 8(1), 17-23. Title/abstract excluded: Inappropriate study design.

Donner-Banzhoff, N., Kreienbrock, L., & Baum, E. (1996). Family practitioners' intervention against smoking in Germany and the UK: Does remuneration affect preventive activity? Sozial- und Praventivmedizin, 41(4), 224-230. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Doran, C. M., Duszynski, K. M., Beilby, J. J., & Mattick, R. P. (2006). Use of pharmacotherapies for the management of addictive behaviours in Australian clinical practice. Addictive Behaviors, 31(11), 1947-1958. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Doran, C. M., Pekarsky, B., Gordon, M., & Sanson-Fisher, R. W. (1998). General practitioners' role in preventive medicine: Scenario analysis using smoking as a case study. Addiction, 93(7), 1013-1022. Title/abstract excluded: Inappropriate study design.

Doran, C. M., Valenti, L., Robinson, M., Britt, H., & Mattick, R. P. (2006). Smoking status of Australian general practice patients and their attempts to quit. Addictive Behaviors, 31(5), 758-766. Title/abstract excluded: Inappropriate study design.

Doumit, G., Gattellari, M., Grimshaw, J., & O'Brien, M. A. (2007). Local opinion leaders: effects on professional practice and health care outcomes. Cochrane database of systematic reviews (Online)(1), CD000125. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Dubey, V., Mathew, R., Iglar, K., Moineddin, R., & Glazier, R. (2006). Improving preventive service delivery at adult complete health check-ups: The Preventive health Evidence-based Recommendation Form (PERFORM) cluster randomized controlled trial. BMC Family Practice, 7, 44. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Duncan, C., Stein, M. J., & Cummings, S. R. (1991). Staff Involvement and Special Follow-Up Time Increase Physicians' Counseling About Smoking Cessation: A Controlled Trial. American Journal of Public Health, 81(7), 899-901. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Duran, A., Runkle, I., Matia, P., de Miguel, M. P., Garrido, S., Cervera, E., et al. (2008). Family physician and endocrinologist coordination as the basis for diabetes care in clinical practice. BMC Endocrine Disorders, 8(-). Title/abstract excluded: Incorrect population.

Dyer, T. A., & Robinson, P. G. (2006). General health promotion in general dental practice - The involvement of the dental team Part 1: A review of the evidence of

effectiveness of brief public health interventions. British Dental Journal, 200(12), 679-685. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Eakin, E. G., Lichtenstein, E., Severson, H. H., Stevens, V. J., Vogt, T. M., & Hollis, J. F. (1998). Use of tailored videos in primary care smoking cessation interventions. Health Education Research, 13(4), 519-527. Title/abstract excluded: Inappropriate study design.

Earles, J., Folen, R. A., Ma, M., Kellar, M., Geralde, R., & Dydek, G. (2002). Clinical effectiveness of sustained-release bupropion and behavior therapy for tobacco dependence in a clinical setting. Military Medicine, 167(11), 923-925. Title/abstract excluded: Inappropriate study design.

Easton, A., Husten, C., Malarcher, A., Elon, L., Caraballo, R., Ahluwalia, I., et al. (2001). Smoking cessation counseling by primary care women physicians: Women Physicians' Health Study. Women & Health, 32(4), 77-91. Title/abstract excluded: Incorrect intervention.

Edwards, D., Freeman, T., Litt, J., & Roche, A. M. (2006). GPs' confidence in and barriers to implementing smoking cessation activities: Compared to dentists, dental hygienists and pharmacists. Australian Journal of Primary Health, 12(3), 117-125. Title/abstract excluded: Inappropriate study design.

Ellerbeck, E. F., Choi, W. S., McCarter, K., Jolicoeur, D. G., Greiner, A., & Ahluwalia, J. S. (2003). Impact of patient characteristics on physician's smoking cessation strategies. Preventive Medicine: An International Journal Devoted to Practice and Theory, 36(4), 464-470. Title/abstract excluded: Incorrect intervention.

Everett, K., Odendaal, H. J., & Steyn, K. (2005). Doctors' attitudes and practices regarding smoking cessation during pregnancy. South African Medical Journal, 95(5), 350-354. Title/abstract excluded: Inappropriate study design.

Farmer, A. P., Legare, F., Turcot, L., Grimshaw, J., Harvey, E., McGowan, J. L., et al. (2008). Printed educational materials: effects on professional practice and health care outcomes. Cochrane database of systematic reviews (Online)(3), CD004398. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Feenstra, T. L., Hamberg-Van Reenen, H. H., Hoogenveen, R. T., & Rutten-Van Molken, M. P. M. H. (2005). Cost-effectiveness of face-to-face smoking cessation interventions: A dynamic modeling study. Value in Health, 8(3), 178-190. Title/abstract excluded: Inappropriate study design.

Fidler, W., & Lambert, T. W. (2001). A prescription for health: a primary care based intervention to maintain the non-smoking status of young people. Tobacco Control, 10(1), 23-26. Title/abstract excluded: Incorrect population.

Fiore, M. C., Pierce, J. P., Remington, P. L., & Fiore, B. J. (1990). Cigarette smoking: The clinician's role in cessation, prevention, and public health. Disease-a-Month, 36(4), 181-242. Title/abstract excluded: Inappropriate study design.

Fiscella, K., & Franks, P. (1996). Cost-effectiveness of the transdermal nicotine patch as an adjunct to physicians' smoking cessation counseling. Journal of the American Medical Association, 275(16), 1247-1251. Title/abstract excluded: Inappropriate study design.

Fleming, M. F., Barry, K. L., Manwell, L. B., Johnson, K., & London, R. (1997). Brief physician advice for problem alcohol drinkers. A randomized controlled trial in community-based primary care practices. JAMA : the journal of the American Medical Association, 277,

161


1039-1045. Title/abstract excluded: Inappropriate outcomes.

Fleming, M. F., Manwell, L. B., Barry, K. L., Adams, W., & Stauffacher, E. A. (1999). Brief physician advice for alcohol problems in older adults: a randomized community-based trial. Journal of Family Practice, 48(5), 378-384. Title/abstract excluded: Inappropriate outcomes.

Fleming, M. F., Mundt, M. P., French, M. T., Manwell, L. B., Stauffacher, E. A., & Barry, K. L. (2000). Benefit-cost analysis of brief physician advice with problem drinkers in primary care settings. Medical Care, 38(1), 7-18. Title/abstract excluded: Inappropriate outcomes.

Floyd, R. L., Belodoff, B., Sidhu, J., Schulkin, J., Ebrahim, S. H., & Sokol, R. J. (2001). A survey of obstetrician-gynecologists on their patients' use of tobacco and other drugs during pregnancy. Prenatal and Neonatal Medicine, 6(4), 201-207. Title/abstract excluded: Inappropriate study design.

Franklin, J. L., Williams, A. F., Kresch, G. M., Painter, J. T., Torges, K., White, J., et al. (1992). Smoking cessation interventions by family physicians in Texas. Texas medicine, 88(11), 60-64. Title/abstract excluded: Inappropriate study design.

Freemantle, N., Harvey, E. L., Wolf, F., Grimshaw, J., Grilli, R., & Bero, L. A. (1997). WITHDRAWN: Printed educational materials: effects on professional practice and health care outcomes. Cochrane database of systematic reviews (Online)(2), CD000172. Title/abstract: Included. Full paper excluded: Other.

Freemantle, N., Harvey, E. L., Wolf, F., Grimshaw, J., Grilli, R., & Bero, L. A. (2000). Printed educational materials: effects on professional practice and health care outcomes. Cochrane database of systematic reviews (Online)(2), CD000172. Title/abstract: Included. Full paper excluded: Other.

Garg, A., Serwint, J. R., Higman, S., Kanof, A., Schell, D., Colon, I., et al. (2007). Self-efficacy for smoking cessation counseling parents in primary care: An office-based intervention for pediatricians and family physicians. Clinical Pediatrics, 46(3), 252-257. Title/abstract excluded: Inappropriate outcomes.

Geller, A. C., Lash, T. L., Siegel, B., Annas, G. D., & Prout, M. N. (2005). The primary care pediatrician's influence on medical student's performance of smoking assessments and counseling. Preventive Medicine: An International Journal Devoted to Practice and Theory, 41(1), 47-52. Title/abstract excluded: Incorrect intervention.

Gerbert, B., Berg, S. S., Mancuso, M., Caspers, N., McPhee, S., Null, D., et al. (2003). Using innovative video doctor technology in primary care to deliver brief smoking and alcohol intervention. Health promotion practice, 4, 249-261. Title/abstract excluded: Incorrect population.

Gilbert, H., Nazareth, I., & Sutton, S. (2007). Assessing the feasibility of proactive recruitment of smokers to an intervention in general practice for smoking cessation using computer-tailored feedback reports. Family Practice, 24(4), 395-400. Title/abstract excluded: Incorrect population.

Gilbert, H., Nazareth, I., Sutton, S., Morris, R., & Godfrey, C. (2008). Effectiveness of computer-tailored Smoking Cessation Advice in Primary Care (ESCAPE): a Randomised Trial. Trials, 9, 23. Title/abstract excluded: Inappropriate study design.

Gilbert, J. R., Wilson, D. M., Best, J. A., Taylor, D. W., Lindsay, E. A., Singer, J., et al. (1989). Smoking cessation in primary care. A randomized controlled trial of nicotine-bearing chewing gum. The Journal of family practice, 28, 49-55. Title/abstract excluded: Incorrect population.

Gilbert, J. R., Wilson, D. M., Singer, J., Lindsay, E. A., Willms, D. G., Best, J. A., et al. (1992). A family physician smoking cessation program: an evaluation of the role of follow-up visits. American Journal of Preventive Medicine, 8, 91-95. Title/abstract excluded: Incorrect population.

Godfrey, C., Parrott, S., Coleman, T., & Pound, E. (2005). The cost-effectiveness of the English smoking treatment services: evidence from practice. Addiction, 100 Suppl 2, 70-83. Title/abstract excluded: Inappropriate study design.

Goldberg, R. J., Ockene, I. S., Ockene, J. K., Merriam, P., & Kristeller, J. (1993). Physicians' attitudes and reported practices toward smoking intervention. Journal of Cancer Education, 8(2), 133-139. Title/abstract excluded: Inappropriate study design.

Goldstein, M. G., DePue, J. D., Monroe, A. D., Lessne, C. W., Rakowski, W., Prokhorov, A., et al. (1998). A population-based survey of physician smoking cessation counseling practices. Preventive Medicine: An International Journal Devoted to Practice and Theory, 27(5, Pt.1), 720-729. Title/abstract excluded: Inappropriate study design.

Goldstein, M. G., & Niaura, R. (1988). Intervention by general practitioners to reduce smoking. British medical journal (Clinical research ed.), 296(6618), 358-359. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Goldstein, M. G., Whitlock, E. P., & DePue, J. (2004). Multiple behavioral risk factor interventions in primary care: Summary of research evidence. American Journal of Preventive Medicine, 27(SUPPL.), 61-79. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Good, M. J., Frazier, L. M., Wetta-Hall, R., Ablah, E., & Molgaard, C. A. (2004). Kansas office-based nurses' evaluation of patient tobacco cessation activities. Journal of Community Health Nursing, 21(2), 77-85. Title/abstract excluded: Inappropriate study design.

Gorgojo Jimenez, L., Gonzalez Enriquez, J., & Salvador Llivina, T. (2003). Efficacy, effectiveness and cost-effectiveness of interventions for smoking cessation IPE-03/40 (Public report). . Anales del Instituto de Estudios Madrilenos 76. Retrieved from http://www.crd.york.ac.uk/CRDWeb/ShowRecord.asp?View=Full&ID=32004000127 Title/abstract excluded: Inappropriate study design.

Grandes, G., Cortada, J. M., & Arrazola, A. (2000). An evidence-based programme for smoking cessation: Effectiveness in routine general practice. British Journal of General Practice, 50(459), 803-807. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Grandes, G., Cortada, J. M., Arrazola, A., & Laka, J. P. (2003). Predictors of long-term outcome of a smoking cessation programme in primary care. British Journal of General Practice, 53(487), 101-107. Title/abstract excluded: Inappropriate study design.

Green, L. (2005). Helping smokers to stop: advice for pharmacists in England. Retrieved from http://www.nice.org.uk/aboutnice/whoweare/aboutthehda/hdapublications/helping_smokers_to_stop_advice_for_pharmacists_in_england.jsp Title/abstract excluded: Inappropriate study design.

Green, L. W., Orleans, C. T., Ottoson, J. M., Cameron, R., Pierce, J. P., & Bettinghaus, E. P. (2006). Inferring Strategies for Disseminating Physical Activity Policies, Programs, and Practices from the Successes of Tobacco Control. American Journal of Preventive Medicine, 31(4 SUPPL.), 66-81. Title/abstract excluded: Inappropriate study design.

Greydanus, D. E., & Patel, D. R. (2003). Substance abuse in adolescents: A complex conundrum for the clinician.

162


Pediatric Clinics of North America, 50(5), 1179-1223. Title/abstract excluded: Inappropriate study design.

Groner, J. A., Ahijevych, K., Grossman, L. K., & Rich, L. N. (2000). The impact of a brief intervention on maternal smoking behavior. Pediatrics, 105(1 Pt 3), 267-271. Title/abstract excluded: Incorrect population.

Hall, S., Vogt, F., & Marteau, T. M. (2005). A short report: Survey of practice nurses' attitudes towards giving smoking cessation advice. Family Practice, 22(6), 614-616. Title/abstract excluded: Inappropriate study design.

Hallett, R. (1983). Intervention against smoking and its relationship to general practitioners' smoking habits. Journal of the Royal College of General Practitioners, 33(254), 565-567. Title/abstract excluded: Inappropriate study design.

Hansen, L. J., Olivarius, N. D. F., Siersma, V., Drivsholm, T., & Andersen, J. S. (2004). Individualised-treatment goals in diabetes care. Scandinavian Journal of Primary Health Care, 22(2), 71-77. Title/abstract excluded: Inappropriate outcomes.

Hartmann, K. E., & Thorp, J. (1996). Scripted intervention: A map for smoking cessation counseling. Primary Care Update for Ob/Gyns, 3(6), 228-232. Title/abstract excluded: Inappropriate study design.

Hartmann, K. E., Wechter, M. E., Payne, P., Salisbury, K., Jackson, R. D., & Melvin, C. L. (2007). Best practice smoking cessation intervention and resource needs of prenatal care providers. Obstetrics and Gynecology, 110(4), 765-770. Title/abstract excluded: Inappropriate study design.

Haug, K., Fugelli, P., Aaro, L. E., & Foss, O. P. (1994). Is smoking intervention in general practice more successful among pregnant than non-pregnant women. Family Practice, 11(2), 111-116. Title/abstract excluded: Incorrect intervention.

Health Development Agency (NHS) (2000). Smoking cessation guidelines for health professionals: an update. Retrieved from http://www.nice.org.uk/aboutnice/whoweare/aboutthehda/hdapublications/smoking_cessation_guidelines_for_health_professionals_an_update.jsp Title/abstract excluded: Inappropriate study design.

Health Development Agency (NHS) (2003). Standard for training in smoking cessation treatments. Retrieved from http://www.nice.org.uk/aboutnice/whoweare/aboutthehda/hdapublications/standard_for_training_in_smoking_cessation_treatments.jsp Title/abstract excluded: Inappropriate study design.

Health Development Agency (NHS) (2004). Helping smokers stop: a guide for the dental team. Retrieved from http://www.nice.org.uk/aboutnice/whoweare/aboutthehda/hdapublications/helping_smokers_stop_a_guide_for_the_dental_team.jsp Title/abstract excluded: Inappropriate study design.

Health Development Agency (NHS) (2004). Smoking cessation training survey: report on findings. Retrieved from http://www.nice.org.uk/aboutnice/whoweare/aboutthehda/hdapublications/smoking_cessation_training_survey_report_on_findings.jsp Title/abstract excluded: Inappropriate study design.

Heath, J., Andrews, J., Kelley, F. J., & Sorrell, J. (2004). Caught in the middle: experiences of tobacco-dependent nurse practitioners. Journal of the American Academy of Nurse Practitioners, 16(9), 396-401. Title/abstract excluded: Inappropriate study design.

Hebert, J. R., Kristeller, J., Ockene, J. K., Landon, J., Luippold, R., Goldberg, R. J., et al. (1992). Patient characteristics and the effect of three physician-delivered smoking interventions. Preventive Medicine, 21(5), 557-573. Title/abstract excluded: Incorrect population.

Helwig, A. L., Swain, G. R., & Gottlieb, M. (1998). Smoking cessation intervention: the practices of maternity care providers. The Journal of the American Board of Family Practice / American Board of Family Practice, 11(5), 336-340. Title/abstract excluded: Inappropriate study design.

Herbert, R. (2008). What's new in Nicotine & Tobacco Research? Nicotine & Tobacco Research, 10(2), 247-252. Title/abstract excluded: Inappropriate study design.

Herbert, R., Coleman, T., & Britton, J. (2005). U.K. general practitioners' beliefs, attitudes, and reported prescribing of nicotine replacement therapy in pregnancy. Nicotine and Tobacco Research, 7(4), 541-546. Title/abstract excluded: Inappropriate study design.

Hickner, J., Cousineau, A., & Messimer, S. (1990). Smoking cessation during pregnancy: strategies used by Michigan family physicians. The Journal of the American Board of Family Practice / American Board of Family Practice, 3(1), 39-42. Title/abstract excluded: Inappropriate study design.

Hilberink, S. R., Jacobs, J. E., Bottema, B. J. A. M., De Vries, H., & Grol, R. P. T. M. (2005). Smoking cessation in patients with COPD in daily general practice (SMOCC): Six months' results. Preventive Medicine, 41(5-6), 822-827. Title/abstract excluded: Incorrect population.

Hill, D. (1983). The family physician's influence on patients' smoking habits. Australian family physician, 12(9), 678-679. Title/abstract excluded: Inappropriate study design.

Hoch, E., Muehlig, S., Hofler, M., Lieb, R., & Wittchen, H. U. (2004). How prevalent is smoking and nicotine dependence in primary care in Germany? Addiction, 99(12), 1586-1598. Title/abstract excluded: Inappropriate study design.

Hoffman, L. H., Strutton, D. R., Stang, P. E., & Hogue, S. L. (2002). Impact of smoking on respiratory illness-related outpatient visits among 50- to 75-year-olds in the United States. Clinical Therapeutics, 24(2), 317-324. Title/abstract excluded: Inappropriate study design.

Hogg, W., Baskerville, N., & Lemelin, J. (2005). Cost savings associated with improving appropriate and reducing inappropriate preventive care: cost-consequences analysis. BMC Health Serv Res, 5(1), 20. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Hollis, J. F., Lichtenstein, E., Mount, K., Vogt, T. M., & Stevens, V. J. (1991). Nurse-assisted smoking counseling in medical settings: minimizing demands on physicians. Preventive Medicine, 20, 497-507. Title/abstract excluded: Incorrect intervention.

Hollis, J. F., Lichtenstein, E., Vogt, T. M., Stevens, V. J., & Biglan, A. (1993). Nurse-assisted counseling for smokers in primary care. Ann Intern Med, 118(7), 521-525. Title/abstract: Included. Full paper excluded: Inappropriate comparator.

Hollis, J. F., Polen, M. R., Whitlock, E. P., Lichtenstein, E., Mullooly, J. P., Velicer, W. F., et al. (2005). Teen reach: outcomes from a randomized, controlled trial of a tobacco reduction program for teens seen in primary medical care. Pediatrics, 115(4), 981-989. Title/abstract excluded: Incorrect population.

Horn, K., Dino, G., Hamilton, C., Noerachmanto, N., & Zhang, J. (2008). Feasibility of a smoking cessation intervention for teens in the emergency department: reach, implementation fidelity, and acceptability. American journal of critical care : an official publication, American Association of Critical Care Nurses, 17, 205-216. Title/abstract excluded: Incorrect population.

163


Hughes, D. M., McLeod, M., Garner, B., & Goldbloom, R. B. (1991). Controlled trial of a home and ambulatory program for asthmatic children. Pediatrics, 87, 54-61. Title/abstract excluded: Incorrect intervention.

Hughes, J. R., Gust, S. W., Keenan, R. M., Fenwick, J. W., & Healy, M. L. (1989). Nicotine vs placebo gum in general medical practice. Journal of the American Medical Association, 261(9), 1300-1305. Title/abstract excluded: Incorrect population.

Humair, J. P., & Ward, J. (1998). Smoking-cessation strategies observed in videotaped general practitioners. American Journal of Preventive Medicine, 14(1), 1-8. Title/abstract excluded: Inappropriate study design.

Husten, C. G., & Manley, M. W. (1990). How to help your patients stop smoking. American Family Physician, 42(4), 1017-1026. Title/abstract excluded: Inappropriate study design.

Institute for Clinical Systems Improvement (ICSI) (2004). Tobacco use prevention and cessation for adults and mature adolescents. from Bloomington (MN): Institute for Clinical Systems Improvement (ICSI): http://www.guideline.gov/summary/summary.aspx?doc_id=5454&nbr=003731&string=smoking Title/abstract excluded: Inappropriate study design.

Irvin, C. B., Wyer, P. C., & Gerson, L. W. (2000). Preventive care in the emergency department, Part II: Clinical preventive services - An emergency medicine evidence-based review. Academic Emergency Medicine, 7(9), 1042-1054. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Ivers, R. G. (2003). A review of tobacco interventions for indigenous Australians. Australian and New Zealand Journal of Public Health, 27(3), 294-299. Title/abstract: Included. Full paper excluded: Inappropriate study design .

Ivers, R. G. (2004). An evidence-based approach to planning tobacco interventions for Aboriginal people. Drug and Alcohol Review, 23(1), 5-9. Title/abstract excluded: Incorrect intervention.

Jackson, G., Bobak, A., Chorlton, I., Fowler, G., Hall, R., Khimji, H., et al. (2001). Smoking cessation: A consensus statement with special reference to primary care. International Journal of Clinical Practice, 55(6), 385-392. Title/abstract excluded: Inappropriate study design.

Jackson, P. H., Stapleton, J. A., Russel, M. A. H., & Merriman, R. J. (1989). Nicotine gum use and outcome in a general practitioner intervention against smoking. Addictive Behaviors, 14(3), 335-341. Title/abstract excluded: Inappropriate study design.

Jackson, P. H., Stapleton, J. A., & Russell, M. A. H. (1986). Predictors of outcome in a general practitioner intervention against smoking. Preventive Medicine, 15(3), 244-253. Title/abstract excluded: Inappropriate study design.

Jackson, R. S., Leininger, L. S., Harris, R. P., & Kaluzny, A. D. (1994). Implementing continuous quality improvement in primary care: implications for preventive services. J Ambul Care Manage, 17(3), 8-14. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Jaen, C. R., Crabtree, B. F., Zyzanski, S. J., Goodwin, M. A., & Stange, K. C. (1998). Making time for tobacco cessation counseling. Journal of Family Practice, 46(5), 425-428. Title/abstract excluded: Inappropriate study design.

Jaen, C. R., McIlvain, H., Pol, L., Phillips, R. L., Flocke, S., & Crabtree, B. F. (2001). Tailoring tobacco counseling to the competing demands in the clinical encounter. Journal of Family Practice, 50(10), 859-863. Title/abstract excluded: Inappropriate study design.

Jaen, C. R., Stange, K. C., Tumiel, L. M., & Nutting, P. (1997). Missed opportunities for prevention: Smoking cessation counseling and the competing demands of practice. Journal of Family Practice, 45(4), 348-354. Title/abstract excluded: Inappropriate study design.

Jamrozik, K., Vessey, M., Fowler, G., Wald, N., Parker, G., & Van Vunakis, H. (1984). Controlled trial of three different antismoking interventions in general practice. British Medical Journal, 288, 1499-1503. Title/abstract excluded: Incorrect population.

Jamtvedt, G., Young, J. M., Kristoffersen, D. T., O'Brien, M. A., & Oxman, A. D. (2006). Audit and feedback: effects on professional practice and health care outcomes. Cochrane database of systematic reviews (Online)(2), CD000259. Title/abstract: Included. Full paper excluded: Other.

John, J. H., Yudkin, P., Murphy, M., Ziebland, S., & Fowler, G. H. (1997). Smoking cessation interventions for dental patients--attitudes and reported practices of dentists in the Oxford region. British Dental Journal, 183(10), 359-364. Title/abstract excluded: Inappropriate study design.

Johnston, J. M., Chan, S. S. C., Chan, S. K. K., Lam, T. H., Chi, I., & Leung, G. M. (2005). Training nurses and social workers in smoking cessation counseling: A population needs assessment in Hong Kong. Preventive Medicine, 40(4), 389-406. Title/abstract excluded: Inappropriate study design.

Jorgensen, T., Borch-Johnsen, K., Thomsen, T. F., Ibsen, H., Glumer, C., & Pisinger, C. (2003). A randomized non-pharmacological intervention study for prevention of ischaemic heart disease: Baseline results Inter99 (1). European Journal of Cardiovascular Prevention and Rehabilitation, 10(5), 377-386. Title/abstract excluded: Incorrect population.

Kai, T., Maki, T., Takahashi, S., & Warner, D. O. (2008). Perioperative tobacco use interventions in Japan: A survey of thoracic surgeons and anaesthesiologists. British Journal of Anaesthesia, 100(3), 404-410. Title/abstract excluded: Inappropriate study design.

Kaplan, C. P., Perez-Stable, E. J., Fuentes-Afflick, E., Gildengorin, V., Millstein, S., & Juarez-Reyes, M. (2004). Smoking Cessation Counseling with Young Patients: The Practices of Family Physicians and Pediatricians. Archives of Pediatrics and Adolescent Medicine, 158(1), 83-90. Title/abstract excluded: Inappropriate study design.

Katz, D. A., Brown, R. B., Muehlenbruch, D. R., Fiore, M. C., & Baker, T. B. (2004). Implementing Guidelines for Smoking Cessation: Comparing the Efforts of Nurses and Medical Assistants. American Journal of Preventive Medicine, 27(5), 411-416. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Katz, D. A., Muehlenbruch, D. R., Brown, R. B., Fiore, M. C., & Baker, T. B. (2002). Effectiveness of a clinic-based strategy for implementing the AHRQ smoking cessation guideline in primary care. Preventive Medicine: An International Journal Devoted to Practice and Theory, 35(3), 293-302. Title/abstract excluded: Incorrect intervention.

Kerner, J. F., Strickman, N. A., & Bleadingheiser, L. J. (1983). Attitudes and behaviors of primary care physicians with respect to cancer prevention. Journal of Psychosocial Oncology, 1(3), 51-63. Title/abstract excluded: Inappropriate study design.

Kharicha, K., Iliffe, S., Harari, D., Swift, C., Gillmann, G., & Stuck, A. E. (2007). Health risk appraisal in older people 1: are older people living alone an "at-risk" group? The British journal of general practice : the journal of the Royal College of General Practitioners, 57, 271-276. Title/abstract excluded: Inappropriate outcomes.

164


Kim, J. R., Lee, M. S., Hwang, J. Y., & Lee, J. D. (2005). Efficacy of a smoking cessation intervention using the AHCPR guideline tailored for Koreans: a randomized controlled trial. Health promotion international, 20, 51-59. Title/abstract excluded: Incorrect population.

Klerman, L. V. (2004). Protecting children: Reducing their environmental tobacco smoke exposure. Nicotine and Tobacco Research, 6(SUPPL. 2), S239-S252. Title/abstract excluded: Inappropriate study design.

Knight, R. A., & Hay, D. A. (1989). The relevance of the health belief model to Australian smokers. Social science & medicine, 28, 1311-1314. Title/abstract excluded: Inappropriate study design.

Kottke, T. E., Battista, R. N., DeFriese, G. H., & Brekke, M. L. (1988). Attributes of successful smoking cessation interventions in medical practice. A meta-analysis of 39 controlled trials. Journal of the American Medical Association, 259(19), 2883-2889. Title/abstract excluded: Inappropriate study design.

Kottke, T. E., Solberg, L. I., Conn, S., Maxwell, P., Thomasberg, M., Brekke, M. L., et al. (1990). A comparison of two methods to recruit physicians to deliver smoking cessation interventions. Archives of Internal Medicine, 150(7), 1477-1481. Title/abstract excluded: Inappropriate study design.

Kotz, D., Wagena, E. J., & Wesseling, G. (2007). Smoking cessation practices of Dutch general practitioners, cardiologists, and lung physicians. Respiratory Medicine, 101(3), 568-573. Title/abstract excluded: Inappropriate study design.

Kotz, D., Wesseling, G., Huibers, M. J., & van Schayck, O. C. (2007). Efficacy of confrontational counselling for smoking cessation in smokers with previously undiagnosed mild to moderate airflow limitation: study protocol of a randomized controlled trial. BMC public health, 7, 332. Title/abstract excluded: Incorrect population.

Kreuter, M. W., Chheda, S. G., & Bull, F. C. (2000). How does physician advice influence patient behavior? Evidence for a priming effect. Archives of Family Medicine, 9, 426-433. Title/abstract excluded: Incorrect population.

Kristeller, J. L., & Ockene, J. K. (1996). Tobacco curriculum for medical students, residents and practicing physicians. Indiana medicine : the journal of the Indiana State Medical Association, 89(2), 199-204. Title/abstract excluded: Inappropriate study design.

Kulig, J. W. (2005). Tobacco, alcohol, and other drugs: The role of the pediatrician in prevention, identification, and management of substance abuse. Pediatrics, 115(3), 816-821. Title/abstract excluded: Inappropriate study design.

Lai, D. T. C., & Lau, D. M. W. (2005). A case study of perceptions about smoking and smoking cessation interventions among members of an uniform disciplinary force in Hong Kong. Hong Kong Practitioner, 27(9), 334-338. Title/abstract excluded: Inappropriate study design.

Lancaster, T., Dobbie, W., Katrina, V., Yudkin, P., Murphy, M., & Fowler, G. (1999). Randomized trial of nurse-assisted strategies for smoking cessation in primary care. British Journal of General Practice, 49(440), 191-194. Title/abstract excluded: Incorrect population.

Lancaster, T., & Stead, L. (2004). Physician advice for smoking cessation. Cochrane database of systematic reviews (Online : Update Software), -(4), CD000165. Title/abstract excluded: Incorrect population.

Lancaster, T., & Stead, L. F. (2005). Individual behavioural counselling for smoking cessation. Lancaster.Tim., Stead.Lindsay.F.Individual.behavioural counselling for smoking.cessation.Cochrane Database of

Systematic Reviews: Reviews 2005 Issue.2 John.Wiley.& Sons., Ltd.Chichester, UK DOI.: 10.1002./14651858.CD001292.pub2. Title/abstract excluded: Incorrect population.

Lawn, S., & Condon, J. (2006). Psychiatric nurses' ethical stance on cigarette smoking by patients: Determinants and dilemmas in their role in supporting cessation. International Journal of Mental Health Nursing, 15(2), 111-118. Title/abstract excluded: Inappropriate study design.

Lawrance, K., & Lawler, S. A. (2008). Campus physicians' tobacco interventions with university students: A descriptive study of 16 Ontario university clinics. Patient Education and Counseling, 70(2), 187-192. Title/abstract excluded: Inappropriate study design.

Lee, A. H. (2008). A pilot intervention for pregnant women in Sichuan, China on passive smoking. Patient Education and Counseling, 71(3), 396-401. Title/abstract excluded: Inappropriate study design.

Lennox, A. S., Osman, L. M., Reiter, E., Robertson, R., Friend, J., McCann, I., et al. (2001). Cost effectiveness of computer tailored and non-tailored smoking cessation letters in general practice: randomised controlled trial. Bmj, 322(7299), 1396. Title/abstract excluded: Incorrect intervention.

Lichtenstein, E., & Hollis, J. (1992). Patient referral to a smoking cessation program: who follows through? J Fam Pract, 34(6), 739-744. Title/abstract excluded: Inappropriate outcomes.

Lin, J. W., Chu, P. L., Liou, J. M., & Hwang, J. J. (2007). Applying a multiple screening program aided by a guideline-driven computerized decision support system - A pilot experience in Yun-Lin, Taiwan. Journal of the Formosan Medical Association, 106(1), 58-68. Title/abstract excluded: Inappropriate study design.

Lindsay, E. A., Ockene, J. K., Hymowitz, N., Giffen, C., Berger, L., & Pomrehn, P. (1994). Physicians and smoking cessation. A survey of office procedures and practices in the Community Intervention Trial for Smoking Cessation. Archives of Family Medicine, 3(4), 341-348. Title/abstract excluded: Inappropriate study design.

Lindsay, E. A., Wilson, D. M., Best, J. A., Willms, D. G., Singer, J., Gilbert, J. R., et al. (1989). A randomized trial of physician training for smoking cessation. American Journal of Health Promotion, 3(3), 11-18. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Litaker, D., Flocke, S. A., Frolkis, J. P., & Stange, K. C. (2005). Physicians' attitudes and preventive care delivery: Insights from the DOPC study. Preventive Medicine: An International Journal Devoted to Practice and Theory, 40(5), 556-563. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Litt, J. (2005). Smoking and GPs: time to cough up: successful interventions in general practice. Australian family physician, 34(6), 425-429. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Litt, J. (2005). What's new in smoking cessation? Australian Prescriber, 28(3), 73-75+79. Title/abstract excluded: Inappropriate study design.

Livaudais, J. C., Kaplan, C. P., Haas, J. S., Perez-Stable, E. J., Stewart, S., & Jarlais, G. D. (2005). Lifestyle behavior counseling for women patients among a sample of California physicians. Journal of women's health (2002), 14(6), 485-495. Title/abstract excluded: Inappropriate study design.

Longo, D. R., Stone, T. T., Phillips, R. L., Everett, K. D., Kruse, R. L., Jaen, C. R., et al. (2006). Characteristics of smoking cessation guideline use by primary care

165


physicians. Missouri medicine., 103(2), 180-184. Title/abstract excluded: Inappropriate study design.

Lowenstein, S. R., Tomlinson, D., Koziol-McLain, J., & Prochazka, A. (1995). Smoking habits of emergency department patients: an opportunity for disease prevention. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine, 2(3), 165-171. Title/abstract excluded: Inappropriate study design.

Lowry, R. J., Hardy, S., Jordan, C., & Wayman, G. (2004). Using social marketing to increase recruitment of pregnant smokers to smoking cessation service: A success story. Public Health, 118(4), 239-243. Title/abstract excluded: Inappropriate study design.

Majeskie, M. R., Fiore, M. C., & Baker, T. B. (1998). Quality in quantity: in support of consistent, physician-delivered smoking intervention. Tobacco Control, 7(4), 422-423. Title/abstract: Included. Full paper excluded: Inappropriate study design .

Makni, H., O'Loughlin, J. L., Tremblay, M., Gervais, A., Lacroix, C., Dery, V., et al. (2002). Smoking prevention counseling practices of Montreal general practitioners. Archives of Pediatrics and Adolescent Medicine, 156(12), 1263-1267. Title/abstract excluded: Inappropriate study design.

Marcy, T. W., Kaplan, B., Connolly, S. W., Michel, G., Shiffman, R. N., & Flynn, B. S. (2008). Developing a decision support system for tobacco use counselling using primary care physicians. Informatics in Primary Care, 16(2), 101-109. Title/abstract excluded: Inappropriate study design.

Marcy, T. W., Skelly, J., Shiffman, R. N., & Flynn, B. S. (2005). Facilitating adherence to the tobacco use treatment guideline with computer-mediated decision support systems: Physician and clinic office manager perspectives. Preventive Medicine: An International Journal Devoted to Practice and Theory, 41(2), 479-487. Title/abstract excluded: Inappropriate study design.

Marcy, T. W., Thabault, P., Olson, J., Tooze, J. A., Liberty, B., & Nolan, S. (2003). Smoking status identification: Two managed care organizations' experiences with a pilot project to implement identification systems in independent practice associations. American Journal of Managed Care, 9(10), 672-676. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Martin, T. N., Irving, R. J., Sutherland, M., Sutherland, K., & Bloomfield, P. (2005). Improving secondary prevention in coronary bypass patients: closing the audit loop. Heart (British Cardiac Society), 91(4), 456-459. Title/abstract excluded: Inappropriate study design.

Martinez-Raga, J., Keaney, F., Sutherland, G., Perez-Galvez, B., & Strang, J. (2003). Treatment of nicotine dependence with bupropion SR: Review of its efficacy, safety and pharmacological profile. Addiction Biology, 8(1), 13-21. Title/abstract excluded: Incorrect population.

McBride, P. E., Plane, M. B., Underbakke, G., Brown, R. L., & Solberg, L. I. (1997). Smoking screening and management in primary care practices. Archives of Family Medicine, 6(2), 165-172. Title/abstract excluded: Incorrect intervention.

McCurry, N., Thompson, K., Parahoo, K., O'Doherty, E., & Doherty, A. M. (2002). Pregnant women's perception of the implementation of smoking cessation advice. Health Education Journal, 61(1), 20-31. Title/abstract excluded: Inappropriate study design.

McDaniel, A. M., Benson, P. L., Roesener, G. H., & Martindale, J. (2005). An integrated computer-based system to support nicotine dependence treatment in primary care. Nicotine & Tobacco Research, 7(Suppl1), S57-

S66. Title/abstract: Included. Full paper excluded: Incorrect population.

McEwen, A., Akotia, N., & West, R. (2001). General practitioners' views on the English national smoking cessation guidelines. Addiction, 96(7), 997-1000. Title/abstract excluded: Inappropriate study design.

McEwen, A., Preston, A., & West, R. (2002). Effect of a GP desktop resource on smoking cessation activities of general practitioners. Addiction (Abingdon, England), 97(5), 595-597. Title/abstract: Included. Full paper excluded: Inappropriate study design .

McEwen, A., & West, R. (2001). Smoking cessation activities by general practitioners and practice nurses. Tobacco Control, 10(1), 27-32. Title/abstract excluded: Inappropriate study design.

McEwen, A., West, R., Owen, L., & Raw, M. (2005). General practitioners' views on and referral to NHS smoking cessation services. Public Health, 119(4), 262-268. Title/abstract excluded: Inappropriate study design.

McEwen, A., West, R., & Preston, A. (2006). Triggering anti-smoking advice by GPs: Mode of action of an intervention stimulating smoking cessation advice by GPs. Patient Education and Counseling, 62(1), 89-94. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

McLeod, D., Benn, C., Pullon, S., Viccars, A., White, S., Cookson, T., et al. (2003). The midwife's role in facilitating smoking behaviour change during pregnancy. Midwifery, 19(4), 285-297. Title/abstract excluded: Inappropriate study design.

McMenamin, S. B., Schauffler, H. H., Shortell, S. M., Rundall, T. G., & Gillies, R. R. (2003). Support for smoking cessation interventions in physician organizations: Results from a national study. Medical Care, 41(12), 1396-1406. Title/abstract: Included. Full paper excluded: Incorrect population.

McVea, K., Crabtree, B. F., Medder, J. D., Susman, J. L., Lukas, L., McIlvain, H. E., et al. (1996). An ounce of prevention? Evaluation of the 'Put Prevention into Practice' program. J Fam Pract, 43(4), 361-369. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Meredith, L. S., Yano, E. M., Hickey, S. C., & Sherman, S. E. (2005). Primary care provider attitudes are associated with smoking cessation counseling and referral. Medical Care, 43(9), 929-934. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Messimer, S. R., Hickner, J. M., & Henry, R. C. (1989). A comparison of two antismoking interventions among pregnant women in eleven private primary care practices. J Fam Pract, 28(3), 283-288. Title/abstract excluded: Incorrect population.

Meulepas, M. A., Braspenning, J. C., de Grauw, W. J., Lucas, A. E., Harms, L., Akkermans, R. P., et al. (2007). Logistic support service improves processes and outcomes of diabetes care in general practice.[see comment]. Family Practice, 24, 20-25. Title/abstract excluded: Inappropriate outcomes.

Meulepas, M. A., Jacobs, J. E., Smeenk, F. W. J. M., Smeele, I., Lucas, A. E. M., Bottema, B. J. A. M., et al. (2007). Effect of an integrated primary care model on the management of middle-aged and old patients with obstructive lung diseases. Scandinavian Journal of Primary Health Care, 25(3), 186-192. Title/abstract excluded: Incorrect population.

Meyer, B. A., Meyer, T. J., Howes, N., Ruhlen, S., & Pickett, T. E. (1997). Health assessment for partners of pregnant women: a pilot study of four survey methods. The Journal of the American Board of Family Practice / American Board of Family Practice, 10(3), 192-198. Title/abstract excluded: Inappropriate study design.

166


Meyer, C., Ulbricht, S., Baumeister, S., Schumann, A., Ruge, J., Bischof, G., et al. (2008). Supporting GPS in the provision of smoking cessation intervention - From research to daily routine. Addiction, 103(2), 309. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Meyer, C., Ulbricht, S., Baumeister, S. E., Schumann, A., Ruge, J., Bischof, G., et al. (2008). Proactive interventions for smoking cessation in general medical practice: A quasi-randomized controlled trial to examine the efficacy of computer-tailored letters and physician-delivered brief advice. Addiction, 103(2), 294-304. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Meyer, C., Ulbricht, S., Schumann, A., Hannover, W., Hapke, U., Rumpf, H. J., et al. (2003). Inteventions fostering the motivation to quit for smokers in general practice. Suchtmedizin.in Forschung.und Praxis, 5, 134-136. Title/abstract excluded: Inappropriate study design.

Milch, C. E., Edmunson, J. M., Beshansky, J. R., Griffith, J. L., & Selker, H. P. (2004). Smoking cessation in primary care: A clinical effectiveness trial of two simple interventions. Preventive Medicine, 38(3), 284-294. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Montagna, R. A., & Hupcey, J. E. (2000). Increasing smoking cessation counseling by advanced practice nurses. Clinical excellence for nurse practitioners : the international journal of NPACE, 4(4), 224-230. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Mounsey, A. L., Bovbjerg, V., White, L., & Gazewood, J. (2006). Do students develop better motivational interviewing skills through role-play with standardised patients or with student colleagues? Medical Education, 40(8), 775-780. Title/abstract excluded: Inappropriate study design.

Mudde, A. N., De Vries, H., & Dolders, M. G. T. (1995). Evaluation of a Dutch community-based smoking cessation intervention. Preventive Medicine, 24(1), 61-70. Title/abstract excluded: Inappropriate study design.

Muehlig, S., Hoch, E., Hoefler, M., Pittrow, D., & Wittchen, H. U. (2003). Aims, design and methods of the smoking and nicotine dependence awareness and screening (SNICAS) study. International Journal of Methods in Psychiatric Research, 12(4), 208-228. Title/abstract excluded: Inappropriate study design.

Mullins, R., Livingston, P., & Borland, R. (1999). A strategy for involving general practitioners in smoking control. Australian and New Zealand Journal of Public Health, 23(3), 249-251. Title/abstract: Included. Full paper excluded: Inappropriate study design .

Murray, R. L., Coleman, T., Antoniak, M., Fergus, A., Britton, J., & Lewis, S. A. (2008). The potential to improve ascertainment and intervention to reduce smoking in primary care: a cross sectional survey. BMC Health Serv Res, 8, 6. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Murray, R. L., Coleman, T., Antoniak, M., Stocks, J., Fergus, A., Britton, J., et al. (2008). The effect of proactively identifying smokers and offering smoking cessation support in primary care populations: a cluster-randomized trial. Addiction, 103(6), 998-1006; discussion 1007-1008. Title/abstract: Included. Full paper excluded: Inappropriate comparator.

Nagykaldi, Z., Mold, J. W., & Aspy, C. B. (2005). Practice facilitators: a review of the literature. Family medicine, 37(8), 581-588. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Naidoo, B., Warm, D., Quigley, R., & Taylor, L. (2004). Smoking and public health: a review of reviews of interventions to increase smoking cessation, reduce

smoking initiation and prevent further uptake of smoking (evidence briefing). Retrieved from http://www.nice.org.uk/niceMedia/documents/smoking_evidence_briefing.pdf Title/abstract excluded: Inappropriate study design.

National Institute for Health and Clinical Excellence (NICE) (2006). Brief interventions and referral for smoking cessation in primary care and other settings. London (UK):Public health intervention; no. 1, 36p. Retrieved from http://www.guideline.gov/summary/summary.aspx?doc_id=9740&nbr=005214&string=smoking Title/abstract excluded: Inappropriate study design.

National Institute for Health and Clinical Excellence (NICE) (2008). Smoking cessation services in primary care, pharmacies, local authorities and workplaces, particularly for manual working groups, pregnant women and hard to reach communities. London (UK): Public health guidance; no. 10, 87p. Retrieved from http://www.guideline.gov/summary/summary.aspx?doc_id=12286&nbr=006361&string=smoking Title/abstract excluded: Inappropriate study design.

Nebot, A. M., Soler, V. M., Martín, C. C., Birulés, P. M., Oller, C. M., Sala, C. E., et al. (1989). [Effectiveness of the physician's advice to quit smoking: evaluation of the impact a year after the fact]. Revista clínica.española., 184, 201-205. Title/abstract excluded: Incorrect population.

Nguyen, B. H., Nguyen, K. P., Mcphee, S. J., Nguyen, T., Tran, D. Q., & Jenkins, C. N. H. (2000). Promoting cancer prevention activities among Vietnamese physicians in California. Journal of Cancer Education, 15(2), 82-85. Title/abstract excluded: Inadequate sample size.

Nieman, L. Z., Velasquez, M. M., Groff, J. Y., Cheng, L., & Foxhall, L. E. (2005). Implementation of a smoking cessation counseling module in a preceptorship program. Family Medicine, 37(2), 105-111. Title/abstract: Included. Full paper excluded: Incorrect population.

O'Brien, M. A., Rogers, S., Jamtvedt, G., Oxman, A. D., Odgaard-Jensen, J., Kristoffersen, D. T., et al. (2007). Educational outreach visits: effects on professional practice and health care outcomes. Cochrane database of systematic reviews (Online)(4), CD000409. Title/abstract: Included. Full paper excluded: Other.

Ockene, J. K. (1999). Primary care-based smoking interventions. Nicotine & tobacco research : official journal of the Society for Research on Nicotine and Tobacco, 1 Suppl 2(-), S189-S193. Title/abstract excluded: Inappropriate study design.

Ockene, J. K., Hosmer, D. W., Williams, J. W., Goldberg, R. J., Ockene, I. S., Biliouris, T., et al. (1987). The relationship of patient characteristics to physician delivery of advice to stop smoking. Journal of general internal medicine : official journal of the Society for Research and Education in Primary Care Internal Medicine, 2(5), 337-340. Title/abstract excluded: Inappropriate study design.

Ockene, J. K., Kristeller, J., Goldberg, R., Amick, T. L., Pekow, P. S., Hosmer, D., et al. (1991). Increasing the efficacy of physician-delivered smoking interventions: a randomized clinical trial. Journal of general internal medicine : official journal of the Society for Research and Education in Primary Care Internal Medicine, 6, 1-8. Title/abstract excluded: Incorrect population.

Ockene, J. K., Kristeller, J., Pbert, L., Hebert, J. R., Luippold, R., Goldberg, R. J., et al. (1994). The physician-delivered smoking intervention project: can short-term interventions produce long-term effects for a general outpatient population? Health Psychology, 13(3), 278-281. Title/abstract excluded: Incorrect population.

Ockene, J. K., Lindsay, E., Berger, L., & Hymowitz, N. (1990). Health care providers as key change agents in the

167


Community Intervention Trial for Smoking Cessation (COMMIT). International Quarterly of Community Health Education. Special Issue: Community Intervention Trial for Smoking Cessation: COMMIT, 11(3), 223-237. Title/abstract excluded: Inappropriate study design.

Ockene, J. K., Quirk, M. E., Goldberg, R. J., Kristeller, J. L., Donnelly, G., Kalan, K. L., et al. (1988). A residents' training program for the development of smoking intervention skills. Archives of Internal Medicine, 148(5), 1039-1045. Title/abstract excluded: Inappropriate outcomes.

Okuyemi, K. S., Nollen, N. L., & Ahluwalia, J. S. (2006). Interventions to facilitate smoking cessation. American Family Physician, 74(2), 262-271+276. Title/abstract excluded: Inappropriate study design.

O'Loughlin, J., Makni, H., Tremblay, M., Lacroix, C., Gervais, A., Dery, V., et al. (2001). Smoking cessation counseling practices of general practitioners in Montreal. Preventive Medicine, 33(6), 627-638. Title/abstract excluded: Inappropriate study design.

Ornstein, S. M., Garr, D. R., Jenkins, R. G., Musham, C., Hamadeh, G., & Lancaster, C. (1995). Implementation and evaluation of a computer-based preventive services system. Family medicine, 27(4), 260-266. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Oxcheck Study Group (1994). Effectiveness of health checks conducted by nurses in primary care: results of the OXCHECK study after one year. Imperial Cancer Research Fund OXCHECK Study Group. Bmj, 308(6924), 308-312. Title/abstract excluded: Incorrect intervention.

Oxman, A. D., Thomson, M. A., Davis, D. A., & Haynes, R. B. (1995). No magic bullets: a systematic review of 102 trials of interventions to improve professional practice. Cmaj, 153(10), 1423-1431. Title/abstract: Included. Full paper excluded: Other.

Page, A. R., Walters, D. J., Schlegel, R. P., & Best, J. A. (1986). Smoking cessation in family practice: the effects of advice and nicotine chewing gum prescription. Addictive Behaviors, 11, 443-446. Title/abstract excluded: Incorrect population.

Park, E., Eaton, C. A., Goldstein, M. G., DePue, J., Niaura, R., Guadagnoli, E., et al. (2001). The development of a decisional balance measure of physician smoking cessation interventions. Preventive Medicine: An International Journal Devoted to Practice and Theory, 33(4), 261-267. Title/abstract excluded: Inappropriate study design.

Park, E. R., DePue, J. D., Goldstein, M. G., Niaura, R., Harlow, L. L., Willey, C., et al. (2003). Assessing the transtheoretical model of change constructs for physicians counseling smokers. Annals of Behavioral Medicine, 25(2), 120-126. Title/abstract excluded: Inappropriate study design.

Park, E. R., Kleimann, S., Pelan, J. A., & Shields, A. E. (2007). Anticipating clinical integration of genetically tailored tobacco dependence treatment: Perspectives of primary care physicians. Nicotine & Tobacco Research, 9(2), 271-279. Title/abstract excluded: Inappropriate study design.

Park, E. R., MacDonald Gross, N. A., Goldstein, M. G., DePue, J. D., Hecht, J. P., Eaton, C. A., et al. (2002). Physician recruitment for a community-based smoking cessation intervention. Journal of Family Practice, 51(1), 70. Title/abstract excluded: Inappropriate study design.

Parker, D. R., Evangelou, E., & Eaton, C. B. (2005). Intraclass correlation coefficients for cluster randomized trials in primary care: The cholesterol education and research trial (CEART). Contemporary Clinical Trials, 26(2), 260-267. Title/abstract excluded: Inappropriate study design.

Parrott, S., Godfrey, C., Raw, M., West, R., & McNeill, A. (1998). Guidance for commissioners on the cost effectiveness of smoking cessation interventions. Thorax, 53(SUPPL. 5), S1A-S38A. Title/abstract excluded: Inappropriate study design.

Patel, D. R., & Greydanus, D. E. (2000). Office interventions for adolescent smokers. Adolescent medicine (Philadelphia, Pa.), 11(3), 577-588. Title/abstract excluded: Inappropriate study design.

Pbert, L., Adams, A., Quirk, M., Hebert, J. R., Ockene, J. K., & Luippold, R. S. (1999). The patient exit interview as an assessment of physician-delivered smoking intervention: A validation study. Health Psychology, 18(2), 183-188. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Pbert, L., Moolchan, E. T., Muramoto, M., Winickoff, J. P., Curry, S., Lando, H., et al. (2003). The state of office-based interventions for youth tobacco use. Pediatrics, 111(6 Pt 1), e650-e660. Title/abstract excluded: Inappropriate study design.

Perez-Stable, E. J., Juarez-Reyes, M., Kaplan, C. P., Fuentes-Afflick, E., Gildengorin, V., & Millstein, S. G. (2001). Counseling smoking parents of young children: Comparison of pediatricians and family physicians. Archives of Pediatrics and Adolescent Medicine, 155(1), 25-31. Title/abstract excluded: Inappropriate study design.

Perry, B. L. (2000). Smoking cessation: Successful intervention. Primary Care Update for Ob/Gyns, 8(1), 36-39. Title/abstract excluded: Inappropriate study design.

Persson, L. G., & Hjalmarson, A. (2006). Smoking cessation in patients with diabetes mellitus: Results from a controlled study of an intervention programme in primary healthcare in Sweden. Scandinavian Journal of Primary Health Care, 24(2), 75-80. Title/abstract: Included. Full paper excluded: Incorrect population.

Pieterse, M. E., Boekema, E. R., Seydel, E. R., & Wiegman, O. (1992). Smoking cessation via the general practitioner: effects of a minimal contact intervention programme [abstract]. Patient Education & Counseling, 20, 107. Title/abstract excluded: Incorrect population.

Pilnick, A., & Coleman, T. (2006). Death, depression and 'defensive expansion': closing down smoking as an issue for discussion in GP consultations. Soc Sci Med, 62(10), 2500-2512. Title/abstract excluded: Inappropriate study design.

Pine, D., Sullivan, S., Conn, S. A., & David, C. (1999). Promoting tobacco cessation in primary care practice. Primary Care - Clinics in Office Practice, 26(3), 591-610. Title/abstract excluded: Inappropriate study design.

Pine, D., Sullivan, S., Sauser, M., & David, C. (1997). Effects of a systematic approach to tobacco cessation in a community-based practice. Archives of Family Medicine, 6(4), 363-367. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Pinget, C., Martin, E., Wasserfallen, J. B., Humair, J. P., & Cornuz, J. (2007). Cost-effectiveness analysis of a European primary-care physician training in smoking cessation counseling. European Journal of Cardiovascular Prevention and Rehabilitation, 14(3), 451-455. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Pizzo, A. M., Chellini, E., Grazzini, G., Cardone, A., & Badellino, F. (2003). Italian general practitioners and smoking cessation strategies. Tumori, 89(3), 250-254. Title/abstract excluded: Inappropriate study design.

Pletsch, P. K., Devine, E. C., Payne, J., & London, R. (2001). Adoption of a Smoking Cessation Guideline in Primary Care: A Case Study. Clinical Excellence for

168


Nurse Practitioners, 5(5), 31-38. Title/abstract excluded: Inappropriate study design.

Pohl, J. M. (2000). Smoking cessation and low-income women: theory, research, and interventions. Nurse practitioner forum, 11(2), 101-108. Title/abstract excluded: Inappropriate study design.

Pohl, J. M., & Caplan, D. (1998). Smoking cessation: using group intervention methods to treat low-income women. Nurse Practitioner, 23(12), 13, 17-18, 20. Title/abstract excluded: Inappropriate study design.

Pollack, K. I., Arredondo, E. M., Yarnall, K. S. H., Lipkus, I., Myers, E., McNeilly, M., et al. (2001). How do residents prioritize smoking cessation for young "high-risk" women? Factors associated with addressing smoking cessation. Preventive Medicine: An International Journal Devoted to Practice and Theory, 33(4), 292-299. Title/abstract excluded: Inappropriate study design.

Pollak, K. I., McBride, C. M., Scholes, D., Grothaus, L. C., Civic, D., & Curry, S. J. (2002). Women's reports of smoking cessation advice during reproductive health visits and subsequent smoking cessation. American Journal of Managed Care, 8(10), 837-844. Title/abstract excluded: Inappropriate study design.

Potter, B., & Fleming, M. F. (2003). Obstetrics and gynecology resident education in tobacco, alcohol, and drug use disorders. Obstetrics and Gynecology Clinics of North America, 30(3), 583-599. Title/abstract excluded: Inappropriate study design.

Prochaska, J. J., Teherani, A., & Hauer, K. E. (2007). Medical students' use of the stages of change model in tobacco cessation counseling. Journal of General Internal Medicine, 22(2), 223-227. Title/abstract excluded: Inappropriate study design.

Prochazka, A., Koziol-McLain, J., Tomlinson, D., & Lowenstein, S. R. (1995). Smoking cessation counseling by emergency physicians: opinions, knowledge, and training needs. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine, 2(3), 211-216. Title/abstract excluded: Inappropriate study design.

Proude, E. M., Conigrave, K. M., Britton, A., & Haber, P. S. (2008). Improving alcohol and tobacco history taking by junior medical officers. Alcohol and Alcoholism, 43(3), 320-325. Title/abstract: Included. Full paper excluded: Incorrect population.

Quinn, V. P., Stevens, V. J., Hollis, J. F., Rigotti, N. A., Solberg, L. I., Gordon, N., et al. (2005). Tobacco-cessation services and patient satisfaction in nine nonprofit HMOs. American Journal of Preventive Medicine, 29(2), 77-84. Title/abstract excluded: Inappropriate study design.

Quirk, M., Ockene, J., Kristeller, J., Goldberg, R., Donnelly, G., Amick, T., et al. (1991). Training family practice and internal medicine residents to counsel patients who smoke: Improvement and retention of counseling skills. Family Medicine, 23(2), 108-111. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Ramos, M. M. (2005). Effectiveness of advanced group and individual interventions in tackling tobacco dependency in primary care. Atencion Primaria, 36, 462-465. Title/abstract excluded: Incorrect population.

Raw, M., McNeill, A., & Coleman, T. (2005). Lessons from the English smoking treatment services. Addiction, 100 Suppl 2, 84-91. Title/abstract excluded: Inappropriate study design.

Raw, M., McNeill, A., & West, R. (1998). Smoking cessation guidelines for health professionals - A guide to effective smoking cessation interventions for the

health care system. Thorax, 53(SUPPL. 5), S1-S19. Title/abstract excluded: Inappropriate study design.

Registered Nurses Association of Ontario (RNAO) (2002). Toolkit: implementation of clinical practice guidelines. from Toronto (ON): Registered Nurses Association of Ontario (RNAO): http://www.rnao.org/Page.asp?PageID=924&ContentID=823 Title/abstract excluded: Inappropriate study design.

Registered Nurses Association of Ontario (RNAO) (2007). Integrating smoking cessation into daily nursing practice. from Toronto (ON): Registered Nurses Association of Ontario (RNAO): http://www.guideline.gov/summary/summary.aspx?doc_id=11503&nbr=005957&string=smoking Title/abstract excluded: Inappropriate study design.

Rennard, S. I. (1999). Smoking cessation interventions: Making them work in your practice. Consultant, 39(12), 3384-3392. Title/abstract excluded: Inappropriate study design.

Rennard, S. I. (2008). Update on smoking cessation interventions for the primary care physician. Introduction. The American journal of medicine, 121(4 Suppl 1), S1-S2. Title/abstract excluded: Inappropriate study design.

Rice, V. H. (2006). Nursing intervention and smoking cessation: Meta-analysis update. Heart and Lung: Journal of Acute and Critical Care, 35(3), 147-163. Title/abstract excluded: Incorrect population.

Rice, V. H., & Stead, L. F. (2004). Nursing interventions for smoking cessation. Cochrane database of systematic reviews (Online : Update Software), -(1), CD001188. Title/abstract excluded: Inappropriate study design.

Rice, V. H., & Stead, L. F. (2008). Nursing interventions for smoking cessation. Cochrane database of systematic reviews (Online), -(1), CD001188. Title/abstract excluded: Inappropriate study design.

Richmond, R., & Heather, N. (1990). General practitioner interventions for smoking cessation: Past results and future prospects. Behaviour Change, 7(3), 110-119. Title/abstract excluded: Inappropriate study design.

Richmond, R., & Webster, I. (1985). Evaluation of general practitioners' use of a smoking intervention programme. International Journal of Epidemiology, 14(3), 396-401. Title/abstract excluded: Inappropriate study design.

Richmond, R. L. (1996). Retracing the steps of Marco Polo: From clinical trials to diffusion of interventions for smokers. Addictive Behaviors, 21(6), 683-697. Title/abstract excluded: Inappropriate study design.

Richmond, R. L., & Anderson, P. (1994). Research in general practice for smokers and excessive drinkers in australia and the UK. I. Interpretation of results. Addiction, 89(1), 35-40. Title/abstract excluded: Inappropriate study design.

Richmond, R. L., & Anderson, P. (1994). Research in general practice for smokers and excessive drinkers in Australia and the UK. III. Dissemination of interventions. Addiction, 89(1), 49-62. Title/abstract excluded: Inappropriate study design.

Richmond, R. L., Austin, A., & Webster, I. W. (1986). Three year evaluation of a programme by general practitioners to help patients to stop smoking. British Medical Journal, 292, 803-806. Title/abstract excluded: Incorrect population.

Richmond, R. L., Kehoe, L. A., & Webster, I. W. (1993). Multivariate models for predicting abstention following intervention to stop smoking by general practitioners. Addiction, 88(8), 1127-1135. Title/abstract excluded: Incorrect population.

Richmond, R. L., Makinson, R. J., Kehoe, L. A., Giugni, A. A., & Webster, I. W. (1993). One-year evaluation of three smoking cessation interventions administered by

169


general practitioners. Addictive Behaviors, 18(2), 187-199. Title/abstract excluded: Incorrect population.

Richmond, R. L., & Webster, I. W. (1985). A smoking cessation programme for use in general practice. The Medical journal of Australia, 142, 190-194. Title/abstract excluded: Incorrect population.

Riemsma, R. P., Pattenden, J., Bridle, C., Sowden, A. J., Mather, L., Watt, I. S., et al. (2002). A systematic review of the effectiveness of interventions based on a stages-of-change approach to promote individual behaviour change. Health Technol Assess, 6(24), 1-231. Title/abstract excluded: Inappropriate outcomes.

Rigotti, N. A. (2003). Commentary: Putting the research into practice. BMJ: British Medical Journal, 327(7428), 1395. Title/abstract excluded: Inappropriate study design.

Ritvo, P. G., Irvine, M. J., Lindsay, E. A., Kraetschmer, N., Blair, N., & Shnek, Z. M. (1997). A critical review of research related to family physician-assisted smoking cessation interventions. Cancer Prevention and Control, 1(4), 289-303. Title/abstract excluded: Incorrect population.

Robertson, J. O., Lincoff, A. M., Wolski, K., & Topol, E. J. (2006). Planned Versus Provisional Use of Glycoprotein IIb/IIIa Inhibitors in Smokers Undergoing Percutaneous Coronary Intervention. American Journal of Cardiology, 97(12), 1679-1684. Title/abstract excluded: Incorrect population.

Robson, J., Boomla, K., Fitzpatrick, S., Jewell, A. J., Taylor, J., Self, J., et al. (1989). Using nurses for preventive activities with computer assisted follow up: a randomised controlled trial. BMJ, 298(6671), 433-436. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Roche, A. M., & Freeman, T. (2004). Brief interventions: Good in theory but weak in practice. Drug and Alcohol Review, 23(1), 11-18. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Rohrbaugh, M. J., Shoham, V., Trost, S., Muramoto, M., Cate, R. M., & Leischow, S. (2001). Couple dynamics of change-resistant smoking: toward a family consultation model. Family process, 40(1), 15-31. Title/abstract excluded: Inappropriate study design.

Rollnick, S., Butler, C. C., & Stott, N. (1997). Helping smokers make decisions: The enhancement of brief intervention for general medical practice. Patient Education and Counseling, 31(3), 191-203. Title/abstract excluded: Inappropriate study design.

Rosal, M. C., Ockene, J. K., Hurley, T. G., Kalan, K., & Hebert, J. R. (1998). Effectiveness of nicotine-containing gum in the physician-delivered smoking intervention study. Preventive Medicine, 27(2), 262-267. Title/abstract excluded: Incorrect population.

Rosen, I. M., & Maurer, D. M. (2008). Reducing tobacco use in adolescents. American Family Physician, 77(4), 483-492. Title/abstract excluded: Inappropriate study design.

Roski, J. (1998). Changing practice patterns as a result of implementing the Agency for Health Care Policy and Research guideline in 20 primary care clinics. Tobacco Control, 7 Suppl, S19-20; discussion S24-15. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Roski, J., Jeddeloh, R., Roemhild, H., Werb, P., Koele, R., & Sherwood, N. (1998). Implementing smoking cessation practice guidelines in primary care [abstract]. Abstract.Book./Association for Health Services Research, 15, 206. Title/abstract excluded: Full-text not available from any source.

Rowan, M. S., Coambs, R. B., Jensen, P., Balderston, M., MacKenzie, D., & Kothari, A. (1998). Smoking cessation, physicians, and medical office staff.

Clinical tobacco intervention in Prince Edward Island. Canadian Family Physician, 44(-), 2433-2440. Title/abstract excluded: Inappropriate study design.

Russell, M. A., & Jarvis, M. J. (1985). Theoretical background and clinical use of nicotine chewing gum. NIDA research monograph, 53(-), 110-130. Title/abstract excluded: Inappropriate study design.

Russell, M. A., Merriman, R., Stapleton, J., & Taylor, W. (1983). Effect of nicotine chewing gum as an adjunct to general practitioner's advice against smoking. British Medical Journal, 287, 1782-1785. Title/abstract excluded: Incorrect population.

Russell, M. A. H., Stapleton, J. A., Feyerabend, C., Wiseman, S. M., Gustavsson, G., Sawe, U., et al. (1993). Targeting heavy smokers in general practice: Randomised controlled trial of transdermal nicotine patches. British Medical Journal, 306(6888), 1308-1312. Title/abstract excluded: Incorrect population.

Russell, M. A. H., Stapleton, J. A., Hajek, P., Jackson, P. H., & Belcher, M. (1988). District programme to reduce smoking: Can sustained intervention by general practitioners affect prevalence? Journal of Epidemiology and Community Health, 42(2), 111-115. Title/abstract excluded: Inappropriate study design.

Russell, M. A. H., Stapleton, J. A., Jackson, P. H., Hajek, P., & Belcher, M. (1987). District programme to reduce smoking: Effect of clinic supported brief intervention by general practitioners. British Medical Journal, 295(6608), 1240-1244. Title/abstract excluded: Incorrect population.

Ryder, D. (1988). Brief intervention by general practitioners against smoking. British medical journal (Clinical research ed.), 296(6615), 132-133. Title/abstract excluded: Inappropriate study design.

Sachs, D. P. (1995). Effectiveness of the 4-mg dose of nicotine polacrilex for the initial treatment of high-dependent smokers. Archives of Internal Medicine, 155, 1973-1980. Title/abstract excluded: Incorrect population.

Sachs, D. P., Säwe, U., & Leischow, S. J. (1993). Effectiveness of a 16-hour transdermal nicotine patch in a medical practice setting, without intensive group counseling. Archives of Internal Medicine, 153, 1881-1890. Title/abstract excluded: Incorrect population.

Sahu, A., Edwards, R., Harrison, R. A., Thornton, J., & Kelly, S. P. (2008). Attitudes and behaviour of ophthalmologists to smoking cessation. Eye, 22(2), 246-250. Title/abstract excluded: Inappropriate study design.

Salkeld, G., Phongsavan, P., Oldenburg, B., Johannesson, M., Convery, P., Graham, C. P., et al. (1997). The cost-effectiveness of a cardiovascular risk reduction program in general practice. Health Policy, 41, 105-119. Title/abstract excluded: Inappropriate study design.

Sanchez, R. J., & Khalil, L. (2005). Badger Heart Program: Health screenings targeted to increase cardiovascular awareness in women at four northern sites in Wisconsin. Wisconsin Medical Journal, 104(6), 24-29. Title/abstract excluded: Inappropriate study design.

Sanders, D., Fowler, G., Mant, D., Fuller, A., Jones, L., & Marzillier, J. (1989). Randomized controlled trial of anti-smoking advice by nurses in general practice. Journal of the Royal College of General Practitioners, 39(324), 273-276. Title/abstract excluded: Incorrect population.

Sanders, D., Peveler, R., Mant, D., & Fowler, G. (1993). Predictors of successful smoking cessation following advice from nurses in general practice. Addiction, 88, 1699-1705. Title/abstract excluded: Inappropriate study design.

170


Sanz, P. B., de Miguel, D. J., Anegon, B. M., García, C. M., Gómez, S. E., & Fernández-Domínguez, J. F. (2006). [Effectiveness of a programme of intensive tobacco counselling by nursing professionals]. Atencion primaria / Sociedad Española.de Medicina de Familia.y Comunitaria., 37, 266-272. Title/abstract excluded: Incorrect population.

Sanz, P. B., de Miguel, D. J., Aragón, B. M., González-González, A. I., Cortes, C. M., & Vázquez, I. (2003). [Effectiveness of non-pharmacological primary care methods for giving up tobacco dependency]. Atencion primaria / Sociedad Española.de Medicina de Familia.y Comunitaria., 32, 366-370. Title/abstract excluded: Incorrect population.

Saywell, R. M., Jay, S. J., Lukas, P. J., Casebeer, L. L., Mybeck, K. C., Parchman, M. L., et al. (1996). Indiana family physician attitudes and practices concerning smoking cessation. Indiana medicine : the journal of the Indiana State Medical Association, 89(2), 149-156. Title/abstract excluded: Inappropriate study design.

Schnoll, R. A., Rukstalis, M., Wileyto, E. P., & Shields, A. E. (2006). Smoking Cessation Treatment by Primary Care Physicians An Update and Call for Training. American Journal of Preventive Medicine, 31(3), 233-239. Title/abstract: Included. Full paper excluded: Inappropriate study design.

Schroeder, K., Lawlor, D. A., Montaner, D., & Ebrahim, S. (2006). Self-reported smoking cessation interventions were not associated with quitting in older women. Journal of Clinical Epidemiology, 59(6), 622-622. Title/abstract excluded: Inappropriate study design.

Secker-Walker, R. H., Solomon, L. J., Flynn, B. S., & Dana, G. S. (1994). Comparisons of the smoking cessation counseling activities of six types of health professionals. Preventive Medicine: An International Journal Devoted to Practice and Theory, 23(6), 800-808. Title/abstract excluded: Inappropriate study design.

Segnan, N., Ponti, A., Battista, R. N., Senore, C., Rosso, S., Shapiro, S. H., et al. (1991). A randomized trial of smoking cessation interventions in general practice in Italy. Cancer causes & control : CCC, 2(4), 239-246. Title/abstract excluded: Incorrect population.

Sesney, J. W., Kreher, N. E., Hickner, J. M., & Webb, S. (1997). Smoking cessation interventions in rural family practices: an UPRNet study. J Fam Pract, 44(6), 578-585. Title/abstract excluded: Inappropriate study design.

Shakeshaft, A., Nassirimanesh, B., Day, C., & Dolan, K. A. (2005). Perceptions of substance use, treatment options and training needs among Iranian primary care physicians. International Journal for Equity in Health, 4(-). Title/abstract excluded: Inappropriate study design.

Shearer, J., & Shanahan, M. (2006). Cost effectiveness analysis of smoking cessation interventions. Australian and New Zealand Journal of Public Health, 30(5), 428-434. Title/abstract excluded: Inappropriate study design.

Sherman, S. E., Fu, S. S., Joseph, A. M., Lanto, A. B., & Yano, E. M. (2005). Gender differences in smoking cessation services received among veterans. Women's Health Issues, 15(3), 126-133. Title/abstract excluded: Incorrect intervention.

Silagy, C. (2000). Physician advice for smoking cessation. Cochrane database of systematic reviews (Online : Update Software), -(2), CD000165. Title/abstract excluded: Incorrect population.

Silagy, C., Lancaster, T., Gray, S., & Fowler, G. (1994). Effectiveness of training health professionals to provide smoking cessation interventions: systematic review of randomised controlled trials. Quality and

Safety in Health Care, 3(4), 193-198. Title/abstract: Included. Full paper excluded: Other.

Sinclair, H. K., Bond, C. M., Lennox, A. S., Silcock, J., Winfield, A. J., & Donnan, P. T. (1998). Training pharmacists and pharmacy assistants in the stage-of-change model of smoking cessation: a randomised controlled trial in Scotland. Tob Control, 7(3), 253-261. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Sippel, J. M., Osborne, M. L., Bjornson, W., Goldberg, B., & Buist, A. S. (1999). Smoking cessation in primary care clinics. Journal of general internal medicine : official journal of the Society for Research and Education in Primary Care Internal Medicine, 14, 670-676. Title/abstract excluded: Incorrect population.

Slama, K., Karsenty, S., & Hirsch, A. (1995). Effectiveness of minimal intervention by general practitioners with their smoking patients: a randomised, controlled trial in France. Tobacco Control, 4, 162-169. Title/abstract excluded: Incorrect population.

Slama, K., Redman, S., Perkins, J., Reid, A. L., & Sanson-Fisher, R. W. (1990). The effectiveness of two smoking cessation programmes for use in general practice: a randomised clinical trial. BMJ, 300, 1707-1709. Title/abstract excluded: Incorrect population.

Smith, M. Y., Cromwell, J., DePue, J., Spring, B., Redd, W., & Unrod, M. (2007). Determining the cost-effectiveness of a computer-based smoking cessation intervention in primary care. Managed care (Langhorne, Pa.), 16(7), 48-55. Title/abstract excluded: Inappropriate study design.

Smith, S. M., Bury, G., O'Leary, M., Shannon, W., Tynan, A., Staines, A., et al. (2004). The North Dublin randomized controlled trial of structural diabetes shared care. Family Practice, 21(1), 39-45. Title/abstract excluded: Inappropriate study design.

Solberg, L. I., Kottke, T. E., Brekke, M. L., Calomeni, C. A., Conn, S. A., & Davidson, G. (1996). Using Continuous Quality Improvement to Increase Preventive Services in Clinical Practice--Going Beyond Guidelines. Preventive Medicine, 25(3), 259-267. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Solberg, L. I., Maxwell, P. L., Kottke, T. E., Gepner, G. J., & Brekke, M. L. (1990). A systematic primary care office-based smoking cessation program. J Fam Pract, 30(6), 647-654. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Soto Mas, F. G., Papenfuss, R. L., Jacobson, H. E., Hsu, C. E., Urrutia-Rojas, X., & Kane, W. M. (2005). Hispanic physicians' tobacco intervention practices: a cross-sectional survey study. BMC public health [electronic resource]. 5(-), 120. Title/abstract excluded: Inappropriate study design.

Soumerai, S. B., McLaughlin, T. J., & Avorn, J. (1989). Improving drug prescribing in primary care: a critical analysis of the experimental literature. The Milbank quarterly, 67(2), 268-317. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Spencer, E., Swanson, T., Hueston, W. J., & Edberg, D. L. (1999). Tools to improve documentation of smoking status: continuous quality improvement and electronic medical records. Archives of Family Medicine 8(1), 18-22. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Spencer, L. (2005). An academic detailing intervention to disseminate physician-delivered smoking cessation counseling: Smoking cessation outcomes of the physicians counseling smokers project - Commentary. American Journal of Health Promotion, 19(5), 388-389. Title/abstract: Included. Full paper excluded: Inappropriate study design .

171


Spitz, M. R., Chamberlain, R. M., Sider, J. G., & Fueger, J. J. (1992). Cancer prevention practices among Texas primary care physicians. Journal of Cancer Education, 7(1), 55-60. Title/abstract excluded: Inappropriate study design.

Stacey, F., Heasman, P. A., Heasman, L., Hepburn, S., McCracken, G. I., & Preshaw, P. M. (2006). Smoking cessation as a dental intervention - Views of the profession. British Dental Journal, 201(2), 109-113. Title/abstract excluded: Inappropriate study design.

Stack, N. M., & Zillich, A. J. (2007). Implementation of inpatient and outpatient tobacco-cessation programs. American Journal of Health-System Pharmacy, 64(19), 2074-2079. Title/abstract excluded: Incorrect population.

Stead, L. F., Bergson, G., & Lancaster, T. (2008). Physician advice for smoking cessation. Cochrane Database of Systematic Reviews, -(2). Title/abstract excluded: Inappropriate study design.

Steinberg, M. B., Nanavati, K., Delnevo, C. D., & Abatemarco, D. J. (2007). Predictors of self-reported discussion of cessation medications by physicians in New Jersey. Addictive Behaviors, 32(12), 3045-3053. Title/abstract excluded: Inappropriate study design.

Steiro, A., Smedslund, G., & Thune Hammerstrøm, K. (2007). Smoking cessation in different socioeconomic groups. Norwegian Knowledge Centre for the Health Services (NOKC). Retrieved from http://www.inahta.org/Publications/Briefs-Checklist-Impact/20072/3677/ Title/abstract excluded: Inappropriate outcomes.

Steptoe, A., Doherty, S., Rink, E., Kerry, S., Kendrick, T., & Hilton, S. (1999). Behavioural counselling in general practice for the promotion of healthy behaviour among adults at increased risk of coronary heart disease: Randomised trial. British Medical Journal, 319(7215), 943-947. Title/abstract excluded: Incorrect population.

Stevens, M. M., Olson, A. L., Gaffney, C. A., Tosteson, T. D., Mott, L. A., & Starr, P. (2002). A pediatric, practice-based, randomized trial of drinking and smoking prevention and bicycle helmet, gun, and seatbelt safety promotion. Pediatrics, 109, 490-497. Title/abstract excluded: Incorrect population.

Stevens, V. J., Severson, H., Lichtenstein, E., Little, S. J., & Leben, J. (1995). Making the most of a teachable moment: a smokeless-tobacco cessation intervention in the dental office. Am J Public Health, 85(2), 231-235. Title/abstract excluded: Incorrect population.

Strayer, S. M., Rollins, L. K., & Martindale, J. R. (2006). A handheld computer smoking intervention tool and its effects on physician smoking cessation counseling. Journal of the American Board of Family Medicine, 19(4), 350-357. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Su, H., Gordon, M., Roake, J., & Lewis, D. (2006). Management of risk factors: a survey of New Zealand vascular surgeons. The New Zealand medical journal., 119(1231), U1905. Title/abstract excluded: Inappropriate study design.

Sullivan, P., & Kothari, A. (1997). Right to bill may affect amount of tobacco counselling by MDs. Canadian Medical Association Journal, 156(2), 241-243. Title/abstract excluded: Inappropriate study design.

Thomas, K., Yaphe, J., & Matalon, A. (2007). Current primary care physician interventions to promote smoking cessation in Israel: An observational study. Israel Medical Association Journal, 9(9), 645-648. Title/abstract excluded: Inappropriate study design.

Thompson, R. S., Michnich, M. E., Friedlander, L., Gilson, B., Grothaus, L. C., & Storer, B. (1988). Effectiveness of smoking cessation interventions integrated into

primary care practice. Medical Care, 26(1), 62-76. Title/abstract excluded: Incorrect population.

Thorndike, A. N., Rigotti, N. A., Stafford, R. S., & Singer, D. E. (1998). National patterns in the treatment of smokers by physicians. Journal of the American Medical Association, 279(8), 604-608. Title/abstract excluded: Inappropriate study design.

Tomar, S. L. (2001). Dentistry's role in tobacco control. The Journal of the American Dental Association, 132 Suppl(-), 30S-35S. Title/abstract excluded: Inappropriate study design.

Tsevat, J. (1992). Impact and cost-effectiveness of smoking interventions. Am J Med, 93(1A), 43S-47S. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Tyc, V. L., Hudson, M. M., Hinds, P., Elliott, V., & Kibby, M. Y. (1997). Tobacco use among pediatric cancer patients: Recommendations for developing clinical smoking interventions. Journal of Clinical Oncology, 15(6), 2194-2204. Title/abstract excluded: Inappropriate study design.

U.S. Preventive Services Task Force (USPSTF) (2003). Counseling to prevent tobacco use and tobacco-caused disease: recommendation statement. from Rockville (MD): Agency for Healthcare Research and Quality (AHRQ): http://www.guideline.gov/summary/summary.aspx?doc_id=4314&nbr=003268&string=smoking Title/abstract excluded: Inappropriate study design.

University of Michigan Health System (2006). Smoking cessation. from Ann Arbor (MI): University of Michigan Health System: http://www.guideline.gov/summary/summary.aspx?doc_id=9800&nbr=005247&string=smoking Title/abstract excluded: Inappropriate study design.

Urso, P. (2003). Match the best smoking cessation intervention to your patient. The Nurse practitioner, Suppl(-), 12-16, 18. Title/abstract excluded: Inappropriate study design.

Van den Bruel, A., Cleemput, I., Van Linden, A., Schoefs, D., Ramaekers, D., & Bonneux, L. (2004). Effectiveness and cost-effectiveness of treatments for smoking cessation. Belgian Health Care Knowledge Centre (KCE). KCE Reports Vol. 1A Ref. PF04-26.02A. Retrieved from http://www.crd.york.ac.uk/crdweb/ShowRecord.asp?View=Full&ID=32005000669 Title/abstract excluded: Inappropriate outcomes.

Van Schayck, O. C. P., Pinnock, H., Ostrem, A., Litt, J., Tomlins, R., Williams, S., et al. (2008). IPCRG Consensus statement: Tackling the smoking epidemic - Practical guidance for primary care. Primary Care Respiratory Journal, 17(3), 185-193. Title/abstract excluded: Inappropriate study design.

Varona, P., Bonet, M., Fernandez, N., Canizares, M., Garcia Roche, R., Ibarra, A. M., et al. (2005). Use of medical counseling for the prevention and control of smoking in the municipality of old Havana. Journal of Urban Health, 82(1), 71-75. Title/abstract excluded: Inappropriate study design.

Veterans Administration, D. o. D. (2004). VA/DoD clinical practice guideline for the management of tobacco use. from Washington (DC): Department of Veteran Affairs: http://www.guideline.gov/summary/summary.aspx?doc_id=6107&nbr=003962&string=tobacco Title/abstract excluded: Inappropriate study design.

Vetter, N. J., & Ford, D. (1990). Smoking prevention among people aged 60 and over: a randomized controlled trial. Age and ageing, 19, 164-168. Title/abstract excluded: Incorrect population.

172


Vogt, F., Hall, S., & Marteau, T. M. (2005). General practitioners' and family physicians' negative beliefs and attitudes towards discussing smoking cessation with patients: A systematic review. Addiction, 100(10), 1423-1431. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Vogt, F., Hall, S., & Marteau, T. M. (2007). General practitioners' beliefs about effectiveness and intentions to recommend smoking cessation services: Qualitative and quantitative studies. BMC Family Practice, 8(39). Title/abstract excluded: Inappropriate study design.

Volmink, J., Laubscher, J., & Furman, S. (1996). The SASPREN primary care survey - Who consults the family doctor? South African Medical Journal, 86(3), 241-245. Title/abstract excluded: Inappropriate study design.

Waage, H. P., Vatten, L. J., Opedal, E., & Hilt, B. (1997). Smoking intervention in subjects at risk of asbestos-related lung cancer. American Journal of Industrial Medicine, 31(6), 705-712. Title/abstract excluded: Incorrect population.

Wachtel, T. M., Kucia, A. M., & Greenhill, J. A. (2008). Secondary prevention for acute coronary syndrome in rural South Australia: are drugs best? What about the rest? Rural and remote health, 8(4), 967. Title/abstract excluded: Inappropriate outcomes.

Walsh, S. E., Singleton, J. A., Worth, C. T., Krugler, J., Moore, R., Wesley, G. C., et al. (2007). Tobacco cessation counseling training with standardized patients. Journal of dental education, 71(9), 1171-1178. Title/abstract excluded: Inappropriate study design.

Ward, J., & Sanson-Fisher, R. (1996). Does a 3-Day Workshop for Family Medicine Trainees Improve Preventive Care? A Randomized Control Trial. Preventive Medicine, 25(6), 741-747. Title/abstract: Included. Full paper excluded: Inadequate sample size.

Warnakulasuriya, S. (2002). Effectiveness of tobacco counseling in the dental office. Journal of dental education, 66(9), 1079-1087. Title/abstract excluded: Inappropriate study design.

Wensing, M., van der Weijden, T., & Grol, R. (1998). Implementing guidelines and innovations in general practice: which interventions are effective? The British journal of general practice, 48(427), 991-997. Title/abstract: Included. Full paper excluded: Inappropriate outcomes.

Werner, M. J. (1995). Principles of brief intervention for adolescent alcohol, tobacco, and other drug use. Pediatric Clinics of North America, 42(2), 335-349. Title/abstract excluded: Inappropriate study design.

Werrett, J. A. (2005). Exploring the views of parents and health professionals on interventions to reduce infant exposure to environmental tobacco smoke: A pilot study. International Journal of Health Promotion and Education, 43(3), 68-74. Title/abstract excluded: Inappropriate study design.

West, R., McNeill, A., & Raw, M. (2000). Smoking cessation guidelines for health professionals: An update. Thorax, 55(12), 987-999. Title/abstract excluded: Inappropriate study design.

White, M., Bush, J., Kai, J., Bhopal, R., & Rankin, J. (2006). Quitting smoking and experience of smoking cessation interventions among UK Bangladeshi and Pakistani adults: The views of community members and health professionals. Journal of Epidemiology and Community Health, 60(5), 405-411. Title/abstract excluded: Incorrect population.

Whitlock, E. P., Vogt, T. M., Hollis, J. F., & Lichtenstein, E. (1997). Does gender affect response to a brief clinic-based smoking intervention? American Journal of Preventive Medicine, 13(3), 159-166. Title/abstract excluded: Incorrect population.

Wiggins, M., Oakley, A., Roberts, I., Turner, H., Rajan, L., Austerberry, H., et al. (2005). Postnatal support for mothers living in disadvantaged inner city areas: a randomised controlled trial. Journal of Epidemiology and Community Health, 59(4), 288-295. Title/abstract excluded: Incorrect population.

Williams, G. C., McGregor, H., Sharp, D., Kouides, R. W., Levesque, C. S., Ryan, R. M., et al. (2006). A self-determination multiple risk intervention trial to improve smokers' health. Journal of General Internal Medicine, 21(12), 1288-1294. Title/abstract excluded: Incorrect population.

Wilson, A., Hippisley-Cox, J., Coupland, C., Coleman, T., Britton, J., & Barrett, S. (2005). Smoking cessation treatment in primary care: prospective cohort study. Tob Control, 14(4), 242-246. Title/abstract excluded: Inappropriate study design.

Wilson, A., McDonald, P., Hayes, L., & Cooney, J. (1992). Health promotion in the general practice consultation: a minute makes a difference. (General Practice). British Medical Journal, v304(n6821), p227(224). Title/abstract: Included. Full paper excluded: Inadequate sample size.

Wilson, A., Sinfield, P., Rodgers, S., Hammersley, V., & Coleman, T. (2006). Drugs to support smoking cessation in UK general practice: are evidence based guidelines being followed? Qual Saf Health Care, 15(4), 284-288. Title/abstract excluded: Inappropriate study design.

Wilson, D. H., Wakefield, M. A., & Eastman, A. (1990). "Sick of smoking": a minimal intervention cessation programme for use by general practitioners. Proceedings from the 7th.World Conference.on Tobacco and Health, 1, 645-649. Title/abstract excluded: Inappropriate study design.

Wilson, D. H., Wakefield, M. A., Steven, I. D., Rohrsheim, R. A., Esterman, A. J., & Graham, N. M. H. (1990). 'Sick of smoking': Evaluation of a targeted minimal smoking cessation intervention in general practice. Medical Journal of Australia, 152(10), 518-521. Title/abstract excluded: Incorrect population.

Wilson, D. M., Taylor, D. W., Gilbert, J. R., Best, J. A., Lindsay, E. A., Willms, D. G., et al. (1988). A randomized trial of a family physician intervention for smoking cessation. Jama, 260(11), 1570-1574. Title/abstract: Included. Full paper excluded: Incorrect publication date range.

Wilson, D. M. C., Lindsay, E. A., Best, J. A., Gilbert, J. R., Willms, D. G., & Singer, J. (1987). A smoking cessation intervention program for family physicians. Canadian Medical Association Journal, 137(7), 613-619. Title/abstract excluded: Inappropriate study design.

Winickoff, J. P., Hillis, V. J., Palfrey, J. S., Perrin, J. M., & Rigottit, N. A. (2003). A smoking cessation intervention for parents of children who are hospitalized for respiratory illness: The Stop Tobacco Outreach Program. Pediatrics, 111(1), 140-145. Title/abstract excluded: Incorrect population.

Woller, S. C., Smith, S. S., Piasecki, T. M., Jorenby, D. E., Helberg, C. P., Love, R. R., et al. (1995). Are clinicians intervening with their patients who smoke? A 'real-world' assessment of 45 clinics in the upper midwest. Wisconsin Medical Journal, 94(5), 266-272. Title/abstract excluded: Inappropriate study design.

Wood-Baker, R. (2002). Outcome of a smoking cessation programme run in a routine hospital setting. Internal Medicine Journal, 32(1-2), 24-28. Title/abstract excluded: Inappropriate study design.

Wright, J., Warren, E., Reeves, J., Bibby, J., Harrison, S., Dowswell, G., et al. (2003). Effectiveness of multifaceted implementation of guidelines in primary care. Journal of Health Services Research and Policy,

173


8(3), 142-148. Title/abstract: Included. Full paper excluded: Incorrect intervention.

Wyne, A. H., Chohan, A. N., Al Moneef, M. M., & Al Saad, A. S. (2006). Attitudes of general dentists about smoking cessation and prevention in child and adolescent patients in Riyadh, Saudi Arabia. The journal of contemporary dental practice [electronic resource]. 7(1), 35-43. Title/abstract excluded: Inappropriate study design.

Wynn, A., Coleman, T., Barrett, S., & Wilson, A. (2002). Factors associated with the provision of anti-smoking advice in general practice consultations. Br J Gen Pract, 52(485), 997-999. Title/abstract excluded: Inappropriate study design.

Young, J. M., Girgis, S., Bruce, T. A., Hobbs, M., & Ward, J. E. (2008). Acceptability and effectiveness of opportunistic referral of smokers to telephone cessation advice from a nurse: a randomised trial in Australian general practice. BMC Family Practice, 9(-), 16. Title/abstract excluded: Incorrect population.

Young, J. M., & Ward, J. E. (2001). Implementing guidelines for smoking cessation advice in Australian general practice: Opinions, current practices, readiness to change and perceived barriers. Family Practice, 18(1),

14-20. Title/abstract excluded: Inappropriate study design.

Zapka, J. G., Fletcher, K. E., Ma, Y., & Pbert, L. (1997). Physicians and smoking cessation. Evaluation & the Health Professions, 20(4), 407-427. Title/abstract excluded: Inappropriate study design.

Zwar, N. (2008). Smoking cessation--what works? Australian family physician, 37(1-2), 10-14. Title/abstract excluded: Inappropriate study design.

Zwar, N., Hermiz, O., Hasan, I., Comino, E., Middleton, S., Vagholkar, S., et al. (2008). A cluster randomised controlled trial of nurse and GP partnership for care of chronic obstructive pulmonary disease. BMC pulmonary medicine, 8(-), 8. Title/abstract excluded: Inappropriate study design.

Zwar, N., Richmond, R., Borland, R., Stillman, S., Cunningham, M., & Litt, J. (2005). Smoking cessation guidelines for Australian general practice. Australian family physician, 34(6), 461-466. Title/abstract excluded: Inappropriate study design.

Zwar, N. A., & Richmond, R. L. (2006). Role of the general practitioner in smoking cessation. Drug and Alcohol Review. Special Issue: Smoking Cessation Research, 25(1), 21-26. Title/abstract excluded: Inappropriate study design

.

174


175


Appendix F: Quality Checklists for Appraising Interventions

Method of treatment assignment � Correct, blinded randomisation method described OR randomised, double-blind

method stated AND group similarity documented � Blinding and randomisation stated but method not described OR suspect technique (e.g.

allocation by drawing from an envelope) � Randomisation claimed but not described and investigator not blinded. � Randomisation not mentioned

Control of selection bias after treatment assignment � Intention to treat analysis AND full follow-up � Intention to treat analysis AND <15% loss to follow-up � Analysis by treatment received only OR no mention of withdrawals � Analysis by treatment received AND no mention of withdrawals OR more than 15%

withdrawals/loss to follow-up/post-randomisation exclusions Blinding

� Blinding of outcome assessor AND patient and care giver � Blinding of outcome assessor OR patient and care giver � Blinding not done

Outcome assessment (if blinding was not possible) � All patients had standardised assessment � No standardised assessment OR not mentioned

Source: NHMRC (1999) How to review the evidence: systematic identification and review of the scientific literature. Canberra: NHMRC. Modified from I Chalmers, Cochrane Handbook, available on the Cochrane Library CD-ROM

176


177


Appendix G: Data Extraction Tables Systematic reviews Table 25: Anderson et al. (2004) Citation Anderson, P., & Jané-Llopis, E. (2004). How can we increase the involvement of

primary health care in the treatment of tobacco dependence? A meta-analysis. Addiction, 99(3), 299-312.

Level of evidence Level I Country The Netherlands Objective To systematically review studies testing the effectiveness of educational and practice

base strategies to increase the involvement of primary health-care practitioners in the treatment of tobacco dependence.

Study type/design Systematic review/meta-analysis of 19 trials; 16 RCTs, two controlled clinical trials (CCTs), and one interrupted time-series design (ITS), smoking specific and general prevention studies.

Search strategy Databases searched: MEDLINE, EMBASE, CINAHL and the Cochrane Library (1966– 2001). Search terms included ‘intervention studies’, ‘evaluation trials’, ‘continuing education’, ‘reminder systems’, ‘guidelines’ ‘screening programmes’; primary health care; ‘physicians family’; ‘primary health care’, ‘family practice’, ‘nurse clinicians’; ‘tobacco’; ‘smoking cessation’, tobacco use disorder’ and ‘nicotine’.

Type of included studies

Randomised or controlled clinical trials, controlled before and after trials and interrupted time-series designs that reported measures of practitioners’ behaviour. Also, some studies reported bio-chemically verified patient quit rates. Studies focused on the engagement of health-care practitioners’ behaviour in dealing with tobacco or dealing with broader prevention activities that included tobacco; directed to all patients 18 years and older; published between 1966 and 2001; reported in English or Spanish.

Types of participants

In 17 trials, the participants were primary healthcare physicians (including five trials within physician vocational or residency training contexts) and in two trials the participants were practice-based nurses. All trials were undertaken in high-income countries (eight in the USA).

Type of intervention Broadly: Programmes aimed to change practitioners’ behaviour in relation to prevention or treatment and included a tobacco-related outcome measure. Interventions included: (1) Educational interventions (12 trials): included the distribution of educational materials, educational meetings or outreach visits, and the provision of audit and feedback (to both practitioners in training and/or established practitioners). (2) Practice-based interventions (four trials): included revision of professional roles, changes in the medical records systems, and the use of quality monitoring systems. (3) Outreach visits (five trials): when there was a personal visit by a trained person to health practitioners in their work setting. (4) Multi-component interventions (ten trials): when programmes included more than one type of intervention component within an educational or practice-based intervention, or when combined at least one educational and one practice-based intervention category. All single-component programmes were non-outreach programmes. The programmes for practitioners in training were largely non-outreach, single-component, tobacco-specific and educational.

Outcomes Screening rates. Advice-giving. Bio-chemically verified patient quit rates.

Data analyses & statistics

A meta-analysis, using a random effects model, of 24 programmes identified in 19 trials. Effect sizes were adjusted by inverse variance weights to control for studies’ sample sizes.

178


Description of included studies (Author/year; country; study design & duration; general prevention including tobacco [G] or smoking-specific [S] intervention type; outcome; % improvement for intervention group & 95% CI).

Bonevski et al. (1999); Australia; RCT; 3 months; [G] single-component educational; ‘ask’; 1.1 (- 3.4–5.5). Borgiel et al. (1999); Canada; RCT; 12 months; [G] single-component educational; ‘ask’ and/or ‘brief advice’; 4.0 (0.8–7.1). Cockburn et al. (1992); Australia; RCT; 12 months; [S] single-component guideline implementation; cessation support; 5.4 (-7.1–18.0). Cohen et al. (1989); USA; 12 months; [S] single-component/multi-component; reminders/nicotine gum/reminders & nicotine gum at no cost; quit rates; 4.8 (2.0–7.6)/ 3.4 (0.8–5.9)/ 3.6 (1.1–6.2) respectively. Cummings et al. (1989); USA; RCT; 9 months; [S] Multi-component education; ‘brief advice’, 6.0 (1.7–10.3); quit rates 1.1 (-0.1–2.3). Jolly et al. (1999); UK; RCT; [G] Multi-component practice-based (education/nurses’ roles); quit rates; -1.4 (-13.4–10.5). Kottke et al. (1992);USA; CCT; 18 months [S]; Multi-component outreach educational and practice based; ‘brief advice’; 14.6 (8.7–20.5). Lennox et al. (1998); UK; RCT; 14 months [S]; single-component educational; ‘ask’/’brief advise’; 4.5 (-0.7–9.7). Lennox et al. (1998); UK; RCT 14 months [S]; single-component educational; ‘ask/brief advise’; 4.5 (-0.7–9.7). McPhee et al. (1991); USA; RCT; 12 months [G]; multi-component computerized reminder/cancer education; ‘ask’/cessation support (counselled); 12.6 (8.7–16.5)/ 17.1 (10.4–23.9) respectively. Moher (2001); UK; RCT; 18 months [G]; single-component register and systematic recall of patients to general practitioners; ‘ask’ Richmond et al. (1998); Australia; RCT [S]; single-component education workshop; delivery of smoking cessation intervention; 8.9 (0.0–17.7). Robson et al. (1989); UK; RCT; 5 years [G]; multi-component ‘call and recall’; ‘smoking status recorded’; 15.7 (12.4–18.9). Rodney et al. (1985); USA; RCT 12 months [G]; single-component change in design of medical record; ‘smoking status recorded’; 0.1 (-6.0–6.2). Secker-Walker et al. (1992); USA; RCT [S]; single-component education (obstetrics & family practices); ‘brief advice’; 39.3 (28.2–50.4). Spencer et al. (1999); USA; Interrupted Time Series (ITS) 19 months [S]; multi-component computerized screening, audit and feedback, multi-disciplinary team; ‘smoking status recorded’/‘brief advice’; 62.2 (53.2–71.1)/ 31.7 (20.5–42.8). Strecher et al. (1991); USA; RCT 6 months [S]; single-component educational (16.2 (7.8–24.7)) vs. single-component chart-based reminder (15.3 (6.7–23.9) vs. multi-component educational and chart-based reminder (13.3 (4.5–22.0)); ‘brief advise’; (quit rates n.s.). Ward et al. (1996); Australia; RCT 13 weeks [G]; single-component educational three-day workshop; ‘ask’; 14.7 (8.6–20.9). Weingarten et al. (1989); Israel; RCT [G]; single-component computerized vs. manual records; ‘smoking status recorded’; 26.8 (14.5–39.1). Wilson et al. (1992); UK; CCT [G]; single-component prolonging length of each general practice visit by three minutes; ‘ask’/ ‘brief advice’; 2.8 (1.2–4.4)/ 11.4 (4.7–18.1).

179


A (2) Adequate/Reported: The aim/scope or the review was well defined and was limited to educational and practice-base strategies. B (2) Adequate/Reported: Search strategy used included/described. C (2) Adequate/Reported: Inclusion criteria appropriate/un-biased. D (2) Adequate/Reported: The authors reported in detail on the quality assessment methodology. E (2) Adequate/Reported: The results of the individual studies were appropriately summarised by meta-analysis and narrative. F (2) Adequate/Reported: Yes, appropriate statistical methods used: random effects model, effect sizes were adjusted by inverse variance weights to control for studies’ sample sizes. G (2)Adequate/Reported: Yes, Q statistic.

*Review quality See below for “A-G” quality criteria questions†

‡§TOTAL: 14 points; Good. Results (relevant to scope of current review)

The searches identified 696 papers, of which 58 were retrieved and 19 were included in the final analysis. Five of 19 trials were conducted prior to 1990 and ten of 19 trials reported smoking related outcomes within trials of ‘general prevention’ programmes [G]. Three trials were conducted pre-1990 and were general prevention programmes. Eight trials were conducted post-1990 and were smoking specific [S] and reported ‘ask’ and or ‘brief advice’ (two trials reporting only quit rates). – Overall, the programmes led to a 15% (95% CI = 7–22) increase in practitioners’ screening rates (mean control rates of 48%); a 13% (95% CI = 9–18) increase in practitioners’ advice-giving rates (mean control rates of 51%), and a 4.7% (95% CI = 2.5–6.9) increase in bio-chemically verified patient quit rates (mean control rates of 16.9%). – For practitioners in training: programmes were effective in changing their patients’ quit rates but not their own screening rates. – For established practitioners, programmes were effective in changing their screening and advice-giving rates, but not their patients’ quit rates. – Generally: multi-component programmes were more effective than single-component programmes and tobacco-specific programmes were more effective than general prevention programmes for both screening and advising to quit. – Programmes combining educational and practice-based measures were more effective than educational or practice-based alone.

Authors’ conclusions

Primary health-care practitioners can be engaged in the treatment of tobacco dependence to increase equally their screening and advice giving rates and their patients’ quit rates. The provision of educational interventions for practitioners in training in combination with systematic outreach practice based support for established practitioners is likely to be an effective strategy to increase the delivery of smoking cessation treatments in primary care settings and increase smoking quit rates. These findings suggest that the behaviours of practitioners in training might be more malleable than those of established practitioners. Established practitioners might be less likely to be influenced by educational interventions alone but more likely to use supporting practice materials such as computerised auditing systems, and to become involved in smoking cessation when stimulated by an outreach strategy.

180


Reviewers notes This comprehensive meta-analysis reports data pertaining to the effectiveness of two

types of interventions aimed to increase the delivery of smoking cessation treatment in primary care settings, namely, education and practice based (organisational) strategies.

Relevance to study question

The review does not include other types of interventions, for example: incentives or interventions aimed at HCPs other than physicians (and to a lesser extent nurses).

* As per NHMRC Interim Levels of Evidence (NHMRC 2005) for Evaluating Intervention Studies †The quality of systematic reviews was assessed using the following questions: (A) Was a clinical question clearly defined? (B) Was an adequate search strategy used? (C) Were the inclusion criteria appropriate and applied in an unbiased way? (D) Was a quality assessment of included studies undertaken? (E) Were the characteristics and results of the individual studies appropriately summarised? (F) Were the methods for pooling the data appropriate? (G) Were sources of heterogeneity explored?

‡ For each individual answer, the following scores were assigned: Adequate/reported = 2 Inadequate = 1 Unknown/not reported = 0 § The following thresholds for study quality have been applied: – An overall study score of 1-4 is rated Poor – An overall study score of 5-10 is rated Fair – An overall study score of 11-14 is rated Good

Abbreviations: CBA = controlled before-after trial; CCT = controlled clinical trials; CI = confidence interval; C-RCTs = cluster randomised controlled trials; EHR = electronic health record; EPOC = effective practice organisation of care; GP = General Practitioner; HCP = health care professional; HMO = Health Maintenance Organisation; HRA = health risk appraisal; ITS = interrupted time-series design; RCT = randomised controlled trial; TS = time-series design; V.A. = Veterans Affairs.

181


Table 26: Dickey et al. (1999) Citation Dickey, L. L., Gemson, D. H., & Carney, P. (1999). Office system interventions

supporting primary care-based health behaviour change counselling. American Journal of Preventive Medicine, 17(4), 299-308.

Level of evidence Level II Country U.S.A. Objective To systematically review the literature on the effectiveness of the two office system

interventions of tools (including questionnaires, appraisals, health risk assessment checklists and flow charts, chart stickers, reminder letters, and education handbooks, brochures, and pamphlets) and teamwork (the coordination and delegation of tasks between HCPs) to improve behaviour-change counselling in primary care.

Study type/design Systematic review of 33 trials: Seven trials of smoking specific [S] ‘office tools’ interventions; 17 trials of general prevention [G] ‘office tools’ interventions; six trials of smoking specific [S] ‘teamwork’ interventions; three trials of general prevention [G] ‘teamwork’ interventions.

Search strategy Databases searched: MEDLINE from 1980, and secondary sources from references lists.


One Meta-analysis, thirteen randomised controlled trials (RCTs), five controlled clinical trials (CCTs), ten time-series designs (TS), and four pre-test/post-test trials that reported measures of practitioners’ behaviour. Note: while some studies did report patient outcomes (for example, bio-chemically verified quit rates) such outcomes were stated by the authors to be outside the scope of the review.


Primary healthcare physicians, practice-based nurses, other health care professionals and administrative staff.

Type of intervention Broadly, programmes aimed to change practitioners’ behaviour in relation providing prevention counselling. Office tools: health risk assessment questionnaires and appraisals, prompting and reminding (chart stickers, checklists, flow charts, reminder letters), and education (manuals and handbooks). Teamwork: the coordination and delegation of tasks between all staff (clinical and administrative).

Outcomes Counselling rates for smoking cessation as reported by patient survey and/or chart review in 13 ‘smoking specific’ trials (including one meta-analysis), and in 12 trials, counselling rates for smoking cessation reported within ‘general prevention’ programmes, and in eight trials, other outcomes reported within ‘general prevention’ programmes that did not relate to smoking counselling.


Narrative synthesis.

Description of included studies (Author; year; country; study design & duration; general prevention (including smoking/cancer prevention) [G] or smoking-specific [S] or general prevention but not reporting smoking separately [Gns]; intervention type; outcome; results.

Battista et al. (1991); Canada; CBA; 6 months; [G] two-component– educational/chart flow sheet; increases of 4% to 19% in counselling rates for 9 of 12 prevention areas. Overall improvement in the unit scores was statistically significant (p < .05). Belcher, (1990); U.S.A. (V.A.); RCT; 5 years; [G] multi-component– provider training, feedback, team work, and chart flow sheet; no significant change in counselling rates for smoking and alcohol use. Bull & Jamrozik, (1998); Australia; CCT; 12 months; [Gns] single component– assessment questionnaire/patient pamphlets; at 6 months, significantly higher percentages of intervention patients reported being physically active (38% vs. 30%); at 12 months, (36% vs. 31%), but n.s. Chang et al. (1995); U.S.A.; TS; 1,3 & 8 months; [S] single component– chart stickers (smoking status); documentation of smoking status increased from 33% to 83% after 8 months (p = .0001). More than twice as many smokers received counselling. Cohen et al. (1989); U.S.A.; RCT; 12 months; [S] two component– chart stickers (smoking status and setting a quit date) and provider education sessions; patients whose charts had stickers had 15% smoking cessation rate at one year vs. 2.7% rate for controls (p < .05). Cummings et al. (1989); U.S.A.(HMO); RCT; 12 months; [S] multi-component– chart reminder stickers, prescription form, exam and waiting room posters, patient self-help booklet and physician education sessions; significantly increased counselling services (p < .05) but after one year, bio-chemically confirmed abstinence was increased 1% for intervention but not control (95% CI, -0.1% to +2.3%) n.s. Cummings et al. (1989); U.S.A. (private practice); RCT; 12 months; [S] multi-component– chart reminder stickers, prescription form, exam and waiting room

182


posters, patient self-help booklet and physician education sessions; significantly increased counselling services – “Brief advice” 44% (control) vs. 64% ( exp); “Cessation” set quit dates 5% (control) vs. 29% ( exp). ; made follow-up appointments 11% (control) vs. 19% (exp) (p < .05) but after 12 months, bio-chemically confirmed abstinence was increased for intervention patients (3.2%) and control (2.5%) (95% CI for the 0.7% difference: -1.7 to +3.1%) n.s. Dietrich et al. (1992); U.S.A. (private practice); RCT; 12 months; [G] multi-component– chart flow sheet, chart stickers (smoking status), posters, brochures, patient-held health diaries, and team work (division of responsibilities). After 12 months, providers with office system intervention had a significantly increased rate of counselling for dietary fat consumption 49% (baseline) vs. 56% (follow-up) and non-significant increases in fibre consumption 40% (baseline) vs. 48% (follow-up) and smoking cessation counselling 83% (baseline) vs. 84% (follow-up), compared with physicians receiving education only. Geiger et al. (1993); U.S.A.; TS; [Gns] single component– health promotion visits; significantly greater rates of counselling for diet, exercise, substance abuse, sexual practices, injury prevention, and dental care than patients seen before intervention. Gemson & Sloan, (1995); U.S.A.; CBA; 9 months; [G] multi-component– provider handbook of preventive services, provider education, posters, and patient-held health diaries; physicians reported significant increases in counselling rates for smoking, nutrition/weight and exercise counselling (confirmed by patient reports, except smoking); no significant increases in counselling at comparison hospital. Gemson & Sloan, (1995); U.S.A. (work-site medical clinic); RCT; 6 months; [Gns] single-component– health risk appraisal (HRA); employees receiving HRA feedback and counselling had significantly greater improvements in physical activity and appraised age after six months; non-significant greater improvements in cholesterol, blood pressure, weight, and seatbelt use. Johns et al. (1992); Student health centre; TS; [Gns] single component–reminder chart checklist for 11 major behavioural risk factors; during intervention period documentation of counselling increased significantly for 13 of 18 providers; counselling behaviour returned to baseline when checklist was removed. Kreuter & Strecher, (1996); U.S.A. (family medicine practice); RCT; 6 months; [G] single-component– health risk appraisal (HRA); patients receiving tailored feedback vs. typical HRA feedback were more likely to change at least one of seven risk behaviours. Logsdon et al. (1984; 1989); U.S.A. (multi-specialty group practices); CBA; 12 months; [G] multi- component– provider handbook, patient handbook, patient risk profile, visit encounter form, prevention prescription pad, provider education sessions; intervention patients at risk reported significant improvements in four of five risk behaviours (exercise, using seat belts, lost 5 lbs, reduced heavy drinking) but non-significant increase in quitting smoking (16.2% vs.11.5%). McIlvain et al. (1992); TS; 3-6 months; [S] two component– chart reminder stickers and provider education sessions; by patient report, physicians counselled significantly more smokers at three months; by six months counselling was still increased, but not significantly (the effect of the intervention was restricted to two clinics using chart stickers). McPhee et al. (1991); U.S.A. (family medicine); RCT; 12 months; [G] multi-component– computer-generated physician and patient reminder notices and health education brochures; intervention providers had significantly higher increases in counselling for all four behaviour areas (smoking assessment, 10.2%; smoking counselling, 17.3%; diet assessment, 12.3%; exercise counselling, 13.9%). Madlon-Kay et al. (1995); U.S.A. (family medicine); CBA; [G] multi-component– patient questionnaire (seatbelt use, exercise, nutrition, stress, smoking, alcohol and drugs, dental health, safe sex, pregnancy prevention); significantly better documentation of counselling for patients using questionnaire vs. those not using questionnaire in exercise, stress, seatbelt use, dental health, and alcohol and drug use. Ornstein et al. (1995); U.S.A. (family medicine); TS; 2 years; [G] single component– computer-generated onscreen physician reminders and printed patient reminders; after 15 months, counselling rates for dental, diet, injury prevention, passive smoking, exercise, sexuality, and substance abuse increased 13% to 15% each. Patterson et al. (1989); U.S.A. (family medicine); TS; 6 months; [G] two-component– provider handbook and provider education; over 6 months, significantly increased compliance with recommended protocols for smoking, obesity, and general health promotion; non-significant increase in compliance for hypertension care. Pinto et al. (1998); U.S.A. (family medicine); RCT; 8 months; [Gns] multi-component– provider handbook, office poster, physical activity assessment, patient manual, provider education sessions; intervention physicians (vs. control physicians) no difference in reported frequency of counselling activities.

183


Robinson et al. (1995); U.S.A. (family medicine); TS; 2 months; [S] single-component– chart stamp (smoking status); after 2 months, counselling rate of smokers had increased (47% to 86%). Shank et al. (1989); U.S.A. (family medicine); TS; 3 years; [G] multi-component– chart flow sheet, provider education and feedback; significant increases in smoking history taking (35% to 52%), alcohol history taking (30% to 51%), exercise/sleep counselling (8% to 22%), and diet counselling (9% to 29%); non-significant changes in contraception, parental guidance, and seatbelt use counselling. Smith et al. (1985); U.S.A. (family medicine); RCT; 6 months; [G] single-component– health risk appraisal (HRA); no significant differences in weight, blood pressure, cholesterol, health risk age, or reported rate of seatbelt use, alcohol use, smoking, or exercise. Strecher et al. (1991); U.S.A. (family medicine); RCT 4 group design; 6 months; [S] two-component– chart prompt vs. provider education vs. both vs. control; providers receiving education plus prompt, or education alone, reported significantly increased smoking-cessation counselling, but providers receiving prompt alone or no intervention did not. Providers in intervention groups advised more patients to quit smoking (76% to 79%) than did control providers (69%). Calfas et al. (1996); CCT; [Gns] single-component– team work; intervention group patients demonstrated significantly higher rates of increases in physical activity than control group patients. Duncan et al. (1991); U.S.A. (HMO); CCT; [S] single-component– team-work: training without support staff, vs. those who received group training with their staff; those trained to work with staff members counselled more patients (63% vs. 27%), set quit dates for more patients (28% vs. 16%), sent reminders to more patients (15% vs. 2%), and provided more self-help booklets (34% vs. 10%). Hollis et al. (1991); U.S.A.; CCT; [S] single-component– team-work: nurse-assisted smoking counselling vs. physician only; demonstrated that nursing staff can be of significant assistance in actually delivering smoking cessation counselling. Kottke et al. (1992); U.S.A.; CCT; 6 months (typical rather than volunteer clinics); [S] single-component– team-work to integrate smoking cessation advice into primary care practice; intervention cohort was significantly more likely to report that someone had asked them if they smoked (39.8% +/- 12.3% vs. 26.0% +/- 12.2%; p < .05), that their physician asked them to quit if they were currently smoking (40.5% +/- 12.1% vs. 26.4% +/- 14.6%; p < .05). Kottke et al. (1988); Meta-analysis of 39 controlled trials; [S] single-component– team work: physician and non-physician interveners vs. physician or non-physician interveners; trials using both physician and non-physician interveners had a significantly higher mean smoking-cessation rate at 6-month follow-up (23%) than trials using physicians (5%) or non-physicians (9%) alone. Long et al. (1996); A multi-site field trial; [Gns] single-component– team work: training for primary care providers and office staff; training was effective in preparing the providers to counsel (reported 70% compliance with the recommendations given). Ockene et al. (1996); U.S.A. (HMO); RCT 3 groups; [Gns] single-component– team-work: usual care, physician training in lipid intervention, and physician training in lipid intervention plus practice management provided by support staff; physician training, including use of practice management tools, did not result in counselling rates different from usual care, although significant improvements in counselling rates were found when staff were involved in implementing practice management tools. Ockene et al. (1991); U.S.A.; RCT; [S] single-component– team work: three physician-delivered smoking-cessation interventions in combination with follow-up contact from behavioural counsellors; the follow-up telephone counselling by behavioural counsellors did not contribute to smoking behaviour changes. Solberg et al. (1990); U.S.A.; TS; 2 years; [S] single-component– team-work (organised identification, progress records, brief physician messages, follow-up, and division of responsibilities e.g. medical assistant identified smokers); at two years, overall quit rate was approximately 20%, rising to 33% for those tobacco users with more clinic contacts or at least one year from the first to the latest contact

184


A (2) Adequate/Reported: The aim/scope or the review was well defined and was limited to trial of interventions aimed at changing HCPs behaviours. B (1) Inadequate: Search strategy used included and described, but briefly. C (0) Not reported: Inclusion criteria not fully defined but appropriate exclusion of studies aimed to evaluate the effectiveness of provider counselling at the patient level. D (2) Adequate/Reported: The authors reported briefly on the quality assessment methodology. The authors opted to provide the necessary caveats on a case-by-case basis due to the diverse methodology, modest scope and lack of randomisation. E (2) Adequate/Reported: The results and characteristics of the individual studies were reported (briefly) and summarised by narrative synthesis. F (2) Adequate/Reported: Not applicable. G (2) Adequate/Reported: Not applicable.


‡§TOTAL: 11 points; Good. Note: this review is in the form of a concise nine page published journal article, rather than a full Systematic Review manuscript. Thus, some of the above criteria may not have been fully detailed due to the editing requirements of the journal publication.

Results (relevant to scope of current review)

Eight of 33 trials were conducted prior to 1990. Counselling rates for smoking cessation were reported for 13 ‘smoking specific’ trials (including one meta-analysis) and within 12 trials of ‘general prevention’ programmes (eight trials reported other outcomes that did not relate to smoking counselling). Of the thirteen ‘smoking specific’ studies, 92% reported a statistically significant increase in the provision of smoking cessation counselling. Of twelve ‘general prevention’ trials, 58% reported a statistically significant increase in the provision of smoking cessation counselling (along with a range of other provider behaviour-change outcomes, some significant, some not).


–Smoking cessation clinical trials have demonstrated the effectiveness of tools and teamwork for increasing counselling effectiveness. – Generally: multi-component programmes are more effective than single-component programmes and programmes implementing office tools and teamwork were more effective than office tools or teamwork alone. –Chart prompts are well accepted by providers, and their use has been shown to be related to improved counselling rates. –Electronic tools for prompting, tracking, and education hold promise for making much needed improvements in the efficiency and effectiveness of counselling. Evidence is less conclusive as to whether use of manual tracking systems, such as patient chart flow sheets and card files, leads to improved rates of behaviour-change counselling –The most effective type of teamwork and/or tools probably varies according to provider, staff, patient, and practice characteristics. In general, however, effectiveness is not only related to the type, but also to the intensity and number of different office interventions used.

Reviewers notes In this and other studies, office staff have been found to be vital for performing a variety of administrative actions, such as arranging patient appointments, disseminating and reviewing behavioural risk questionnaires, applying stickers and reminders to patient charts, distributing educational materials, and performing follow-up mail and telephone contacts. Sharing the responsibility for office-based smoking cessation activities among all staff appears to be a promising strategy. However, perhaps unsurprisingly, the effects of interventions aimed at increasing the rates of smoking cessation counselling are often attenuated when the intervention forms part of a ‘general prevention’ programme. Thus, the generalisability of smoking specific interventions to ‘real world’ settings is questionable in some cases.

185



Reports data pertaining to the effectiveness of two types of interventions aimed to increase the delivery of smoking cessation treatment in primary care settings: namely, ‘office tools’ and ‘teamwork’. Included studies comprise a collection of smoking specific, general prevention and general prevention-not-reporting-smoking trials. This is a typical example of the diverse literature in this field.




186


Table 27: Fiore et al. (2008) Citation Fiore, M., Jaén, C., Baker, T., et al. (2008). Treating Tobacco Use and Dependence:

2008 Update. From U.S. Department of Health and Human Services. Public Health Service

Level of evidence Level I Country U.S.A. Objective This evidence-based clinical practice guideline contains strategies and

recommendations designed to assist clinicians, tobacco dependence treatment specialists, health care administrators, insurers, and purchasers in delivering and supporting effective treatments for tobacco use and dependence. Eleven new topics were chosen by the 2008 Guideline Panel for updated meta-analysis, including (relevant to this review) topic eleven: “Effectiveness of systems interventions, including provider training and the combination of training and systems interventions”.

Study type/design Evidence-based clinical practice guideline up-date: based on three systematic reviews/meta-analyses of randomised controlled trials (and some other study designs as appropriate) conducted up to 1996, 1999 and up-dated in 2007, smoking specific.

Search strategy Databases searched: 11 electronic databases. Type of included studies

Randomised controlled trials. Cluster randomised controlled trials.


HCPs, General Practitioners, nurses, pharmacists and pharmacy assistants, and other office staff.

Type of intervention HCP training and reminder systems (charting systems) aimed to change practitioners’ behaviour in relation to providing smoking cessation interventions in primary care settings. Training: Included didactic presentation of material, group discussions, and role playing; from formal training during residency to on-site clinician training within the community. Reminders: Any charting system, such as chart reminder stickers or treatment algorithms attached to the chart.

Outcomes Clinician assessment of smoking status (Ask), provision of treatment (Assist/Cessation), and arranging for treatment follow-up (Arrange/Cessation). Note: Also reported on abstinence outcomes.


The primary meta-analytic model used in this and the previous two Guidelines was logistic regression using random effects modelling. Some studies involved randomisation at the clinician/clinic level, and screening of such studies considered factors such as data non-independence, the evaluation of pre-intervention or baseline status, and the number and types of higher level units.

Description of included studies (Author/year; country; study design & duration; intervention type; outcome

Cornuz et al. (2002): Switzerland, medical clinics, cluster randomised controlled trial, 35 HCPs, 12 months, training (two half-day workshops), ‘A’, ‘B’ and ‘C’. Pbert et al. (2006): U.S.A. paediatric clinics, 61 HCPs, RCT, 12 months, training and reminders, ask ‘A’. Cummings et al. (1989): U.S.A primary care Health Maintenance Organisation; RCT, 12 months; 81 HCPs; training- group tutorial 3X1 hour (personalised health advice, quit date, one follow up maintenance visit, self help materials, nicotine gum); cessation support ‘C’. Cummings et al. (1989): U.S.A primary care private practice; RCT, 12 months; 44 HCPs; training- group tutorial 3X1 hour (personalised health advice, quit date, one follow-up visit, self help materials, nicotine gum); cessation support ‘C’. Sinclair et al. (1998): U.K., rural community pharmacies; RCT, 62 community pharmacies (including both pharmacists and assistants); training two-hour workshop based on stages of change model; ask ‘A’.

187


Guideline quality appraisal (AGREE instrument) Scope and Purpose The overall objective(s) of the guideline is (are) specifically described.

Strongly Agree � Strongly Disagree The clinical question(s) covered by the guideline is (are) specifically described.

Strongly Agree � Strongly Disagree The patients to whom the guideline is meant to apply are specifically described.

Strongly Agree � Strongly Disagree

Comments The aim/scope or the Clinical Practice Guideline was well defined and includes a chapter on ‘System interventions’ (chapter 5). The aim, as stated in chapter five, was to evaluate the effectiveness of interventions intended to increase rates of delivering tobacco dependence treatments, quit attempts, and successful smoking cessation.

Stakeholder Involvement The guideline development group includes individuals from all the relevant professional groups.

Strongly Agree � Strongly Disagree The patients’ views and preferences have been sought.


Comments Yes. With respect to Chapter 5 ‘System Interventions’, the Clinician or other Health Care Professional constitutes the population, rather the patient per se.

The target users of the guideline are clearly defined.

Strongly Agree � Strongly Disagree The guideline has been piloted among target users.

Strongly Agree � Strongly Disagree Rigour of development Systematic methods were used to search for evidence.


Comments The three searches identified approximately 8,700 articles (includes approximately 2,700 articles added to the literature since publication of the 2000 Guideline) and more than 300 articles were identified for possible inclusion in a meta-analysis. More than 600 additional articles were examined in detail by the Panel and were used in the formulation

Domain score 12 100

%

Domain score 15 91%

188


recommendations that were not supported by meta-analyses. Five articles were included that reported on process outcomes (ask, brief advice, cessation support), where the HCP or clinic was the unit of randomisation. Comments cont. The authors reported in detail on the quality assessment methodology. The results of the individual studies were appropriately summarised by meta-analysis and narrative. Appropriate statistical methods used: considered factors such as data non-independence, the evaluation of pre-intervention or baseline status, and the number and types of higher level units.

The criteria for selecting the evidence are clearly described.


Comments Inclusion criteria appropriate: (a) reported the results of a randomised controlled trials or cluster randomised controlled trials, (b) provided follow-up results at least five months after the quit date, (c) was published in a peer-reviewed journal; (d) was published between January 1975 and June 2007, (e) was published in English and (f) was one of the 11 topics chosen to be included in the 2008 update.

The methods used for formulating the recommendations are clearly described.

Strongly Agree � Strongly Disagree The health benefits, side effects and risks have been considered in formulating the recommendations.

Strongly Agree � Strongly Disagree There is an explicit link between the recommendations and the supporting evidence.

Strongly Agree � Strongly Disagree The guideline has been externally reviewed by experts prior to its publication.


Comments Eighty one peer reviewers, 15 public submissions, and the draft document was published in the public domain.

A procedure for updating the guideline is provided.

Strongly Agree � Strongly Disagree Clarity and presentation The recommendations are specific and unambiguous.

Strongly Agree � Strongly Disagree The different options for management of the condition are clearly presented.

Strongly Agree � Strongly Disagree Key recommendations are easily identifiable.

Strongly Agree � Strongly Disagree The guideline is supported with tools for application.

Domain score 26 90%

189


Strongly Agree � Strongly Disagree Applicability The potential organisational barriers in applying the recommendations have been discussed.

Strongly Agree � Strongly Disagree The potential cost implications of applying the recommendations have been considered.

Strongly Agree � Strongly Disagree The guideline presents key review criteria for monitoring and/or audit purposes.

Strongly Agree � Strongly Disagree Editorial independence The guideline is editorially independent from the funding body.

Strongly Agree � Strongly Disagree Conflicts of interest of guideline development members have been recorded.

Strongly Agree � Strongly Disagree Further comments

This 2008 Guideline broadens the scope from the individual patient level (patient outcomes, e.g. quit rates), to include interventions aimed at changing health professionals’ behaviour (process outcomes, e.g. documenting smoking status). Specifically, chapter five of the Guideline is based on (but not limited to) studies of the effectiveness of systems interventions, including provider training and the combination of training and systems interventions.

Overall assessment: would you recommend these guidelines for use in practice?

Strongly recommend Recommend (with provisos or alterations) Would not recommend Unsure

Adapted from: The AGREE Collaboration (2003). Appraisal of Guidelines for Research & Evaluation (AGREE) Agree Instrument training manual, the AGREE instrument: short form. www.agreecollaboration.org

Domain score 15 91%

Domain score 9 67%

Domain score

83%

190



The three searches identified approximately 8,700 articles (includes approximately 2,700 articles added to the literature since publication of the 2000 Guideline) and more than 300 articles were identified for possible inclusion in a meta-analysis. More than 600 additional articles were examined in detail by the Panel and were used in the formulation recommendations that were not supported by meta-analyses. Five articles were included that reported on process outcomes (ask, brief advice, cessation support), where the HCP or clinic was the unit of randomisation. Results from the 2008 meta analyses: topic 11.

Effectiveness of clinician training combined with charting on asking about smoking status (Ask) (n = 3 studies) Intervention Number of

arms Odds Ratio (95%CI)

Estimated rate (95% CI)

No intervention

3 1.0 58.8

Training & Charting

3 2.1 (1.9–2.4) 75.2 (72.7–77.6)

Effectiveness of clinician training on rates of providing treatment (Assist/Cessation) (n = 2 studies)

Intervention

Number of arms

Odds Ratio (95% CI)


No intervention

2 1.0 36.2

Clinician training

2 3.2 (2.0–5.2) 64.7 (53.1–74.8)

Effectiveness of training combined with charting on setting a quit date (Assist/Cessation) (n = 2 studies) Intervention Number of



No intervention

2 1.0 11.4

Training & Charting

2 5.5 (4.1–7.4) 41.4 (34.4–48.8)

Effectiveness of training combined with charting on providing materials (Assist/Cessation) (n = 2 studies) Intervention Number of



No intervention

2 1.0 8.7

Training & Charting

2 4.2 (3.4–5.3) 28.6 (24.3–33.4)

Effectiveness of training combined with charting on arranging for follow-up (Arrange/Cessation) (n = 2 studies) Intervention Number of



No intervention

2 1.0 6.7

Training & Charting

2 2.7 (1.9–3.9) 16.3 (11.8–22.1)

Based exclusively on ‘intention to treat’ data.


– Clinician training increases the delivery of effective tobacco use treatments. – Further, combining clinician training with a charting system, such as chart reminder stickers or treatment algorithms attached to the chart, increases rates of tobacco use assessment, setting a quit date, providing materials, and arranging for follow-up. – Training should be reinforced throughout the clinicians’ education and practice.

191


Reviewers notes The guidelines list the following recommendations (�strength-of-evidence rating B): Health care delivery administrators, insurers, and purchasers can promote the treatment of tobacco dependence through a systems approach and should, therefore, make tobacco assessment and coverage of treatment a contractual obligation of the health care insurers and/or clinicians who provide services to them. Various institutional policies would facilitate these interventions, including: • Implementing a tobacco user identification system in every clinic. • Providing adequate training, resources, and feedback to ensure that providers consistently deliver effective treatments. • Dedicating staff to provide tobacco dependence treatment and assessing the delivery of this treatment in staff performance evaluations. • Promoting hospital policies that support and provide tobacco dependence services. • Including tobacco dependence treatments (both counselling and medication) identified as effective in this Guideline as paid or covered services for all subscribers or members of health insurance packages. �The three strength-of-evidence ratings

Multiple well-designed randomised clinical trials, directly relevant to the recommendation, yielded a consistent pattern of findings. Evidence from randomised clinical trials supported the recommendation, but the scientific support was not optimal. For instance, few randomised trials existed, the trials that did exist were somewhat inconsistent, or the trials were not directly relevant to the recommendation. Reserved for important clinical situations in which the Panel achieved consensus on the recommendation in the absence of relevant randomised controlled trials.

The guidelines highlight the following areas regarding clinician training for future research: • Effectiveness of training programs for other health disciplines, such as nursing, psychology, dentistry (including hygienists), social work, and pharmacy. • Effective elements in successful training programs (e.g., continuing medical education, interactive components). • Combined effect of multiple systems changes, such as clinician training, reminder systems, clinician feedback, incentive payments, and recruitment of opinion leaders.


This Guideline is the only Clinical Practice Guideline identified in the search that broadens the scope from the individual patient level (patient outcomes e.g. quit rates), to include interventions aimed at changing health professionals’ behaviour (process outcomes, e.g. documenting smoking status). Specifically, chapter five of the Guideline is based on (but not limited to) studies of the effectiveness of systems interventions, including provider training and the combination of training and systems interventions.

192


Table 28: Grimshaw et al. (2004) Citation Grimshaw, J. M., Thomas, R. E., MacLennan, G., Fraser, C., Ramsay, C. R., Vale, L.,

et al. (2004). Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technology Assessment, 8(6), iii-iv, 1-72.

Level of evidence Level I / III (a systematic review of level II, III and III-3 studies) Country U.K. Objective To undertake a systematic review of the effectiveness and costs of different guideline

development, dissemination and implementation strategies. Study type/design Systematic review of 235 trials reporting 309 comparisons, smoking specific and

general prevention studies including smoking. Search strategy Databases searched: MEDLINE, Healthstar, Cochrane Controlled Trial Register,

EMBASE, SIGLE and the specialised register of the Cochrane Effective Practice and Organisation of Care (EPOC) group.


Cluster Randomised Controlled Trials, RCTs, CCTs, CBAs and ITS trials that reported measures of practitioners’ behaviour and/or clinic performance. Also, some studies reported patient outcomes. Thirty-six studies aimed to change one behaviour and the remainder targeted several behaviours (to a maximum of six behaviours). The study population always involved clinicians, and in 14 studies, other healthcare professionals were also involved (most commonly nurses).


Medically qualified healthcare professionals (HCPs).

Type of interventions (∆ Evaluated within smoking trials).

∆Distribution of educational materials: distribution of published or printed recommendations for clinical care, including clinical practice guidelines, audiovisual materials and electronic publications. ∆Educational meetings: healthcare providers who have participated in conferences, lectures, workshops or traineeships. ∆Educational outreach visits: use of a trained person who met with providers in their practice settings to give information with the intent of changing the provider’s practice. Local opinion leaders: use of providers nominated by their colleagues as ‘educationally influential’. ∆Reminders: patient- or encounter-specific information, provided verbally, on paper or on a computer screen, which is designed or intended to prompt a health professional to recall information. ∆Audit and feedback: any summary of clinical performance of healthcare over a specified period. Local consensus processes: inclusion of participating providers in discussion to ensure that they agreed that the chosen clinical problem was important and the approach to managing the problem was appropriate. Patient-mediated interventions: new clinical information (not previously available) collected directly from patients and given to the provider, e.g. depression scores from an instrument. Marketing: use of personal interviewing, group discussion (‘focus groups’), or a survey of targeted providers to identify barriers to change and subsequent design of an intervention that addresses identified barriers.

Outcomes Dichotomous process variables, including asking and/or documenting smoking status ‘A’, providing and/or documenting the provision of brief advise, and the provision and/or documentation of any component of cessation support ‘C’ (quit rates ‘Q’ also reported for some studies).


Single estimates of dichotomous process variables were derived for each study comparison based upon the primary end-point or the median measure across several reported end-points. Separate analyses were undertaken for comparisons of different types of intervention. A planned meta-regression analysis could not be undertaken owing to the large number of different combinations of multifaceted interventions.

193


Description of included studies (Author/year; country; study design & duration; intervention type; outcome

Belcher et al. 1990; U.S.A.; RCT (patients the unit of randomisation), 5 years, general prevention including smoking, revision of professional roles, continuity of care, changes to the site and setting of service delivery, audit and feedback; percentage of patients smoking status recorded for Health Care Clinic vs. Control: pre-intervention 21.00 vs. 28.00, post-intervention 73.00 vs. 28.00 p < .05, ‘A’. Cohen et al. 1987; U.S.A.; RCT, nicotine gum, chart reminders, education; percentage of patients asked by physician about setting a quit date; percentage of patients stopped smoking at 12 months, post-intervention 5.20 vs. 1.50, ‘A’. Dietrich et al. 1992; U.S.A.; RCT 2x2 factorial trial, 12 months; audit and feedback, educational outreach visits, educational materials, recommendations to quit smoking ‘B’ (p = .017) however, counselling ‘C’, N.S. McPhee et al. 1991; U.S.A.; RCT, 12 months; distribution of educational materials, computer generated reminders, patient reminders; percentage of smoking status assessed 32.4% (SD 13.9)(control)/45.0% (16.6)(exp); p = .014 and smokers counselled 41.8% (SD 22.2)(control)/58.8% (23.0)(exp); p = .027. ‘A’ & ‘C’. Messimer et al. 1989; U.S.A.; RCT, 11 private primary care practices; distribution of educational materials and educational meetings to implement the American Lung Association's ‘Because You Love Your Baby’ smoking intervention; quit rates. Ockene et al. 1994; U.S.A.; RCT, provision of intervention algorithm at each patient visit; primary measure: smoking cessation counselling score (patient exit interview); relative deterioration of –1.2% in performance ‘B’. Rosser et al. 1991; Canada; RCT (patients the unit of randomisation); preventive services including smoking status – use of computerised reminders; for smoking status 33.7% in the physician reminder group vs. 14.1% in the randomized control group, p < .001, ‘A’. Wilson et al. 1988; Canada; RCT; family/general practice/community; physician training; percentage of patients reporting that physicians mentioned smoking; +39.1% absolute improvement in performance, ‘A’. A (2) Adequate/Reported: The aim/scope or the review was very comprehensive and well defined and covered any identified intervention strategies for the dissemination and implementation of clinical guidelines. B (2) Adequate/Reported: Comprehensive search strategy used included/described. C (2) Adequate/Reported: Inclusion criteria appropriate/un-biased. D (2) Adequate/Reported: The authors reported in detail on the quality assessment methodology: the Cochrane Effective Practice and Organisation of Care (EPOC) group methodology. E (2) Adequate/Reported: The results of the 235 studies reporting 309 comparisons were appropriately summarised and comparisons were based upon the primary end-point or the median measure across several reported end-points. Results were tabulated and reported by detailed narrative. F (2) Adequate/Reported: Yes appropriate statistical methods used. G (2) Adequate/Reported: Yes.


‡§TOTAL: 14 points; Good.

194



The searches identified 5000 potentially relevant papers, of which, 235 reporting 309 comparisons were included in the final analysis. Of these, 73% of comparisons evaluated multifaceted interventions and eight papers reported smoking related outcomes, one within a trial of a ‘general prevention’ programme. Overall, 67% of studies reported dichotomous process measures and 39% reported continuous process measures (patient outcomes also reported in 22% of studies). The overall results (effect sizes: the median absolute difference across post-intervention measures), largely consistent with the limited number of smoking trials, are as follows: Reminders: The most frequently evaluated single intervention. The results were moderate (effect sizes > 10 and ≤ 20%). Educational materials: May have a modest effect on guideline implementation but the effects may be short lived (effect sizes > 5% and ≤ 10%). Educational outreach: Appeared to be more effective than educational materials (effect sizes > 5% and ≤ 10%). Audit and feedback: Appeared to result in modest effects (effect sizes > 5% and ≤ 10%), although this finding was less robust as there were fewer evaluations of these interventions, and only two of the included smoking studies used audit and feedback as an intervention component. Multifaceted interventions: The combination of educational materials, educational meetings and reminders appears more effective than educational materials and educational meetings alone. However, overall, there appeared to be no evidence of an expected dose-response curve associated with increasing the number of components in an intervention. Summary

Generally ineffective Dissemination of printed educational materials. Didactic educational session (‘lecture and textbook’ style instruction). Mixed effects Audit and feedback. Local opinion leaders. Generally effective Reminders Educational outreach Multifaceted interventions

Results of single interventions and multi-faceted interventions from Cluster Randomised Controlled trials (C-RCTs) Intervention Nu

mber of trials

Range of observed effects*

Median effect size*


5 +3.6 to +17.0

+8.1

Audit and feedback

5 +1.3 to +16.0

+7.0

Reminders 14 -1.0 to +34.0

+14.1

Multi-faceted including educational outreach

13 -4.0 to +17.4

+6.0

*Absolute percentage improvement in guideline compliance

195



There is an imperfect evidence base to support decisions about which guideline dissemination and implementation strategies are likely to be efficient under different circumstances. Decision makers need to use considerable judgement about how best to use the limited resources they have for clinical governance and related activities to maximise population benefits.

Reviewers notes This review is perhaps the most comprehensive review of the effectiveness of different guideline implementation strategies to date. While the number of smoking trials is relatively low, inferences may be drawn judiciously from the overall findings. The identified interventions have been applied in multiple contexts and consider the performance of HCPs in providing a range of preventive services to patients, including smoking cessation advice and cessation support. Therefore, when considering allocating resources to improve preventive activities in ‘real world’ settings, the overall findings may in fact provide guidance that is more generalisable than single intervention ‘smoking only’ trials.


This review is relevant to the study question in that it reports on many trials of a range of interventions aimed at changing HCPs’ behaviour to increase smoking cessation activities and support for their patients.




196


Table 29: Lancaster et al. (2000) Citation Lancaster, T., & Fowler, G. (2000). Training health professionals in smoking cessation.

Cochrane Database Systematic Reviews (3), Art. No.: CD000214. Level of evidence Level I Country U.K., nine studies from U.S.A. and one from Canada. Objective To assess the effectiveness of interventions that train healthcare professionals in

methods for improving the quality of care delivered to patients who smoke. Hypothesis (a) that training healthcare professionals is more effective than no training in increasing the number of smokers who are offered advice about stopping and who subsequently achieve abstinence. Hypothesis (b) that the effect of training can be enhanced by either providing prompts and reminders to healthcare professionals to offer advice about smoking cessation to their patients or encouraging them to offer nicotine replacement therapy as an adjunct to their advice to smokers.

Study type/design Systematic literature review including ten randomised controlled trials, smoking specific.

Search strategy Databases searched: Cochrane Tobacco Addiction Group trials register. Type of included studies

Randomised trials in which the intervention was training of health care professionals in smoking cessation. Trials that reported process outcomes (A, B and/or C), but excluding trials that only reported process outcomes (and not smoking behaviour).


Included trials in which the unit of randomisation was a healthcare practitioner or practice. Almost all the training was provided on a group basis in a tutorial or workshop setting and the duration of training ranged from a single one hour session to a whole day.

Type of intervention Interventions in which HCPs were trained in methods to promote smoking cessation among their patients.

Outcomes Nine of ten studies reported rates of performance of tasks of smoking cessation by health care professionals including offering counselling, setting quit dates, giving follow-up appointments, distributing self-help materials and recommending nicotine gum. Note: Patient outcomes also reported in all studies, i.e. quit rates.


In the absence of consensus on how to pool cluster level data, the authors decided on a narrative approach to synthesising the data, rather than a formal meta-analysis.

197


Description of included studies (Author/year; country; study design & duration; outcome).

Cohen et al. 1989; U.S.A primary care; RCT, 12 months; 114 HCPs; training- group tutorial 1 hour (advice, quit date, follow up check), and training plus prompt, and training plus nicotine gum, and training plus prompt and nicotine gum; quit rate ‘Q’. Cohen et al. 1989; U.S.A dental practices; RCT, 12 months; 44 HCPs; training- group tutorial 1 hour (advice, quit date, follow up check), and training plus prompt, and training plus nicotine gum, and training plus prompt and nicotine gum; quit rate ‘Q’. Cummings et al. 1989; U.S.A primary care Health Maintenance Organisation; RCT, 12 months; 81 HCPs; training- group tutorial 3X 1 hour (personalised health advice, quit date, one follow up maintenance visit, self help materials, nicotine gum); cessation support ‘C’ & ‘Q’. Cummings et al. 1989; U.S.A primary care private practice; RCT, 12 months; 44 HCPs; training- group tutorial 3x 1 hour (personalised health advice, quit date, one follow-up visit, self help materials, nicotine gum); cessation support ‘C’ & ‘Q’. Kottke et al. 1989; U.S.A primary care private practice; RCT, 12 months; 66 HCPs; training- group tutorial six hours (personalised advice, quit date, nicotine gum, self-help materials, follow up) and, training plus self help materials; cessation support ‘C’. Lennox et al. 1998; Scotland, primary care medical practices; RCT, 14 months; 16 practices, one day training workshop based on stages of change model, Ask ‘A’ & quit rate ‘Q’. Sinclair et al. 1998; U.K., rural community pharmacies; RCT, 62 community pharmacies (including both pharmacists and assistants); training: two hour workshop based on stages of change model; Ask ‘A’ & quit rate ‘Q’. Strecher et al. 1991; U.S.A primary care residency training; RCT, 6 months; 250 HCPs; training- group tutorial one hour (simple advise), and training plus prompt; cessation support ‘C’. Wang et al. 1994; Taiwan, residents and physicians in Family Medicine, RCT, 12 months, 27 physicians, training: two lessons in stages of change model and practice guidelines; quit rate ‘Q’. Wilson et al. 1988; Canada; RCT, 12 months; 83 HCPs; training- group tutorial four hours and training plus nicotine gum and advice; cessation support ‘C’. A (2) Adequate/Reported: The aim/scope or the review was well defined. B (2) Adequate/Reported: Searched the Cochrane Tobacco Addiction Group trials register for studies relating to training. C (2) Adequate/Reported: Inclusion criteria appropriate and un-biased, studies using historical controls were excluded. D (2) Adequate/Reported: The authors reported in detail on the quality assessment methodology. E (2) Adequate/Reported: The results of the individual studies were appropriately summarised by narrative. The characteristics of the included were clearly tabulated. F (2) Adequate/Reported: As a minimum, the authors discussed the issue of cluster effects and its implications for interpreting the results of the review in the discussion section. G (2) Adequate/Reported: Yes.


‡§TOTAL: 14 points; Good. Results (relevant to scope of current review)

Cessation ‘C’: Trained were about 1.5 to 2.5 times more likely to counsel patients about smoking than un-trained HCPs, and to initiate other interventions such as setting a quit date, suggesting a follow-up appointment, and offering self-help materials or nicotine gum. Prompts and reminders to HCPs to deploy smoking cessation techniques were used in addition to a training programme in three trials. In each, the prompts increased the frequency of health professional intervention. Patient Outcomes: Additionally, quit rates were reported in all trials. Of eight studies that compared patient smoking behaviour between trained professionals and controls, six found no effect of intervention. In only one of the three trials of prompts and reminders in addition to a training programme was there a significant improvement in abstinence rates.


– Training health professionals to provide smoking cessation interventions had a measurable effect on professional performance. There was no strong evidence that it changed smoking behaviour. – Reminders are a simple measure and the limited evidence from this review indicates that they increase smoking counselling by HCPs.

198


Reviewers notes – This review up-dates and replaces an earlier version (Silagy, 1994) excluding one paper and adding three post-1990 trials. – A modest, but non-robust, effect on patient outcome was also found, suggesting that training alone is unlikely to be an effective strategy for improving quality of care, unless it is linked to organisational changes that support the intervention. – The largest trial did report a statistically significant improvement in patient outcomes. In this trial, the physicians were paid on a per-item basis for each patient follow-up visit, according to the standard reimbursement scheme in the Canadian health care system. Although in most of the trials some form of follow-up was recommended as part of the training, in none of the others was a specific schedule arranged, and few such visits were scheduled. This may be related to the fact that reimbursement for such visits is generally not forthcoming from third party payers in the United States, where most of the remaining studies took place.


This review is relevant to the study question in that it reports on trials of HCP training (and training plus adjuncts such as manual or computerised reminders) aimed at changing HCPs behaviour to increase smoking cessation counselling and support.




199


Original studies Table 30: Amundson et al. (2003) Citation Amundson, G., Solberg, L. I., Reed, M., Martini, E. M., & Carlson, R. (2003). Paying

for quality improvement: compliance with tobacco cessation guidelines. Jt Comm J Qual Saf, 29(2), 59-65.

Level of evidence * Level III-3 Country U.S.A. Research question/aims

To determine the effect of combined measurement feedback and financial incentives (paid for Ask and Advise rates ≥ 80%) on the rates at which network physicians document that tobacco users are identified (A) and advised to quit (B).

Study type/design Interrupted time series without a parallel control group, duration = 36 months, analysed/eligible = 100%, smoking outcomes [nested within other program outcomes, not reported in this paper].

Participant group Participants: 20 medical groups within the Minnesota HealthPartners network-model health plan. Inclusion criteria: All HealthPartners network medical groups with an enrolment of 2500 or more registered patients; participated in the network for more than two years. Exclusion criteria: None specified. Subject disposition: Health care professionals in clinics or groups of between 16-500 individual Health care professionals per group.

Intervention Feedback (also available publicly on a ‘Consumer Choice’ Website), recognition and financial reward to medical groups that achieve specific high levels of performance on selected measures of care quality (including smoking A&B) as defined by the Outcomes Recognition Program (ORP). There were no additional practice based interventions other than providing ORP manuals. Note: other ORP targets were in place at the time of the trial, specifically cervical and breast cancer screening and paediatric immunisation rates, however, these were not reported in this paper.

Comparator No intervention prior to 1997 compared to 1997; 1998; 1999. Outcome definitions Documentation of Ask and Advise in the patient medical records at the last visit. Data analyses & statistics

Analyses: The unit of analysis is the medical group (n = 20). Differences between mean rates over time were assessed statistically including ANOVAs and t-tests. 5929 patient charts. Sample size calculation: Not reported. A. (2) Adequate/Reported: Not a randomised trial. B. (2) Adequate/Reported: Not a randomised trial. C. (0) Not reported D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: Yes. Baseline data provided for the 12 months prior to the implementation of the intervention (based on 14 489 patient charts) of smoking documentation at the last visit: Ask 49% (range 24% - 78% and Advise 32% (range 0% - 71%). F. (2) Adequate/Reported: All medical groups analysed regardless of whether or not groups or individual practices actually implemented office based changes. G. (1) Inadequate: No medical groups dropped out but it is not known/stated how many health care professionals participated/dropped out of the programme.

Study quality (See below for A-G quality criteria questions†, criteria scores‡ and total/ rating§)

TOTAL: 11 points; Good

200


Table 30: Amundson et al. (2003) (continued) Results (within scope of this review)

1996 (pre) 1997 1998 1999 Change 1996-

1999 and p value

Mean Ask rate

49.7% ± 7%

57% ± 9%

68% ± 7% 73% ± 7% 24% p < .001

Mean Advise rate

32% ± 9% 42% ± 8%

54% ± 13%

53% ± 12%

21% p < .005

The overall 73% Ask rate in 1999 represents a relative increase of 44% compared with baseline performance and the 53% Advise rate in 1999 represents a relative increase of 66% compared with baseline performance

Authors conclusions

Data feedback, combined with a financial incentive, appears to be an effective way for a health plan to improve physician compliance with tobacco treatment guidelines.

Reviewers notes There is no control group, therefore it is possible community levels of A&B were increasing at the time of the trial, however national rates have remained relatively static during the period of study. Further, the authors note that in their previous study (Solberg et al. 2000), this level of sustained change within medical groups required significant leadership engagement and commitment.


Yes. Employed an interrupted time series design to investigate the effect of combined measurement feedback and financial incentives on the rates at which network- physicians document tobacco and provide advice to quit. Note: In 2001, the Advise target was dropped and the Ask target was raised to 95% because of good compliance and in 2002, Assist and Arrange targets were introduced because of widespread evidence that that is where improvement is really needed.

* As per NHMRC Interim Levels of Evidence (NHMRC 2005) for Evaluating Intervention Studies †The quality of the study was assessed using the following questions: (A) Was the assignment to the treatment groups really random? (B) Was the treatment allocation concealed? (C) Were the groups similar at baseline in terms of prognostic factors? (D) Were the eligibility criteria specified? (E) Were the point estimates and measure of variability presented for the primary outcome measure? (F) Did the analysis include an intention-to-treat analysis? (G) Were withdrawals and dropouts completely described?


Abbreviations: AHRQ = Agency for Healthcare Research and Quality; BMI = body mass index; CBA = controlled before-after trial; CCT = controlled clinical trials; CFPC = College of Family Physicians of Canada; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FFS = fee for service; FP = family practitioner; GEEs = generalised estimating equations; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; eLinkS = electronic linkage system; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; ORP = Outcomes Recognition Programme; PAR = practice assessment report; PCTs = Primary Care Trusts; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; THIN = The Health Improvement Network; TS = time-series design; UC = usual care; UMMS = University of Massachusetts Medical School; V.A. = Veterans Affairs clinic setting.

201


Table 31: Bentz et al. (2007) Citation Bentz, C. J., Bayley, K. B., Bonin, K. E., Fleming, L., Hollis, J. F., Hunt, J. S., et al.

(2007). Provider feedback to improve 5A's tobacco cessation in primary care: A cluster randomised clinical trial. Nicotine and Tobacco Research, 9(3), 341-349.

Level of evidence * Level II Country U.S.A. Research question/aims

To evaluate the impact of an electronic health record EHR-generated practice feedback on rates of referral to a state-level tobacco Quitline.

Study type/design RCT: Cluster randomised clinical trial, 12 months follow-up, analysed/eligible = 100%, smoking specific.

Participant group Participants: 279 Health Care Professionals (HCPs) from 19 primary care clinics in Oregon caring for a diverse population of 175,000 patients. Inclusion criteria: Clinic uses ‘Logician’, a standard EHR for storage and retrieval of demographic information, diagnoses, medications, laboratory data, vital signs, and visit records; modified to include an electronic Quitline fax-referral capability. Exclusion criteria: None specified. Subject disposition: 150 medical doctors, 14 registered nurses, 8 nurse practitioners, and 107 medical assistants.

Intervention Physician trainers, blinded to study arm, held 30 minute training sessions for providers and staff in all clinics to review evidence-based cessation strategies and the EHR was modified to include an electronic Quitline fax-referral capability and feedback reports rated provider performance in asking, advising, assessing, and assisting with tobacco cessation, compared with a clinic average and an achievable benchmark of care (defined as the average performance of the top 10% of providers being measured).

Comparator Physician trainers, blinded to study arm, held 30 minute training sessions for providers and staff in all clinics to review evidence-based cessation strategies and the EHR was modified to include an electronic Quitline fax-referral capability but no feedback reports.

Outcome definitions The main outcomes were EHR-documented tobacco-related performance of ask (A), advise, assess (B), and assist, and referral of tobacco users to the Oregon Tobacco Quit Line (C) over a 12-month period (both fax referrals and self-referrals [calling the Quitline after receipt of a brochure] were included in the analysis).


Analyses: Baseline continuous variables analysed using weighted t-tests, and categorical variables were compared using chi-square tests. Analysis of all outcome variables was performed using generalised estimating equations to accommodate cluster-level and several a priori defined individual-level covariates, e.g. clinic volume, size of clinic, insurance mix, baseline “ask” rate, the presence of a clinic champion, and case-mix. Sample size calculation: Sample size estimates were powered to detect changes in ask rates from 30% to 65%, advice rates from 30% to 60%, and assist rates from 10% to 30%, requiring 17.4 clinics with 172,680 patients. A. (2) Adequate/Reported: Cluster randomised. B. (2) Adequate/Reported: Physician trainers blinded to study arm and blinded assessment of EMRs. C. (2) Adequate/Reported: Intervention and control clinics were similar on all patient, provider, and clinic-level variables measured. D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: EHR data collected 6 months prior to the start of the study period and prior to implementation of the EHR tracking system showed a baseline smoking prevalence of 15% and a baseline asking rate of 30%. F. (0) Unknown/Not reported G. (2) Adequate/Reported: No clinics dropped out.



202


Table 31: Bentz et al. (2007) (continued) Results (within scope of this review)

During 12 months of follow-up, EHR-documented rates of advising, assessing, and assisting were significantly improved in the intervention clinics (n = 10) compared with the control clinics (n = 9) (p < .001). A higher case-mix index (sicker patients) and presence of a clinic champion (a staff member who publicly advocated use of the fax-referral process in each clinic) were associated with higher rates of referral to a state-level Quitline. In the un-adjusted analysis, there was no demonstrated impact of feedback on EHR documentation of referral or receipt of calls at the Quitline. However, after adjusting for case mix and the presence of a clinic champion, baseline ask rates, insurance mix, clinic size, and clinic volume, the odds ratio for intervention (feedback) was 1.53 (95% CI 1.05–2.23) for total referral rate (note: the 4% referral rate of smokers within the clinics represents a threefold increase, compared with population-based historical data in Oregon). EMR documentation of cessation procedures

Procedure Control Intervention

Crude Odds ratio (95% CI)

p value

(A) Asked rate 88.1% 94.5% 1.072 (1.168–1.076) .05 (B) Advised rate

52.7% 71.6% 1.36 (1.33–1.39) .001

(C) Assisted rate

10.5% 20.1% 1.92 (1.78–2.08) .001

Authors conclusions

The EHR-generated feedback doubled the rate of documentation of assistance with tobacco cessation. The multivariate model showed that the presence of a clinic champion was an important covariate in predicting connections to the Quitline. Interestingly, the clinic champions in this study were not necessarily physicians; lead nurses often were active in exhorting medical assistants to ask about tobacco use and to follow through with faxing referrals. Future studies need to focus on ways to recruit and reinforce the behaviour of clinic champions.

Reviewers notes Both intervention and control clinics were provided convenient EHR tools for documenting tobacco control activities and all had received training and reinforcement for addressing tobacco. The authors reported that it was simple and convenient to modify the EHR, redesign the clinic workflow, gain buy-in from staff, and incorporate the measurement and feedback procedures into the routine of clinic practice.


Yes, relevant. Employed a moderately large, representative, randomised controlled trial with 12 months follow-up to examine the effectiveness of computer generated practice feedback as an intervention to promote ABC. Both control and intervention groups were exposed to the same provider training and the up-dated EMR referral system, therefore the study focused only on feedback as the intervention component of interest.



Abbreviations: CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FP = family practitioner; GP = General Practitioner; EHR = electronic health record; HCP = health care professional; HMO = Health Maintenance Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; PCTs = Primary Care Trusts; PEI = patient exit interview; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; UC = usual care; V.A. = Veterans Affairs clinic setting.

203


Table 32: Borgiel et al. (1999) Citation Borgiel, A. E. M., & Williams, J. I. (1999). Evaluating the effectiveness of 2 educational

interventions in family practice. CMAJ: Canadian Medical Association Journal, 161(8), 965.

Level of evidence * Level II Country Canada Research question/aims

Aims to assess the effectiveness of two educational interventions for improving the quality of care (including charting of smoking status) provided by family physicians: the Practice Assessment Report (PAR) and the Continuing Medical Education Plan (CMEP) with a follow-up visit by a mentor.

Study type/design Cluster randomised controlled trial: duration = 12 months, analysed/eligible = 54%, smoking specific.

Participant group Participants: 57 physicians (27(control) – 29 (exp)). Inclusion criteria: Had been in family practice for five years or more, had been in the same practice for three years or more, had graduated from medical school after 1958 and working at least half-time in practice. Exclusion criteria: Specialists or full-time academics; had been audited within the previous two years; or had participated in a similar physician review program. Subject disposition: Male 66.1%/female 33.9%; 55.4% certificated physicians; 7.1% College of Family Physicians of Canada (CFPC) members and 37.5% non-members; physicians who graduated in 1970 or later 19.6%, physicians who graduated before 1970 80.4%.

Intervention Mentors provided structured information for identifying relative strengths and weakness in the practice, mentors (who were senior and respected physicians in the community) attended an educational workshop with a follow-up phone call at two or three months to review the physician’s commitments and subsequent educational activities.

Comparator No intervention/ Practice assessment report only. Outcome definitions The ‘charting score’ was based on legibility, organization and documentation of patient

data including smoking status (based on 3 sources: patient questionnaires, chart audits and physician questionnaires [pre/post matched questionnaires/chart audits for 2395 patients]).


Analyses: used unpaired t-tests of mean scores to test the effect of the CMEP relative to the PAR on the quality of practice, and paired t-tests to assess the changes within groups between phase one and phase two. Sample size calculation: With a mean difference of 10 points, a standard deviation (SD) of 12 points, a 2-tailed test with an α of .05 and power set at 80%, the required sample size was 24 physicians per group. A. (2) Adequate/Reported: Randomisation by number list, stratified physicians by sex, age and CFPC status into eight permuted blocks. B. (1) Inadequate: Blinding of assessors not clear, however eight specifically trained people all worked to abstract the randomly selected patient records. C. (2) Adequate/Reported: No significant between group differences with respect to sex, CFPC status, practice location and year of graduation. D. (2) Adequate/Reported E. (0) Unknown/Not reported: Yes for ‘global quality of care’ and ‘charting’ but not for ‘ask’ and the provision of ‘brief advice’. F. (0) Unknown/Not reported G. (0) Unknown/Not reported: Unclear. Three physicians dropped out but the reasons were not stated.


TOTAL: 7 points; Fair. Results (within scope of this review)

The percentage of patients reporting that their physician discussed smoking (ask/brief advice); 79.1 %( control) vs. 83.1 %( exp); 4.0% increase (0.8 – 7.1); n.s. Note: no statistically significant differences between groups were observed on any of the study outcomes.

204


Table 32: Borgiel et al. (1999) (continued) Authors conclusions

The literature supports the use of trained physician-educators or opinion leaders for modifying practices. However, mentors may be more effective in educating physicians about specific areas of practice (in which the leader has expertise) than when offering broad general consultations. Interventions based on global assessments of practice are not a good means for identifying learning needs or targeting practice behaviours for change.

Reviewers notes It would appear unrealistic to expect significant changes in practice following the receipt of a general report or structured feedback.


Yes, relevant. Employed a controlled, randomised trial to examine the effectiveness of educational interventions for improving the quality of care (including charting of smoking status) by family physicians. This study is an example of smoking cessation physician training nested within the breadth and scope of family practice. This study demonstrates the significant limitations of this approach.




205


Table 33: Cockburn et al. (1992) Citation Cockburn, J., Ruth, D., Silagy, C., Dobbin, M., Reid, Y., Scollo, M., et al. (1992).

Randomised trial of three approaches for marketing smoking cessation programmes to Australian general practitioners. Bmj, 304(6828), 691-694.

Level of evidence * Level II Country Australia Research question/aims

To compare three approaches for marketing a quit smoking intervention kit to general practitioners

Study type/design Randomised controlled trial, duration = 4 months, analysed/eligible = 84%, smoking specific.

Participant group Participants: 264 randomly selected general practitioners. Inclusion criteria: All general practitioners from the Melbourne metropolitan region. Exclusion criteria: None stated. Subject disposition: Mean age 47 (years); Mean patients/week 154; Mean years in practice 19, (%) female 11; (%) with postgraduate qualifications 30, (%) full time 94.

Intervention Personal delivery and presentation of a smoking cessation kit by an educational facilitator (visit lasted an average of 12.8 (SD 5.8) minutes) with a follow up visit six weeks later. The educational facilitators made appointment to see general practitioners in the same way as pharmaceutical representatives.

Comparator Compared to: delivery to the receptionist by a friendly volunteer courier (unannounced visits) with a follow up phone call six weeks later, or postal delivery with a follow up letter six weeks later.

Outcome definitions – Measured use of components of the kit at four months follow-up. General practitioners offering simple advice to quit (B), giving the patient a comprehensive self help booklet, a "help" card, cessation advice within a structured framework together with a "contract" card whereby the patient makes a written contract with the general practitioners to quit smoking by a certain date (C). – Doctors’ self-report of delivery of smoking cessation interventions via questionnaire which contained a number of subscales. – Cost.


Analyses: Cronbach's coefficient [alpha] statistic was used as an estimate of the reliability of subscales (> .05 used as adequate reliability). Analyses of variance were used to test for significant differences between groups for linear variables and X2 analyses for categorical variables. A significance level of p < .05 was used. Sample size calculation: Not stated A. (2) Adequate/Reported: General practitioners were randomly selected and randomly allocated to receive one of three approaches for distribution of the smoking intervention programme. B. (1) Inadequate: Not specifically stated. C. (2) Adequate/Reported: No significant differences on any of the variables examined. D. (2) Adequate/Reported: Yes, all. E. (0) Unknown/Not reported: No. F. (2) Adequate/Reported: Yes. G. (2) Adequate/Reported: Six doctors refused to receive the kit, and two refused the follow-up evaluation visit.


TOTAL: 11 points; Good Results (within scope of this review)

The overall return rate of the questionnaire was 79%. There was a trend towards general practitioners in this approach being more likely to have used at least one component ‘A&B’ (p = .07). However, there was a significant difference in the number of contract cards used, general practitioners in the educational facilitator group using significantly more contract cards (6.54; SD 12.59) than general practitioners in the other two groups ‘C’ (courier 3.79; SD 10.03 and Post 1.92; SD 6.33), p = .02.

206


Table 33: Cockburn et al. (1992) (continued) Authors conclusions

Doctors receiving the educational facilitator approach were significantly more likely than those receiving the other two approaches to have used one of the "intensive intervention" components from the kit. There were no significant differences in use of "minimal intervention" components of the kit.

Reviewers notes Given the inability of this study to show a cost effective outcome with the educational facilitator approach, this does not seem to be a realistic option for marketing (individually or ‘one-on-one’) smoking interventions to general practitioners.


Yes, relevant. Employed a randomised trial to examine the effectiveness of academic detailing vs. volunteer courier vs. postal dissemination of a smoking cessation resource kit to promote the ‘ABCs’.




207


Table 34: Coleman (2007) Citation Coleman, T., Lewis, S., Hubbard, R., & Smith, C. (2007). Impact of contractual

financial incentives on the ascertainment and management of smoking in primary care. Addiction, 102(5), 803-808.

Level of evidence * Level 3-III Country U.K. Research question/aims

To investigate the impact of the April 2004 contract for UK general practitioners (GPs) on delivery of primary care smoking cessation interventions.

Study type/design Interrupted time series: Secondary data analysis of computerised, longitudinal general practice database, 14 years pre – 18 months post (The Health Improvement Network (THIN) from routine UK primary care medical records), smoking specific.

Participant group Participants: Family physicians’ (medical practices). Inclusion criteria: All primary care GPs in the THIN network database. Exclusion criteria: Non-participating GPs (practices). Subject disposition: Not defined.

Intervention Introduction of substantial financial rewards for compliance with targets across the spectrum of clinical activity including smoking cessation management (approximately 8% payments related to smoking targets). Specifically, targets were set for the ascertainment of patient’s smoking status ‘A’ and recording the delivery of GPs’ brief smoking cessation advice ‘B’ (but not Cessation support/referral).

Comparator No intervention/pre-test. Outcome definitions Incidence of ascertainment of patient’s smoking status ‘A’.

Incidence of recording the delivery of GPs’ brief smoking cessation advice ‘B’. (Incidence was defined as the number of individuals (or current smokers) with a recorded event between 1 January and 31 December in any one year divided by the number of individuals (smokers) available for follow-up in THIN on 1 July of the same year)


Analyses: Analysed data (using Poisson regression) from 1 607 782 patients' medical records that were held within The Health Improvement Network (THIN) database. Calculated for each year between 1990 and 2005 and for each quarter-year from 2003 the incidence of recording of smoking status in medical records and, in smokers, the receipt of GPs' smoking cessation advice and prescriptions for nicotine addiction treatments and prescriptions for nicotine replacement therapy (NRT). Sample size calculation: Not stated/ Not applicable A. (2) Adequate/Reported: Not a randomised trial. B. (2) Adequate/Reported: Not a randomised trial. C. (2) Adequate/Reported: Yes (all GPs pre-post). D. (2) Adequate/Reported: Yes (all GPs pre-post). E. (2) Adequate/Reported: Yes. Quarterly incidence (per 10 000 patients) of smoking status recording and brief advice given: 2003–2005. F. (2) Adequate/Reported: Not applicable. G. (0) Unknown/Not reported: Not reported.



208


Table 34: Coleman (2007) (continued) Results (within scope of this review)

– Compared to the first quarter of 2003, there was an 88% increase in recording of smoking status (rate ratio 1.88, 95% CI 1.87–1.89) and in brief advice to smokers (RR 3.03, 95% CI 2.98–3.09), up to the first quarter of 2004, which was sustained until the first quarter of 2005. – There was no consistent change in the incidence of receiving nicotine replacement therapy (no targets for NRT prescriptions).

Ask Brief Advice and NRT

Authors conclusions

Study findings suggest that the 2004 GP has had an impact on primary care smoking status ascertainment which is additional to the impact of other factors influencing this since 2000. It remains unclear, however, exactly how much the provision of brief advice to smokers has increased but closely monitored targets coupled with financial incentives do appear to change GPs’ clinical behaviours. It is also unclear whether or not the observed changes will be sustained beyond the first year of the contract (and ongoing analysis is required).

Reviewers notes Some of the observed increase in advice-giving may not reflect more advice being given, but rather, more complete recording of advice that GPs already give. It is difficult to assess the true extent to which advice giving has improved because there are no comparable data sources available.


The main strengths of this study are: its large size, data analysed within quarter-year time bands, and the long duration (1990s-2005).




209


Table 35: DePue et al. (2002) Citation DePue, J. D., Goldstein, M. G., Schilling, A., Reiss, P., Papandonatos, G., Sciamanna,

C., et al. (2002). Dissemination of the AHCPR clinical practice guideline in community health centres. (Research Paper). Tobacco Control, 11(4), 329(327).


To investigate the effectiveness of an academic detailing/educational outreach intervention intended to increase the documentation and delivery of smoking interventions via the development of ‘office systems’; including chart prompts, ancillary staff involvement, physician audits with feedback, skills training, and patient education materials.

Study type/design Pre-test/Post-test; one year follow-up, analysed/eligible = 67%, smoking specific. Participant group Participants: 151 HCPs across 14 sites.

Inclusion criteria: 21 health centre sites in Rhode Island. Exclusion criteria: Clinics engaged in other quality assurance accreditation programmes. Subject disposition: Twenty-seven physicians, 53 nurses, one social worker, 70 other staff (including medical assistants, dieticians, counsellors, receptionists, and dentists). From six large, three medium, and five small sites. Seven sites were urban, four suburban, and three rural.

Intervention Three, one hour on-site provider training sessions, on review of effective tobacco interventions, use of office systems, and tobacco counselling skill building by way of an academic detailing/educational outreach approach.

Comparator Baseline vs. follow-up: HCPs their own controls. Outcome definitions Documentation in the last visit (percentage recording): ‘Ask’ notation of smoking

status. ‘Brief advice’ was considered any notation of advice or discussion about quitting smoking. ‘Cessation’ (reported here as a composite of Assist & Arrange) (based on n = 1591 patient charts).


Analyses: Hierarchical mixed effects logistic regression model with random effects. Confidence intervals were calculated at the 95% level. Sample size calculation: Not clear. A. (2) Adequate/Reported: Not a randomised trial. B. (2) Adequate/Reported: Not a randomised trial however chart review forms were anonymous. C. (2) Adequate/Reported: Yes. HCPs were their own controls. D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: Baseline average provider performance rates for ‘Ask’ 30%; ‘Brief Advise’ 19%; ‘Assist/Arrange follow-up (Cessation) 9%/6%, together with the mean change from baseline to study end. F. (0) Unknown/Not reported G. (2) Adequate/Reported: Yes. One first year and six second year sites dropped out because of preparation for an external accreditation.



Average performance rates increased for ‘Ask’ and ‘Advise’: 30% to 44%, and 19% to 26%, pre-post, respectively. Significant increases were found only for some visit types. – Odds ratios for ‘Ask’ post vs. pre-test: OR 3.00; 95% CI 1.19-7.55 – Odds ratios for ‘Advise’ post vs. pre-test: OR 1.29; 95% CI 0.69-2.43 (n.s.) – Odds ratios for ‘Advise’ yearly physical vs. medical follow-up: OR 2.43; 95% CI 1.53-3.87 – ‘Assist’ and ‘Arrange’ follow up with smokers were not significantly impacted.

210


Table 35: DePue et al. (2002) (continued) Authors conclusions

This guideline dissemination effort resulted in quite different provider counselling rates across patient sexes, and visit types. Guideline implementation may require more sustained efforts, with multiple strategies, which are reinforced at higher policy levels, to more fully integrate tobacco interventions into routine primary care practice with all patients who smoke. Since the premise behind documenting smoking status is to prompt provider advice, it was disappointing that Advise rates did not significantly increase over time.

Reviewers notes Based on the ‘4 A's’: Ask all primary care patients about tobacco use, Advise all smokers to quit at every visit, Assist smokers who are ready to quit, and Arrange follow-up to support the quit process. Thus, ‘Cessation’ is reported here as a composite of Assist & Arrange.


Yes, relevant. Employed a quasi-experimental trial to examine an academic detailing/educational outreach programme intended to facilitate the implementation of clinical practice guidelines.




211


Table 36: Dietrich et al. (1992) Citation Dietrich, A. J., O'Connor, G. T., Keller, A., Carney, P. A., Levy, D., & Whaley, F. S.

(1992). Cancer: improving early detection and prevention. A community practice randomised trial. BMJ, 304(6828), 687-691.


To test the impact of physician education and facilitator assisted office system interventions on cancer early detection and preventive services.

Study type/design Randomised 2 x 2 factorial trial of two interventions alone and in combination, 12 months follow-up, general prevention.

Participant group Participants: Office based family physicians and general internists in 98 practices. Inclusion criteria: All medical practices in New Hampshire and Vermont. Exclusion criteria: Physicians who had been at their current practice site for less than 24 months, who were based at a training site, or anticipated leaving their practice within the next year. Subject disposition: Mean age 42, gender 94% male, family medicine 66%, solo practice 49%.

Intervention (1) Education: A day long, expert-led, group education meeting for physicians discussing smoking cessation advice for 120 minutes and other services for 30 to 60 minutes each. (2) Office system intervention: Assistance from a project facilitator in the design and implementation of office routines that support provision of cancer early detection and preventive services, including team work (division of responsibilities for providing services), preventive care flow sheets, external identifiers for smokers' charts, health education posters and brochures, and patient-held health diaries.

Comparator Education vs. office system intervention vs. no intervention/control. Outcome definitions Incidence of recommendations to quit smoking as assessed by patient exit

questionnaires (20-30 patients/physician). Data analyses & statistics

Analyses: Mean proportions for each service from the practices in each study group were compared by two factor analysis of covariance. Sample size calculation: A priori sample size calculations assumed an α statistic of .05, a β statistic of .20, and baseline rates of service indicated that 25 practices per study group would be sufficient to detect a 30% improvement in most services. A. (1) Inadequate: Stated randomisation but exact method not defined. B. (1) Inadequate: Not reported but implied. C. (2) Adequate/Reported: Characteristics of participating physicians tabulated by study group and differences not significant. D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: Baseline rates of service determined by a regional pilot study. F. (0) Unknown/Not reported G. (1) Inadequate: Dropouts described, but no account of the overall participation rate.


TOTAL: 9 points; Fair

212


Table 36: Dietrich et al. (1992) (continued) Results (within scope of this review)

After 12 months, providers implementing the office system intervention had a non-significant increase in smoking cessation advice 83% (baseline) vs. 84% (follow-up), compared with physicians receiving education only 79% (baseline) vs. 73% (follow-up). More recommendations to quit smoking were provided for smokers cared for by physicians receiving the office system intervention in the absence of the education intervention (p = .017). Outcomes Control Educati

on only Office systems only

Education & Office systems

Odds ratio (95% CI)

Ask ‘A’ – – – Not reported Baseline 73% 79% 83% 78% Advic

e ‘B’ 12 months

67% n.s. 73% n.s.

84% § 80% n.s.

Cessation ‘C’ (any)

– – – Not reported

*Quit rate ‘Q’ – – – Not reported Authors conclusions

Community practices assisted by a facilitator in the development and implementation of an office system can substantially improve provision of cancer early detection and preventive services. Counselling to quit smoking increased for the office system only group but not the combined intervention group, possibly because the education sessions addressed controversies and perhaps raised doubts about some services.

Reviewers notes It is not possible to distinguish the impact of the various components of the office system intervention as different practices tailored the intervention components in diverse ways to address their unique needs.


Yes, relevant. Employed a randomised 2 x 2 factorial trial of education vs. office systems, alone and in combination, to examine their effectiveness to change health professionals’ smoking prevention behaviours.




213


Table 37: Fiore et al. (1995) Citation Fiore, M. C., Jorenby, D. E., Schensky, A. E., Smith, S. S., Bauer, R. R., & Baker, T.

B. (1995). Smoking status as the new vital sign: Effect on assessment and intervention in patients who smoke. Mayo Clinic Proceedings, 70(3), 209-213.


To assess the effect of expanding the vital signs to include smoking status.

Study type/design A prospective before-and-after design, duration = 16 months, smoking specific. Participant group Participants: General Internal Medicine Clinic (GIMC) staff at the University of

Wisconsin Hospitals and Clinics. Inclusion criteria: Not stated. Exclusion criteria: Not stated. Subject disposition: 47 physicians and five nurse practitioners, 28,000 patients made 50,000 visits per year to this clinic during 16-month data collection period (between 1991 and 1993).

Intervention Expansion of the vital signs (blood pressure, pulse, respiratory rate, and temperature) to include smoking status. That is making smoking status part of the vital sign assessment and completed by medical assistants before the consultation with a clinician.

Comparator After expanding the vital signs to include smoking status. Outcome definitions Rate of clinicians/other than clinician or both asking about smoking ‘A’.

Percentage of smokers who reported that their clinician advised them that day to quit smoking ‘B’.


Analyses: Using patient exit interviews, smokers were asked whether the clinician had (1) advised them that day to stop smoking ‘B’, and (2) provided specific advice or suggestions that day on how to stop smoking ‘B’. Sample size calculation: Not stated. A. (2) Adequate/Reported: Not a randomised trial. B. (1) Inadequate: physicians were told the interviewers were conducting a brief patient survey of clinic experience, but they were not informed that the survey specifically addressed the assessment of patient smoking status. C. (2) Adequate/Reported: same group before and after the intervention. D. (0) Not Reported E. (1) Inadequately reported F. (0) Unknown/Not reported G. (0) Unknown/Not reported


TOTAL: 6 points; Fair Results (within scope of this review)

Baseline vs. intervention phase of all patients surveyed (smokers plus non-smokers): (All results were statistically significant at p < .0001): The rate of clinicians asking about smoking doubled from 25.5% (baseline) to 52.6% (during the intervention phase) ‘A’. 45.5% vs. 65.7% reported that they have been asked that day by a staff person other than their clinician (whether they smoked). 57.5% vs. 80.8% reported that they had been asked that day by anyone in the clinic (clinician or other staff person) whether they smoked. Significant increases in the percentage of smokers who reported that their clinician advised them that day to quit smoking ‘B’ (from approximately 49% at baseline to 70% during the intervention; p < .01) and in the percentage who reported that their clinician gave them specific advice that day on how to stop smoking (24% at baseline vs. 43% during the intervention; p < .01) ‘C’.

214


Table 37: Fiore et al. (1995) (continued) Authors conclusions

In this initial study, expansion of the vital signs to include smoking status was associated with significantly increased rates of identification of patients who smoke and intervention to inquire about and discourage cigarette smoking. If these findings prove generalisable, inclusion of smoking status as a vital sign will provide a simple, effective, and low-cost method of intervening clinically with the chief avoidable cause of illness and death in our society – use of tobacco.

Reviewers notes Single clinic site. Used same group as their own control. Used patient self report as the main data collection source. Small sample size. Self-report on smoking status.


Yes, assessed the effectiveness of an intervention to improve the rate of asking about smoking status by clinical staff ‘A’, and advised to quit smoking ‘B’.




215


Table 38: Flocke et al. (2006) Citation Flocke, S. A., Gordon, L. E., & Pomiecko, G. L. (2006). Evaluation of a Community

Health Promotion Resource for Primary Care Practices. American Journal of Preventive Medicine, 30(3), 243-251.


To evaluate the effect of a practice-tailored intervention using a web-based educational/referral resource, a Prescription for Health Pad and practice feedback and planning (teamwork) on the outcomes of clinician discussions of health behaviour change, referral to patient education and community resources.

Study type/design Pre-test/post-test study; duration = 12 months, analysed/eligible = 41%, general prevention.

Participant group Participants: ≈ 50 primary care physicians. Inclusion criteria: All Practices within Cuyahoga County were randomly sampled. Exclusion criteria: Institutional review board–related barriers and incongruence between study timeframe and practice availability to participate. Subject disposition: One solo physician practice, and single-specialty family practices ranging in size from two to 12 clinicians, either urban or suburban settings.

Intervention Multi-component intervention: a web-based health behaviour change resource including a database of community programs and patient education materials; a health behaviour prescription pad; practice feedback and planning (teamwork).

Comparator No intervention/pre-test baseline care. Outcome definitions (1) Rates of discussion of smoking cessation (also, diet, exercise, weight

management). (2) Qualitative field notes about practice routines, culture, and implementation efforts were analysed to enhance understanding of the practice change process.


Analyses: Multilevel generalized linear models, controlled for clustering. Sample size calculation: Seven practices and 50 patients in the intervention and control cohorts per practice, an inter-correlation coefficient (ICC) of 0.04, and a p value of .05 (one-tailed), the study has 80% power to detect an odds ratio of 1.5 when the proportion of a positive outcome is .20. A. (2) Adequate/Reported: Not a randomised trial. B. (2) Adequate/Reported: Not a randomised trial. C. (2) Adequate/Reported: Subjects were their own controls. D. (2) Adequate/Reported: Yes, all primary care practices in a specific geographical area. E. (2) Adequate/Reported: Patient report of health behaviour assistance at baseline as assessed by patient exit interview, 12.4%. F. (2) Adequate/Reported: Not applicable. G. (1) Inadequate: No practices dropped out but it is unclear if any physicians dropped out.



(1) The post-intervention cohort reported higher but not statistically significant rates of discussion of smoking: 12.4%(pre-test) vs. 14.0%(post-test); OR 1.09, 95% CI: 0.94–1.26, p = .25 (2) Evaluation of the qualitative field notes suggests that for six of seven practices, the intervention was not appreciably implemented.

Authors conclusions

The intervention tools show promise for increasing clinician discussion of health behaviours, information referral, and referral to community resources. Additional work to increase practice adoption of the tools is required.

Reviewers notes Low participation rates and poor implementation of the intervention components appears congruent with similar studies. That is, the theme of poor implementation even under un-blinded trial conditions with volunteer health care professionals.

216


Table 38: Flocke et al. (2006) (continued) Relevance to study question

Yes, relevant. Employed a pre-post-test trial to examine the organisational support mechanisms of web-based educational/referral; a prescription pad and practice feedback and planning (teamwork) intended to facilitate the delivery of smoking cessation advice.




217


Table 39: Goldstein et al. (2003) Citation Goldstein, M. G., Niaura, R., Willey, C., Kazura, A., Rakowski, W., DePue, J., et al.

(2003). An academic detailing intervention to disseminate physician-delivered smoking cessation counseling: smoking cessation outcomes of the Physicians Counseling Smokers Project. Preventive Medicine: An International Journal Devoted to Practice and Theory, 36(2), 185-196.


The study hypothesis is that physicians in the experimental group will be significantly more likely to adopt and deliver smoking cessation counselling than physicians in the control group and that this would also be reflected in higher levels of smoking cessation at a population level.

Study type/design Quasi-experimental; non-randomised clinical trial; 24 months, analysed/eligible = 80%, smoking specific.

Participant group Participants: Community-based sample of physicians: 259 primary care physicians (88 intervention group vs. 171 control groups) and 4295 adult smokers. Inclusion criteria: Eligible if practiced a primary care specialty (i.e., family medicine, internal medicine, obstetrics/gynaecology), provided primary care for at least 25% of their patients, completed postgraduate training, and planned to practice in the state for the three study years. Exclusion criteria: Full-time hospital-based physicians. Subject disposition: Not fully defined but matched on gender, specialty, and years since graduation from medical school.

Intervention Academic detailing: Multi-component office-based intervention incorporated assessment of physicians’ baseline knowledge and attitudes, identifying clear goals, use of supportive written materials, office-based outreach educational component.

Comparator No intervention/different geographic area/no physician-based intervention but simultaneous delivery of a home-based smoking cessation intervention developed by a group of collaborating investigators.

Outcome definitions The proportion of patients who reported that their physician talked to them about their smoking habit (A), or advised them to quit smoking (B), and/or arranged a follow-up visit to discuss smoking (C), at 12, 18, or 24-month follow-up interview.


Analyses: Multivariate regression analyses were conducted to determine the impact of the intervention on self-reported smoking quit rates, reported by adjusted odds ratios. Controlling for potential confounding variables including age, sex, educational level, health status, and stage of change for smoking cessation at baseline. Sample size calculation: Not reported. A. (2) Adequate/Reported: Not a randomised trial. B. (2) Adequate/Reported: Patients and assessors blind to allocation. C. (1) Inadequate: Not fully reported. D. (1) Inadequate: Inclusion/exclusion criteria were reasonably defined. E. (2) Adequate/Reported: Yes. Baseline rates of ‘A’, ‘B’, ‘C’ and quit rates. F. (1) Inadequate: No. G. (0) Unknown/Not reported: No.



RESULTS 12 months 18 months 24 months Outcome ‘A’ ‘B’ ‘C’ ‘A’ ‘B’ ‘C’ ‘A’ ‘B’ ‘C’ Quit Control 54.7

% 56.7 % 17.8

% 54.0 %

55.9 %

26.8 %

52.0 %

52.2 % 8.9 %

Experimental

54.2 %

54.0 % 17.2 %

61.2 %

60.9 %

35.3 %

53.2 %

52.8 % 10.1 %

OR = 1.35; 95% CI .99-1.83; p = .057

Results (within scope of this review)

– Self-reported ‘A’; ‘B’; ‘C’; at the 18-month assessment, there was a borderline significant difference in patient reports of cessation support (p = .06). – Self-reported smoking quit rates: there was a borderline significant effect for those residing in intervention areas vs. control areas (OR = 1.35; 95% CI .99–1.83; p = .057)

218


Table 39: Goldstein et al. (2003) (continued) Authors conclusions

When considered together, the data suggest that the academic detailing office-based intervention enhanced both the frequency and the effectiveness of primary care physicians’ delivery of smoking cessation interventions.

Reviewers notes Although the percentage of patients who reported that their physician addressed smoking was similar between control and experimental groups, the percentage that quit after receiving counselling differed between the groups (18% control vs. 30% exp; p = .03). Suggests an increase in the effectiveness of the smoking cessation interventions actually delivered to individual patients.


Yes, relevant. Employed a large representative controlled, non-randomised 24 month trial to examine the effectiveness of academic detailing as an intervention to promote ABC.




219


Table 40: Hogg et al. (2008) Citation Hogg, W., Lemelin, J., Graham, I. D., Grimshaw, J., Martin, C., Moore, L., et al.

(2008). Improving prevention in primary care: Evaluating the effectiveness of outreach facilitation. Family Practice, 25(1), 40-48.


To evaluate whether a comprehensive preventive intervention program using outreach facilitators improves preventive care delivery.

Study type/design RCT (Match-paired, cluster-randomised controlled trial), duration = 12 months, analysed/eligible = 15%, general prevention.

Participant group Participants: ≈160 physicians from 54 volunteer primary care practices (3049 patients). Inclusion criteria: Fee-for-service primary care practices in Eastern Ontario, Canada; practices comprising no more than six physicians. Exclusion criteria: Practices comprising more than six physicians and those with an academic affiliation. Subject disposition: Mean number of physicians per group practice 3.05; mean number of patients seen per half day 16.35; percentage of rural practices 31.4.

Intervention Duration 11.5 months; two nurses assigned 13 or 14 practices each, outreach visits to each practice on average once a month for an average of 46 minutes. Visits include: baseline audit, feedback and consensus building (the facilitator explored with the practice ways of integrating preventive care into routine/episodic care visits and shared tools used for improving the quality of preventive care delivery, such as prevention flow sheets, chart flags, sticker reminders, electronic reminders, patient care records), periodic follow-up and consensus building every three to six weeks.

Comparator No intervention Outcome definitions Chart review: The primary outcome measure was the composite index of preventive

performance, defined as the number of appropriate preventive manoeuvres done minus the number of inappropriate manoeuvres done, divided by the total number of eligible preventive manoeuvres (including the prescription of nicotine replacement therapy) as detected through a chart review (a total of 14 preventive manoeuvres). Patient surveys: The mean proportion of eligible patients who received preventive counselling on manoeuvres related to healthy lifestyle (including asking about smoking status and providing smoking cessation counselling) as determined from patient surveys (a total of 14 preventive manoeuvres).


Analyses: Unit of randomisation and analysis was the practice; means, SDs and 95% confidence intervals computed for continuous variables and frequency distributions for categorical variables; chi-square tests and paired t-tests. Sample size calculation: Twenty-four practices were needed for detecting a 10% change in care frequently, an alpha of .05 and 83% power. A. (2) Adequate/Reported: Match-paired, cluster-randomised by external administrator. B. (2) Adequate/Reported: Double blinding of physicians and facilitators to the targeted manoeuvres. C. (2) Adequate/Reported: Intervention and control practices equivalent across seven of the nine characteristics used for matching. D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: Yes two month baseline data collected for all primary outcomes via independent chart review. F. (2) Adequate/Reported: Yes. G. (2) Adequate/Reported: No practices withdrew from the study, recruitment process defined.



220


Table 40: Hogg et al. (2008) (continued) Results (within scope of this review)

Chart review: No difference was detected between the trial’s arms for the primary outcome overall prevention index for prescribing manoeuvres, 2.0% (95% CI: -3.2–7.3) p = .44, including the prescription of nicotine replacement therapy 20.4% (exp) vs. 13.5% (control), mean difference 6.9% (95% CI: -2.7–16.5) p = .16. Patient surveys: A small significant difference between the arms was detected for the secondary outcome’s overall prevention index, 2.8% (95% CI: 0.7–4.8) p = .01. The composite secondary outcome included: Asking about smoking status 35.0% (exp) vs. 33.31% (control), mean difference 1.7% (95% CI: -5.6–8.9) p = .64, and providing smoking cessation counselling 41.9 (exp) vs. 38.1% (control), mean difference 3.7% (95% CI: -3.6–11.0) p = .3.

Authors conclusions

In contrast to similar facilitation trials, this outreach facilitation program did not produce improvements in the delivery of preventive care. This lack of effect may be due to differences in the intervention and context, or the practice’s limited capacity to change. The intervention simultaneously facilitated a high number of manoeuvres, blinded facilitators and physicians to the targeted tests, and had a relatively short intervention period and large number of practices assigned per facilitator, and thus, the facilitators produced a weak intervention – probably explaining the trial’s lack of effect.

Reviewers notes This study essentially replicates a previous study by the authors (Lemelin et al. 2001) yet in a less intensive form (see comparison below) implemented to see if the cost effectiveness of the intervention could be improved. Also, the authors focused on fee-for-service (FFS) practices, and increased the number of practices assigned per facilitator, shortened facilitator training, reduced the intervention length, and targeted a greater number of preventive manoeuvres. Also, the authors note the very low 15% recruitment rate. Lemelin et al. (significant 11% increase over 18 months) vs. Hogg et al. (n.s) Average number of practices per facilitator 7.7 vs.13.5 Length of facilitator training 30 weeks vs. 4 weeks Average number of visits to practices per facilitator 33 vs. 9 Intervention length 18 months vs. 11.5 months Number of preventive manoeuvres facilitated 13 vs. 54 Number of manoeuvres included in outcome measure 13 vs. 25 Practice setting Capitation vs. Fee for Service


Compared to other service models fee-for-services (FFS) practices may have different priorities, particularly in conditions of physician scarcity, in addressing access to care. The generalisability of this study to New Zealand’s capitated/co-payment setting is not known. The New Zealand context perhaps sits between the Lemelin et al. (2001) and Hogg et al. (2008) trial contexts.




221


Table 41: Joseph et al. (2004) Citation Joseph, A. M., Arikian, N. J., An, L. C., Nugent, S. M., Sloan, R. J., & Pieper, C. F.

(2004). Results of a Randomized Controlled Trial of Intervention to Implement Smoking Guidelines in Veterans Affairs Medical Centers: Increased Use of Medications Without Cessation Benefit. Medical Care, 42(11), 1100-1110.


Aimed to test the effect of modest intensity, practical systems changes that might increase the delivery of smoking cessation treatment.

Study type/design Cluster randomised controlled trial; 12 months; randomised 20 Veterans Affairs medical centres to intervention or control conditions (approximately 200 HCPs), analysed/eligible = 25%, smoking specific.

Participant group Participants: All physicians, nurses, psychologists and pharmacists of the included 20 Veterans Affairs medical centres (data from their 5678 patients). Inclusion criteria: Existence of referral-based smoking cessation programme, information resource management capacity for data collection, evidence of commitment to the programme. Exclusion criteria: Facilities serving predominantly psychiatric patients, language/communication barriers, incorrect computerised data/computer systems. Subject disposition: Providers: on average (full time equivalents), 21.9 physicians; 5.9 nurse practitioners/physician assistant; 11.2 nurses; 1.5 pharmacists, per 1000 primary care patient. Patients: predominantly white, unemployed, low income (no significant differences between groups).

Intervention Organisational support: two day off-site training of a facilitator and one key advocate per site, by the principal investigator/interventionist. Two-to- three day on-site visit per site. This ‘academic detailing’ aimed to identify barriers and prioritise goals specific to each site and to design strategies to increase delivery of smoking intervention to all smokers. Recommended strategies included: (1) the use of non-physician personnel for documentation, (2) use of electronic clinical reminder systems, (3) adaptation of note templates to include smoking status, (4) increasing the availability of smoking cessation medications, (5) improving the identification of smoking status in the medical records.

Comparator No-intervention/printed smoking cessation guidelines only. Outcome definitions ‘A’ Recorded smoking status “Do you smoke cigarettes now? yes/no, ‘B’ if the record

indicated that cessation was discussed or the provision of advice or medication, ‘C’ provision of counselling or referral to a smoking cessation clinic.


Analyses: Data regarding smoking status, delivery of treatment, medication use (and smoking cessation) were collected from patient exit surveys and medical record review. For binary variables, McNemar odds on change, Pearson x2; repeated logistic regression. Sample size calculation: With alpha set to 0.05, power 80% to detect odds ratio of 1.48 for ‘asked’ and 2.3 for ‘advise’ with ~280 smokers/group. A. (1) Inadequate: Stated block randomisation but incomplete detail of method. B. (2) Adequate: Unlikely that effective change mechanisms could have been instituted at control clinic sites simply listening to accounts of, or observation of, newly implemented change activities in intervention sites. C. (2) Adequate/Reported: The potential influence of several clinic-level covariates including the proportion of low-income patients, patients’ gender distribution, and clinic size (number of providers, average number of patient visits) was controlled by forming groups of similar clinics and their subsequent block randomisation. D. (2) Adequate: Inclusion and exclusion criteria were defined. E. (2) Adequate: Percentage of patients’ tobacco use status identified at last visit; percentage of smokers who received advice to quit at last visit; percentage of smokers who were offered assistance to quit at last visit, and ‘p’ values. F. (1) Inadequate: Three sites closed/merged and excluded, others analysed as allocated. G. (2) Adequate: The number of withdrawals and reason for withdrawal were stated.


TOTAL: 12 points; Good.

222


Table 41: Joseph et al. (2004) (continued) Results (within scope of this review)

The intervention did not increase participant report of being asked about smoking status or receipt of counselling. It did increase the rate of identification of smoking status in the medical record (p = .0001) and increased documentation of advice but did not reach statistical significance. The intervention did not increase the rate of counselling to stop smoking. Site-level data showed an increase in the number of patients receiving smoking cessation medications but this did not reach significance. Medical record review: Ask: baseline 63.1 %(control)/55.7 %(exp) vs. post-test 60.7%(control)/67.0%(exp); p = .0001. Brief advise: baseline 68%(control) /49%(exp) vs. post-test 69%(control)/ 61%(exp); p = .11 n.s. Cessation support (mean number of patients/1000 receiving medication): baseline 34.4control)/18.73(exp) vs. post-test 27.94(control)/31.26(exp); p = .07 n.s. Patient survey data:Ask: baseline 70.8%(control) /71.6%(exp) vs. post-test 74.3%(control)/76.0%(exp); p = .71 n.s. Brief advice: baseline 70.5%(control)/69.7%(exp) vs. post-test 71.8%(control)/73.9%(exp); p = 0.60 n.s. Cessation support: baseline 53.0%(control)/51.7%(exp) vs. post-test 56.4%(control)/61.2%(exp); p = .32 n.s. Patient survey data, smokers cohort, 12 month follow-up: Smokers in the intervention group were slightly but not statistically significantly more likely to be asked their smoking status ‘A’ at one year following as compared to control group smokers: OR 1.12; 95% CI 0.55-2.30, p = .77.

Authors conclusions

The authors concluded that although smoking cessation promotion activities increased post-intervention, effects on cessation rates are diluted by the relatively low proportion of smokers who seek those treatments. A longer implementation period may have increased the effectiveness of the intervention. Improvements in smoking cessation processes and rates are likely to require more intensive intervention in this population.

Reviewers notes The intervention did not improve smoking cessation rates (13.2%(control) vs. 11.4%(exp); p = .51). Patient report vs. documentation of ‘ABC’ differs. The authors suggest this may be explained by a greater likelihood that smokers recall receipt of counselling vs. the likelihood that providers record counselling activity. Note: The intervention, ‘academic detailing’, only recommended the various strategies for implementation. Individual sites then proceeded to implement these strategies (or not) to varying degrees and in different combinations. Implementation was reported as ranging from ‘moderate to significant progress’ towards these goals. Thus, the exact degree of implementation of the recommended strategies at the clinic level is not known.


Yes, relevant. Employed a controlled, randomised trial to examine organisational support mechanisms intended to facilitate the implementation of clinical practice guidelines recommending screening and treatment of all tobacco users.




223


Table 42: Katz et al. (2004) Citation Katz, D. A., Brown, R. B., Muehlenbruch, D. R., Fiore, M. C., & Baker, T. B. (2004).

Implementing Guidelines for Smoking Cessation: Comparing the Efforts of Nurses and Medical Assistants. American Journal of Preventive Medicine, 27(5), 411-416.


To test the effectiveness of implementing The Agency for Healthcare Research and Quality (AHRQ) Smoking Cessation Clinical Practice Guideline at primary care clinics

Study type/design RCT, duration ≈ 24 months across two-study period; baseline period (12 month; from June 16, 1999 to June 20, 2000), the intervention period (from June 21, 2000, to May 3, 2001), smoking specific.

Participant group Participants: Eight community-based clinics (six family practice clinics and two internal medicine clinics). Inclusion criteria: Community-based clinics staffed by at least five full-time physicians or mid-level clinicians (physician’s assistants or nurse practitioners), assigned intake clinicians to specific physicians or mid-level clinicians (with minimal crossover), did not have an on-site, nurse-based smoking cessation program during the study period, had not recently participated in a smoking cessation trial or prevention trial that addressed smoking cessation (within two years prior to the start date of the trial), did not have a residency training program, and were located within a 60-mile radius of Madison. Exclusion criteria: Not stated. Subject disposition: Eight community-based clinics (six family practice clinics, two internal medicine clinics) in southern Wisconsin Intervention (test sites, n = 4 clinics with 43 intake clinicians) or usual care (control sites, n = 4 clinics with 32 intake clinicians) (Table 2 from the paper). Any clinicians refer to intake clinicians or primary care clinicians (including physicians, nurse practitioners, and physician’s assistants). Intake clinicians included registered nurses, licensed practical nurses, and medical assistants. At all clinics, 2163 consecutive adult patients (at least 18 years old) who smoked an average of one or more cigarettes per day into the study across two study periods.

Intervention A guideline-based intervention (multimodality intervention to implement the AHRQ Smoking Cessation Guideline). The AHRQ Smoking Cessation Clinical Practice Guideline recommends that all clinicians strongly advise their patients who use tobacco to quit. The guideline algorithm for smoking cessation brief assessment and counselling, used in training intake clinicians to implement the key recommendations for the AHRQ Smoking Cessation Guideline at the time of each visit. The intervention includes five components: a tutorial for intake clinicians, group and individual performance feed-back for intake clinicians, use of a modified vital signs stamp, an offer of free nicotine replacement therapy, and proactive telephone counselling.

Comparator No intervention. Staff at control sites received only general information about the AHRQ Guideline.

Outcome definitions Performance on guideline recommendation. Study personnel performed face-to-face exit interviews of all eligible patients immediately after their office visit to assess how well the clinic staff had performed Guideline-recommended activities. Telephone interviews were conducted also after two and six months following the exit interview (i.e. date of enrolment in the study). Primary outcomes: Performance of recommended smoking cessation activities by clinic staff (intake clinicians or primary care clinicians); Abstinence at two and six months after the initial clinic visit (7-day point prevalence defined as the proportion of patients who reported abstinence over the previous seven days; and continuous abstinence (defined as self-reported abstinence at both the two and six-month interviews).

224


Table 42: Katz et al. (2004) (continued) Data analyses & statistics

Analyses: Intake clinicians were surveyed immediately before and after the intervention period. Individual patients were grouped by individual intake clinicians (who were grouped by clinic), constructed three-level hierarchical logistic regression models of performance and cessation outcomes across the test and control sites combined. Statistical analyses performed using Stata, version 7.0 software. All tests were two-sided, and a p value less than or equal to .05 was considered to indicate statistical significance. Sample size calculation: The projected sample size was 1200 smokers per comparison period (i.e., eight clinics x 10 intake clinicians/clinic x 15 smokers/intake clinician) or 2400 smokers for the entire study. The trial was designed to have 80% power, with an alpha error of .05, to detect an intervention effect size of 0.4 for the six-month quit rate. A. (2) Adequate/Reported: Used a random number generator to randomly assign each clinic to receive either the intervention. B. (1) Inadequate: Study personnel were blinded to the treatment group assignment. C. (2) Adequate/Reported D. (2) Adequate/Reported E. (2) Reported F. (0) Not reported G. (1) Inadequate: Data were missing for those intake clinicians who were not working at the start of the intervention period, who were employed as “floating” (or locum tenens) staff and did not attend project meetings, or who refused to complete the staff survey. Overall survey completion rate was 84%.



Performance of recommended smoking cessation activities by clinic staff (A and C): Baseline: Majority of patients at both sites were asked by clinic staff about smoking status during office visit, only about one in four were asked about willingness to quit smoking by staff from each site. Intervention period: patients asked about smoking status (A) at test sites were more than at control sites (87% v s 67%; adjusted OR = 3.1, 95% CI 1.2 to 8.2; p < .001), or about their willingness to quit smoking (73% v s 30%; adjusted OR = 6.4, 95% CI 3.7 to 10.8; p < .001), were given literature about quitting (C) (38% v s 3%; adjusted OR = 21, 95% CI 8.8 to 49; p < .001), were assisted with setting a quit date (C) (27% v s 1%; adjusted OR = 33, 95% CI 11 to 100; p < .001) or were engaged about in a discussion about pharmacotherapy, which included nicotine replacement or bupropion therapy (C) (39% v s 14%; adjusted OR = 3.9, 95% CI 2.5 to 6.3; p < .001). Although the overall % of patients who were advised to quit smoking (B) by any clinic staff was not statistically significantly higher at test sites than at control sites (47% vs. 38%; adjusted OR = 1.3, 95% CI 0.8 to 2.3; p = .29), the % of patients who were advised to quit smoking during the intervention by intake clinicians was markedly, and statistically significantly, higher at test sites than at control sites (31% vs. 10%; adjusted OR=5.1, 95% CI 2.7 to 9.7; p < .001). Abstinence rates: Baseline period: No statistically significant differences between two groups. Intervention period: Abstinence was more reported at the intervention site than control sites: At 2-month follow-up assessment: 16.4% v s 5.8%; adjusted OR = 3.3, 95% CI 1.9 to 5.6; p < .001 At 6-month follow-up assessment: 15.4% v s 9.8%; adjusted OR = 1.7, 95% CI 1.2 to 2.6; p = .009 Continuous abstinence at both 2 and 6 months: 10.9% v s 3.8%; adjusted OR = 3.4, 95% CI 1.8 to 6.3; p < .001

225


Table 42: Katz et al. (2004) (continued) Authors conclusions

Results suggest that intake clinicians can effectively implement the AHRQ Guideline-based strategy to improve the delivery of smoking cessation advice and pharmacotherapy in a time- and cost-efficient manner. Implementation of a guideline-based smoking cessation intervention by intake clinicians in primary care is associated with higher abstinence among smokers.

Reviewers notes One of the authors (Fiore) has served as a consultant for, has given lectures sponsored by, or has conducted research sponsored by GlaxoSmithKline (Research Triangle Park, NC, who supplied the nicotine patches) and was appointed by the University of Wisconsin to a named Chair made possible by an unrestricted gift to the university from GlaxoSmithKline.


Yes, for the A and C parts of the ABC.




226


Table 43: Kottke et al. (1992) Citation Kottke, T. E., Solberg, L. I., Brekke, M. L., Conn, S. A., Maxwell, P., & Brekke, M. J.

(1992). A controlled trial to integrate smoking cessation advice into primary care practice: Doctors Helping Smokers, Round III. J Fam Pract, 34(6), 701-708.


To test whether on-site recruitment, training, and organisational assistance in incorporating a smoking intervention system of documented efficacy into non-volunteer primary care practices would result in higher rates of smoking cessation advice to patients.

Study type/design Non-randomised multi-site clinical trial: 29 sites; 18 months follow-up, analysed/eligible = 86%, smoking specific.

Participant group Participants: Non-volunteer primary care clinics (clinics the unit of analysis [8control) – 10 (exp)]) and included 142 primary care physicians and other support staff). Inclusion criteria: Clinics identified within an independent practice association-type managed health care plan in the Minnesota area. Exclusion criteria: Other clinics. Subject disposition: Not clear (mean number of physicians per clinic 8.5 ± 5.2).

Intervention A clinic-wide intervention system to support physicians: responsibility for identifying smokers, assessing their habits, advising them to quit, negotiating an action, and following up on those who want to quit was widely dispersed among all personnel who come in contact with patients (approximately 30% nurse clinician, 10% of a secretary's time, and 5% of a physician’s time). The clinic support program is tailored to the unique attributes of each clinic. Intervention based on the “social learning model of consultation” (Brown & Schulte 1887). Intervention included on average 5.7 site visits, 24 telephone calls, and 5.6 mailings per practice site over the 18-month period and provided information about a feasible and effective system, help with problem solving, assessment of program progress, and feedback and reinforcement.

Comparator No intervention at control clinics. Outcome definitions Rates of smoking cessation ‘asking’ and ‘advice’ to patients (who had visited their

clinic in the previous six months) as measured by 24 question pre-intervention and post-intervention patient surveys (n = 507(control); n = 466(exp)).


Analyses: The unit of analysis was the clinic. Two tailed t-test. Sample size calculation: Not stated. A. (2) Adequate. Not a randomised trial. B. (2) Adequate. (within the context of the study design) Research assistant blinded to clinic allocation for outcome assessment. C. (1) Inadequate. Groups not well defined, but no group differences in the number of patients per clinic (p = .24); number of sites per clinic (p = .24); number of physicians per clinic (p = .37); patient visits (p = .41) D. (1) Inadequate. Inclusion and exclusion criteria were not well defined. E. (1) Inadequate. Baseline percentages for rates of smoking cessation ‘asking’ and ‘advice’ to patients: i.e. ‘untrained’ GPs’ smoking cessation intervention behaviour provided for pre-intervention period and post-intervention via patient survey for control and intervention groups. F. (2) Adequate. Yes. G. (2) Adequate. Yes. Two clinics (control) and one clinic (exp) dropped out due to a statistically small number of respondents. Analysis based on remaining 18 clinics.


TOTAL: 11 points; Good. Results (within scope of this review)

Ask: baseline 21.9% (± 9.6%) (control) /22.9% (±11.2%) (exp) [p = .84] vs. post-test 26.0% (±12.2%) (control)/39.8% (±12.3%)(exp); p < .05 Brief advice: post-test 26.4% (±14.6%) (control)/40.5% (±12.1%) (exp); p < .05. Cessation: post-test 13.4% (±11.1%) (control)/22.8% (±11.5%) (exp); p < .1 (n.s.)

227


Table 43: Kottke et al. (1992) (continued) Authors conclusions

The intervention takes advantage of repeated face-to-face contacts to recruit and train clinic personnel to implement a systematic smoking intervention program. The intervention significantly increased the rates at which a population of primary care clinics identified their patients who smoked and advised them to quit smoking. However, the authors comment that for any smoking intervention program to persist there must be both a demand for services by patients and payment from some source outside the clinic. There is little reason to believe that attempts to change physician behaviour that do not deal with the practice environment can be successful.

Reviewers notes This study highlights that in a non-volunteer sample, not all clinics are pro-active in smoking cessation interventions. The authors reported that between 40% and 50% of the 29 clinic sites showed very little evidence of implementing a systematic program at two years follow-up. Despite this low rate of implementation, the intervention still demonstrated a statistically significant change in provider behaviour.


Yes, relevant. Employed a controlled non-randomised trial to examine the efficacy of a multi-faceted intervention to promote ABC, importantly, in a non-volunteer population.




228


Table 44: Krist et al. (2008) Citation Krist, A. H., Woolf, S. H., Frazier, C. O., Johnson, R. E., Rothemich, S. F., Wilson, D.

B., et al. (2008). An Electronic Linkage System for Health Behavior Counseling. Effect on Delivery of the 5A's. American Journal of Preventive Medicine, 35(5 SUPPL.), S350-S358.


To evaluate the effects of an electronic linkage system (eLinkS) [instituted to promote health behaviour counselling and to automate patient referrals to community counselling services] on counselling rates.

Study type/design Pre-test/post test: two year-pre and one year-post, analysed/eligible = 87%, general prevention.

Participant group Participants: Nine primary care practices in the Tidewater region of Virginia, 48 clinicians, a total of 5679 patients. Inclusion criteria: Not stated. Exclusion criteria: Not stated. Subject disposition: 23 physicians, 21 residents, 4 mid-level providers, average age 37 years (26-62), 52% women, median years in practice 7.5 (2-31), average number of full-time equivalent in the office 6.3 (0.9-8), 79% clinician only complete use of eLinkS, 21% team approach.

Intervention Using the electronic medical record (EMR) as a platform, eLinkS was designed. It reflects early input from clinicians at study sites, when staff who take patients to examination rooms and obtained vital signs and entered them into medical record the EMR displayed a screen prompt to remind the staff to enter tobacco-use status (as well as height and weight). One-hour training session held at all practices prior to launch. Exposure period was five weeks.

Comparator Current practice (Pre launch). Outcome definitions Delivery of 5As measured by clinicians’ entries in the EMR dialogue box. 5As (A1 ask,

Ad address, A2 advice, A3 assess, A4 agree; and A5 arrange) for health behaviour counselling as recorded by eLinkS for all adult patients seen in the study practices during the exposure period (April 16, 2006 to May 22, 2006).


Analyses: SAS® used for statistical calculation, differences in percentages tested for significance using Fisher’ exact test, logistic regression for each of the 5As. Initial regressor variables those that could theoretically influence eLinkS usage: the patient’s age, gender, health behaviours, co-morbidities, encounter type and complexity, and the physician’s age, gender, training, years in practice, and practice site p ≤ .25 as the entry criterion and p ≤ .10 as the retention criterion. Patients were the unit of analysis, practices selected purposively, and fixed-practice efforts were considered. Sample size calculation: Not indicated. A. (2) Adequate/Reported: Not a randomised trial. B. (2) Adequate/Reported: Not a randomised trial. C. (2) Adequate/Reported: All clinicians (pre and post). D. (2) Adequate/Reported: All clinicians came from the same geographical area, practices share common type of electronic medical record that is managed by a central informatics staff, and have used the EMR for three to ten years, however, 44% were family medicine residents (that is, younger) and comprises more women than non-residents. E. (1) Inadequate: Reported on complete use of eLinkS. F. (0) Not reported G. (1) Inadequate



229


Table 44: Krist et al. (2008) (continued) Results (within scope of this review)

Exposure to eLinkS and delivery of the 5As to the study population: A1 (BMI and tobacco use status assessed) 2117 patients (37.2%), tobacco use only assessed 358 patients (16.9%). Data were missing more often for tobacco-use status than for BMI. Address (Ad): eLinkS prompts appeared for 1860 patients, 1801 occurring automatically and 59 initiated by the clinician. When confronted with these prompts, clinicians elected to open the dialogue box form (address the behaviour) for 912 (49%) patients. Advise (A2). Clinicians advised 537 patients to change their health behaviour, 13% of patients with an unhealthy behaviour (n = 4030) and 29% of patients for whom a prompt appeared (n = 1860). A2 was documented for 20% of smokers, 7% of overweight patients, 17% of obese patients, and 13% of problem drinkers. Agree (A3). Clinicians recorded that 461 patients were ready to improve the targeted behaviour, and 441 were engaged in what to do next. This represents 86% and 82%, respectively, of patients who received clinician advice (n = 537). Assist (A4). Fully 407 patients were referred for intensive counselling, 10% of patients with an unhealthy behaviour or 76% of patients who received A2. The population that received A4 included 12% of obese patients, 3% of overweight patients, and 8% of smokers; no risky drinkers were referred for intensive counselling. Two-thirds of patients who sought intensive counselling chose group counselling as the venue, primarily Weight Watchers. Usual care was requested by 176 patients. Arrange (A5). Arrangements for follow-up with the practice were made for 306 patients. Office visits, or nurses’ telephone calls to monitor progress, were the preferred arrangements.

Authors conclusions

The intervention increased the rate at which patients were referred for intensive behavioural counselling compared to current practice norms. Given the evidence that intensive counselling is more effective in promoting behaviour change, implementing eLinkS could have substantial public health benefits.

Reviewers notes The intervention went live on April 16, 2006, and was discontinued five weeks later (May 22, 2006) when an overwhelming surge in referrals for intensive counselling exceeded available funds. Patients referred to intensive counselling during the exposure period were eligible to receive services for up to nine months (through February 2007). Authors stated that the participating clinicians were not representative of primary care clinicians generally, because large percentage (44%) was family medicine residents, a group that is younger and comprises more women than non-residents. Intensive counselling was free to patients, eliminating cost as a barrier so may have enhanced the referral rate.


Yes, relevant as it assessed an intervention to promote change in behavioural counselling of primary health care providers.




230


Table 45: Lemelin et al. (2001) Citation Lemelin, J., Hogg, W., & Baskerville, N. (2001). Evidence to action: a tailored

multifaceted approach to changing family physician practice patterns and improving preventive care. CMAJ: Canadian Medical Association Journal, 164(6), 757.


To describe a randomised controlled field trial of a tailored multifaceted intervention to improve preventive care in capitation-based family practices. The multifaceted intervention is delivered by nurses trained in prevention facilitation to improve prevention in primary care.

Study type/design RCT, duration = 18 months, analysed/eligible = 48.4%, general prevention. Participant group Participants: Health services organisations (HSOs) (community primary care practices

that have a payment system based on capitation). Inclusion criteria: Not clearly indicated (just mentioned that recruitment involved health services organisations in Ontario). Exclusion criteria: Not clearly stated. Subject disposition: Received facilitation intervention (23 practices), received no intervention (23 practices). Intervention practices vs. control practices (mean number of physicians per group practice 291 vs. 270, mean year of graduation from medical school 1975 vs. 1975, mean % of female physicians per practice 126 vs. 204, mean number of registered nurses per practice 116 per 164).

Intervention Multifaceted intervention delivered by one of three nurse facilitators (prevention facilitators) who completed a 30-week intensive training program before being assigned to the intervention practices. Facilitators met with the lead physician and were introduced to the practice staff. They used seven intervention strategies that were designed to change practice patterns and improve preventive care performance (audit and ongoing feedback, consensus building, opinion leaders and networking, academic detailing and education materials, reminder systems, patient-mediated activities, and patient education materials).

Comparator No intervention. Outcome definitions Three outcome measures were defined:

Overall index of preventive performance (calculated as the proportion of eligible patients who received eight recommended preventive manoeuvres less the proportion of eligible patients who received five inappropriate preventive manoeuvres). One of these manoeuvres involved Smoking cessation counselling (C). Up-to-datedness index (the proportion of recommended manoeuvres done) Inappropriateness index (the proportion of inappropriate manoeuvres done) Outcome data were collected from the time the study began to track the rate of improvement in prevention over time. Analysed preventive performance after 9, 15 and 18 months according to the date of the patient’s last visit, as determined by the chart audit.


Analyses: The unit of intervention and analysis was the medical practice. General linear model analysis of variance to test for differences between study groups in preventive performance index after 9 months, 15 months, and 18 months of intervention. Used Student’s t-test for independent groups to test for significant differences between the two groups. Sample size calculation: There were 20 practices per study group, for a total of 40 practices. This allowed the detection of a mean difference of .09 in the preventive performance index between the intervention and control groups with an α value of .05 and 80% power.

231


Table 45: Lemelin et al. (2001) (continued) A. (2) Adequately reported: Practices randomly assigned to either intervention or control, primary care practice (one to six doctors was the unit of randomisation and unit of analysis). Method of randomisation was not stated. Each facilitator had primary responsibility for up to eight primary care practices. Practice assignment was done according to geographical proximity to the facilitators’ residences. The facilitators had no interaction with the control practices. The latter were told that they were involved in a study on prevention but not told which manoeuvres were being measured. B. (2) Adequate: randomisation and follow-up audits were performed by different auditors, who were blinded as to the status of the practices and the assessment of outcomes. C. (2) Adequate/Reported: Participating physicians were younger than non-participants (p < .001) and were more likely to be female (p = .001). However, the intervention and control practices did not differ significantly in any of the measured demographic characteristics. D. (1) Inadequately reported E. (1) Inadequately reported F. (0) Not reported G. (1) Inadequate: One practice from the intervention group was lost to follow-up because the practice was moved.



Specific to smoking cessation: Performance of smoking cessation counselling as one of three grades A and B preventive manoeuvres before and after the intervention (C): Mean % proportions of eligible patients from intervention practice before and after the intervention were 64.0 vs. 73.0 (as determined by patient survey), 37.6 vs. 41.2 (as determined by chart review), and the absolute difference was 26.4 vs. 31.8. Preventive manoeuvres performed in the intervention and control practices before and after the intervention, as determined from patient chart review, showed a difference change between intervention and control practices of 54 for smoking cessation counselling manoeuvre. Mean % of eligible patients before and after the intervention is 37.6% and 41.2% respectively as compared to 40.5% and 38.7% before and after the intervention in the control group respectively. General prevention results: Before the intervention: Index of preventive performance similar for both groups (31.9% (95%CI 27.3-36.5%) vs. 32.1% (95%CI 27.2-37.0%) respectively). At follow-up: Index of preventive performance similar for both groups (43.2% (95% CI 38.4-48.0%) vs. 31.9% (95% CI 26.8-37.0%) respectively), for an absolute improvement in the intervention group of 11.5% (p < .001). The mean proportion of eligible patients who received the recommended manoeuvres was 62.3% (95%CI 58.2%-66.4%) in the intervention group, as compared with 57.4% (95%CI 54.1-60.7%) in the control group, for an absolute improvement of 7.2% (p = .008). The corresponding values for the inappropriate manoeuvres were 19.1% (95%CI 15.6%-22.6%) and 25.5% (95%CI 20.0%-31.0%), for an absolute improvement of 4.4% (p = .019).

Authors conclusions

The tailored multifaceted intervention delivered by nurse facilitators was effective in modifying physician practice patterns and significantly improved preventive care performance.

232


Table 45: Lemelin et al. (2001) (continued) Reviewers notes One of the strengths of the study was that the primary outcome measure considered a

large number of both recommended and inappropriate preventive manoeuvres supported by the evidence-based guidelines of the Canadian Task Force on Preventive Health Care.


Yes, assessing a multifaceted intervention in delivering counselling for smoking cessation




233


Table 46: Lennox et al. (1998) Citation Lennox, A. S., Bain, N., Taylor, R. J., McKie, L., Donnan, P. T., & Groves, J. (1998).

Stages of Change training for opportunistic smoking intervention by the primary health care team. Part I: randomised controlled trial of the effect of training on patient smoking outcomes and health professional behaviour as recalled by patients. Health Education Journal, 57(2), 140-149.

Level of evidence * Level II Country Scotland Research question/aims

Aim: To assess the impact of a training intervention on health professionals and smoking subjects at one year follow-up.

Study type/design RCT, pragmatic trial, 14 months follow-up, analysed/eligible = 61%, smoking specific. Participant group Participants: Approximately 140 GPs, nurses, and other staff from 16 practices (and

random sample of their patients). Inclusion criteria: All general practices in Aberdeen city. Exclusion criteria: Involved in other health promotion research, staff member on research team, impending large scale staff changes. Subject disposition: Not stated.

Intervention A one-day ‘Stages of Change’ training (including motivational interviewing) workshops for primary health care teams (off-site). The workshops aimed to address three commonly stated barriers to HCPs delivering effective health promotion in primary care settings, specifically, a lack of training, low self-efficacy and a perceived lack of time.

Comparator No intervention/‘untrained’ controls. Outcome definitions – Health professionals’ behaviour as assessed by postal questionnaire (of smoking

patients) at eight and 14 months follow-up. Recall of smoking being mentioned in a consultation since the time of the workshop. – Point prevalence and continuous abstinence.


Analyses: The unit of randomisation was the practice. Pair-matched for list size, staff numbers, social deprivation. Binary outcomes analysed by chi-square tests, logistic and multiple logistic regression analysis. Adjusted for age, sex, deprivation, and intra-practice clustering. Sample size calculation: 80% power assuming 75% follow-up data, 3% difference in abstinence, alpha = .05, 1410 patients required. A. (2) Adequate/Reported: Computer generated numbers. B. (2) Adequate/Reported: Blinded allocation and assessment of principal outcomes. C. (0) Unknown/Not reported: HCPs not defined, no significant differences between groups of patients. D. (1) Inadequate: Criteria not well defined. E. (1) Inadequate: Point estimates for continuous abstinence 5% control, 8% intervention, not specifically stated for ‘ABC’. F. (2) Adequate/Reported: Yes. G. (2) Adequate/Reported: Yes: from 26 practices, four declined and six excluded for staff changes, other workshops, and other study.



234


Table 46: Lennox et al. (1998) (continued) Results (within scope of this review)

Smoking patients in the intervention group were more likely than those in the control group to recall smoking having been mentioned in a consultation during the 14 months follow-up, but no evidence to suggest improved patient outcomes. Patient survey data: GPs Ask/Brief advice: 74.9%(control)/ 79.4%(exp); Difference 4.5%, 95%CI -0.71-9.71; p = .09 n.s. Nurses Ask/Brief advice: 76.2% (control)/ 83.2% (exp); Difference 7.0%, 95%CI -3.58 -17.58; p = .19 n.s. Other HCPs Ask/Brief advice: 68.6%(control)/ 73.7%(exp); Difference 5.1%, 95%CI -15.69-25.86; p = .63 n.s. Patient outcomes, smokers cohort, 14 month follow-up: Continuous abstinence between 8 and 14 months: 4.7%(control)/ 3.6%(exp); Difference -1.1%, 95%CI -3.03-‘-0.83; p = .26 n.s.

Authors conclusions

Patients in the intervention group were more likely than controls to recall smoking having been mentioned in a consultation, but there were no significant effects of the intervention on patient smoking outcomes fourteen months after the workshops. The design, power and analysis make it unlikely that the study failed to detect an effect on patient outcomes. Changing health professionals’ attitudes and self-efficacy does not guarantee sustained changes in preventive behaviours: organisational factors are also important.

Reviewers notes Effectiveness trial not efficacy trial; not all patients necessarily attended the practice during the study. Retrospective power calculations verified that the lack of statistical significance for GPs was not due to lack of power, however, for nurses and other HCPs, this may have been the case.


Yes, relevant. Employed a randomised trial to examine the effectiveness of a training workshop to change health professionals’ smoking prevention behaviours.




235


Table 47: McRobbie et al. (2008) Citation McRobbie, H., Hajek, P., Feder, G., & Eldridge, S. (2008). A cluster-randomised

controlled trial of a brief training session to facilitate general practitioner referral to smoking cessation treatment. Tobacco Control, 17(3), 173-176.

Level of evidence * Level II Country U.K. Research question/aims

To evaluate the effects of a brief GP training session on the number of referrals received by their local specialist smoking cessation services (SSCS).

Study type/design RCT; cluster-randomised controlled trial; 3 months, analysed/eligible = 20%, smoking specific.

Participant group Participants: 91 GPs from three east London primary care trusts (PCTs) (46(control) / 45 (exp)) (from 55 practices). Inclusion criteria: Full-time and part-time GPs within study area. Exclusion criteria: GPs outside the study area. Subject disposition: Not specified.

Intervention A 40 minute interactive training session addressing the rationale and skills for referral of smokers for treatment.

Comparator No-intervention/referral guidance by post. Outcome definitions The number of referrals recorded by the SSCS over three months after the

intervention. Data analyses & statistics

Analyses: Poisson regression, accounting for heterogeneity between practices using a random effect for practice; adjusted for baseline referral rate and number of GPs per practice; intention-to-treat analysis. Sample size calculation: To detect an increase from the baseline referral rate of .6 to 2 referrals per GP over a three-month period: minimum of 40 participants/arm; 85% power, alpha = 5%, two-tailed test; based on an assumption of three GPs per practice and an intra-cluster correlation in referral rates between practices of 0.1. A. (2) Adequate/Reported: Practices within each PCT were randomly allocated to either usual care or intervention arms. B. (2) Adequate/Reported: (adequate within the context) Research assistant blinded to GP allocation for outcome assessment. C. (0) Unknown: Not reported. D. (1) Inadequate: Inclusion and exclusion criteria were not well defined. E. (2) Adequate: Baseline measure or estimates of (‘untrained’) GPs’ smoking cessation intervention behaviour provided for 12 month pre-intervention period. F. (2) Adequate: Yes. G. (2) Adequate: No withdrawals.



Three-month baseline mean number of referrals per GP: 0.6(control) /1.0(exp); post-intervention 1.8(control) /6.4(exp). Incidence rate ratio for referrals from the intervention arm compared to usual care was 4.9 (p < .001; 95% CI 1.7 to 14.7).

Authors conclusions

A brief training session can significantly increase GP referral to smoking cessation services. However, the effect of the intervention dropped off after the first month (still remaining significant at three months) and it is likely that the intervention would need to be reinforced periodically.

Reviewers notes This study demonstrates the efficacy of education as an intervention to promote ABC and highlights the necessity for such interventions to be reinforced and/or combined with prompts and reminders to maintain effectiveness over time. Issues around generalisability include the ‘expert’ status of the educators and volunteerism (the enthusiastic minority of GPs [24%] who volunteered for the study). Short follow-up time. Also, note the single and relatively easy outcome (task) of referral (i.e. this doesn’t actually take much of the GPs’ time as the counselling work is done off site).

236


Table 47: McRobbie et al. (2008) (continued) Relevance to study question

Yes, relevant. Employed a controlled, randomized trial to examine the efficacy of education as an intervention to promote ABC.




237


Table 48: McPhee et al. (1991) Citation McPhee, S. J., Bird, J. A., Fordham, D., Rodnick, J. E., & Osborn, E. H. (1991).

Promoting cancer prevention activities by primary care physicians. Results of a randomized, controlled trial. Jama, 266(4), 538-544.


To test the efficacy of a computerised reminder system, supplemented by cancer education materials, in promoting 11 cancer prevention activities, including screening of smoking status and smoking cessation counselling in community practice primary care settings.

Study type/design RCT; multi-centre; 12 months, duration = 12 months, analysed/eligible = 16%, general prevention.

Participant group Participants: 40 primary care physicians (one physician per practice) (20control) – 20 (exp)). Inclusion criteria: Full-time physicians in small group community-based practices in five northern California counties. Exclusion criteria: Practices of > four physicians, out of geographic area. Subject disposition: Male 75%; 30 family physicians, 10 general internists; 55% in solo practice; mean age 45 years (no significant differences in baseline characteristics between intervention and control groups).

Intervention Computer-generated cancer prevention reminders of overdue screening tests, and smoking and dietary assessment and counselling, supplemented by cancer education materials.

Comparator No-intervention. Outcome definitions Performance scores as percentage compliance with clinical guidelines. Percentage of

patients each physician assessed for smoking status during each 12 month period (pre/post intervention) and percentage of smokers (as assessed by physician) who had been counselled during one year. Data extracted from a random sample of 60 medical records per physician for 12 month pre-intervention and post-intervention.


Analyses: Student t-tests; multiple regression analyses provided estimates of incremental differences in performance scores between intervention and control groups; α set at the .05. Sample size calculation: Sample size of 40; effect size of .30 or greater; 80% power; α = .05. A. (1) Inadequate: Unclear. B. (2) Adequate/Reported: (adequate within the context) Research assistant blinded to study group for outcome assessment. C. (2) Adequate/Reported: Yes, participant characteristics tabulated with no significant differences between groups. D. (2) Adequate: Inclusion and exclusion criteria were defined. E. (2) Adequate: Baseline measure or estimates of (‘untrained’) GPs’ smoking cessation intervention behaviour provided for 12 month pre-intervention period. F. (2) Adequate: Yes. G. (2) Adequate: No withdrawals.



Controlling for pre-intervention performance levels, significant incremental differences in performance scores between intervention and control groups were achieved (1) Smoking status recorded: 32.4% (SD 13.9)(control)/45.0% (16.6)(exp); p = .014 (2) Smoking cessation counseling (delivered to identified smokers): 41.8% (SD 22.2)(control)/58.8% (23.0)(exp); p = .027.

Authors conclusions

The mean performance scores of intervention group physicians were significantly higher than controls for smoking assessment and counselling. The intervention addresses the physician barriers of forgetfulness, time constraints and logistical difficulties. The effectiveness of the reminders suggests that physician forgetfulness may account for a considerable part of the gap between cancer prevention recommendations and actual clinical performance. Computerized reminders can significantly increase physicians' performance of cancer prevention activities in community-based practices.

238


Table 48: McPhee et al. (1991) (continued) Reviewers notes Note: (1) The intervention requires office staff to create new records and/or update

computer records before each eligible patient’s appointment. (2) In part, this study answers the question of whether the differences observed are attributable simply to improved record-keeping or actual clinical behaviours. The same intervention was applied simultaneously across a range of recorded physician tasks, some of which could also be validated by laboratory data, thus giving weight to the conclusion that physicians’ behaviour (rather than recording) was modified.


Yes, relevant. Employed a controlled, randomized trial to examine support mechanisms to facilitate the implementation of smoking cessation interventions by GPs to all tobacco users, in community care settings.




239


Table 49: Millett et al. (2007) Citation Millett, C., J. Gray, et al. (2007). "Impact of a pay-for-performance incentive on support

for smoking cessation and on smoking prevalence among people with diabetes." Cmaj 176(12): 1705-10.

Level of evidence * Level III-2 Country U.K. Research question/aims

To assess the impact of a pay-for-performance incentive in the United Kingdom introduced in 2004 as part of the new general practitioner contract to improve support for smoking cessation and to reduce smoking prevalence among people with diabetes in a multiethnic population.

Study type/design Population-based longitudinal study (cohort) examined pre-and post-intervention, duration = 15 months before 2003 and 2005 study periods, analysed/eligible = 32/36 88.88%, smoking specific.

Participant group Participants: General practices in Wandsworth Primary Care Trust in southwest London, England (36 primary care practices). Inclusion criteria: All practices in the study area are eligible. Patients with diabetes type 1 or 2 (codes for diabetes diagnosis C10 or diabetes care 66A) identified, patients received repeat prescriptions for diabetic medications or whose gHb (glycated haemoglobin) levels > 7.5% also included. Exclusion criteria: Not for general practices, but patients < 18 years, women with gestational diabetes or who received treatment for polycystic ovarian syndrome rather than diabetes were excluded. Subject disposition: 32 practices agreed to participate in the study, with 4248 adults ≥ 18 years with diabetes registered, in both the 2003 and 2005 study periods (2227 men, 2057 women). Missing data 2.9%.

Intervention Pay-for-performance incentive as a method to improve quality of care received by patients introduced in 2004.

Comparator Pre- (2003) and post- the intervention (2005). Outcome definitions Smoking status recorded ‘A’, and smoking cessation advice ‘B’ based on information

recorded on practice computers during the 2003 and 2005 study periods. These were part of the disease indicator within the clinical domain of the Quality and Outcomes Framework in UK GP contract of April 2004. Of 99 points available for diabetes care, five points were allocated to the measures of care for patients with diabetes who smoke and whose notes contain a record that smoking cessation advice has been offered in the last 15 months.


Data for the Wandsworth Prospective Diabetes Study were collected before (June-October 2003) and after (November 2005-January 2006) the introduction of the new GP contract in April 2004. Analyses: McNemar test was used to examine differences in frequency distributions of indicators between 2003 and 2005. Intra-practice correlations for different variables were: recording of smoking status, smoking cessation advice, and actual smoking status. Sample size calculation: Not reported. A. (2) Adequate/Reported: Not a randomised trial. B. (2) Not applicable C. (1) Inadequate: Not fully reported for the general practices characteristics . D. (1) Inadequate: Reasonably reported. E. (2) Adequate: Yes. Baseline rates of ‘A’, ‘B’, and prevalence rates. F. (1) Inadequate: No. G. (0) Unknown/Not reported: No.



240


Table 49: Millett et al. (2007) (continued) Results (within scope of this review)

‘A’ smoking status recorded 15 months before the 2003 and 2005 study periods, % of patients and % change with adjusted OR for age, sex, ethnic background, deprivation group and practice-level clustering with (95% CI). Smoking status was significantly more likely to be ever recorded in 2005 than in 2003 (98.8% v 90.0%, p < .001). Proportion of patients whose smoking status was recorded in the 15 months before the study period increased from 67.6% in 2003 to 86.7% in 2005 (p < .001). ‘B’ proportion of patients with documented smoking cessation advice increased from 48.0% in 2003 83.5% in 2005 p < .001). Smoking prevalence deceased significantly over time from 20% (2003) to 16.2% (2005) (a change in percentage of -3.8% (p < .001) using McNemar test). The decrease was lower among women than among men.

Authors conclusions

Pay-for-performance incentives appear to be effective in increasing the delivery of cessation advice given by primary care physicians and in reducing the prevalence of smoking among patients with diabetes. However, financial incentives are likely to be most effective in reducing the prevalence of smoking when combined with other quality improvement initiatives within a comprehensive tobacco control strategy.

Reviewers notes Characteristics of the practices and individuals who provide the services (GPs and other staff) was not clear. Restricted to patients with diabetes and not the general population. Authors noted that some of the reduction of smoking prevalence may have been attributable to other cessation interventions, to improved recording of smoking status, or to a secular trend of reduced tobacco use. Authors commented that the findings raise the question of whether these incentives result in improved patient outcomes or do they just lead to improved documentation and use of processes of care? There remains the need for a randomised controlled trial to evaluate the impact of the intervention. Smoking status was through patient self-report and was not confirmed objectively. The quality of the cessation advice given was not assessed.


The study is relevant, as the intervention was to improve support for smoking cessation and promote the delivery of cessation advice.




241


Table 50: Moher et al. (2001) Citation Moher, M., Yudkin, P., Wright, L, Turner, R., Fuller, A., Schofield, T., et al. (2001).

Cluster randomised controlled trial to compare three methods of promoting secondary prevention of coronary heart disease in primary care. BMJ, (322), 1-7.

Level of evidence * Level II Country U.K. Research question/aims

To assess the effectiveness of three different interventions for improving the secondary preventive care of patients with coronary heart disease delivered at the level of general practice: audit and feedback, recall to a GP, and recall to a nurse clinic.

Study type/design RCT: un-blinded, cluster randomised controlled trial of three group design: audit group vs. GP recall group vs. Nurse clinic recall group, duration = 18 months, analysed/eligible = 21/79 (26.5%), general prevention.

Participant group Participants: General practices, n = 21. Inclusion criteria: All 79 general practices in Warwickshire were invited. Exclusion criteria: Not stated. Subject disposition: Of 41 general practices that expressed interest [21 recruited, 20 withdrew (15 ineligible as already running nurse-led clinics, five committed to opportunistic care) so expressed interest but not recruited], 38 did not express interest. [(Of the 21 practices the total of patients alive and registered with the practices at follow-up are 1906 patients; (559 patients in audit group, 682 in GP recall group, and 665 in Nurse recall group)].

Intervention Audit and feedback: Summary audit results at a practice meeting. Practices asked to provide usual care and were given no further support during the trial. Recall to GP: Practices given same patient information as audit group but also the names of patients identified as having coronary heart disease. The facilitator discussed and agreed to guidelines for secondary prevention with the practice doctors and gave ongoing support in setting up a register and recall system for regular review of patients with CHD by their GP. Recall to nurse clinic: Practices given same patient information as GP recall group. The trial’s nurse facilitator gave ongoing support to the practices in setting up a register and recall system for systematic review of patients with CHD in a nurse-led clinic. After discussion and agreement to guidelines for secondary prevention, the practice doctors and nurses agreed to the clinic protocols, and the nurses received education to implement it.

Comparator Audit vs. GP recall vs. Nurse recall. Outcome definitions ‘A’ (at baseline and at follow-up at 18 months) A record of smoking habit and, for

smokers, a record of review in the previous two years. Data analyses & statistics

Analyses: ANOVA. Sample size calculation: To account for the clustered design authors applied cluster inflation factor and carried out sensitivity analyses assuming an intra-cluster correlation coefficient between .05 and .06 and a cluster size between 100 and 200. Under any of these assumptions seven practices in each trial arm would give the study a power of at least 82% (maximum 90%), and this trial size was adapted.

242


Table 50: Moher et al. (2001) (continued) A. (2) Adequate/Reported: Blocked random allocation was performed within three strata, randomisation, based on computer generated random numbers was carried out by a statistician blind to the identity of the practice. B. (2) Adequate: The authors stated that a research nurse who was blind to the patient’s allocation group and had no previous involvement with the trial carried out the clinical assessment (including blood pressure measurement and blood sampling for serum cholesterol and cotinine estimations). C. (1) Inadequate: Baseline characteristics were shown in details for the 21 recruited (in terms of list size, whole time equivalent for both GPs and nurses, training practice, computerised practice and location) against practices that showed interest but were not recruited or those that expressed no interest (but no data shown within the three recruited groups). Also baseline characteristics shown in details for patients within the three groups but not for the characteristics of the practices.


D. (1) Inadequate: Only stated that all 79 general practices in Warwickshire were invited to participate in the study. E. (2) Adequate/Reported F. (0) Not reported G. (0) Unknown/Not reported: Not reported.


Smoking status assessment (Ask) increased from baseline among the three groups. This increase was from 73% to 78% (audit group), from 71% to 92% (GP recall group), and from 71% to 95% (Nurse recall group). p -value for this difference in smoking status between the groups (based on means adjusted for baseline) was statistically significant (.001). In details other results are: At baseline about 30% of patients were adequately assessed overall, this proportion increased markedly during the trial in all three groups. This was from the improvement in assessing cholesterol levels, but assessment of smoking status and blood pressure also increased, especially in the GP recall and nurse recall groups. Adequate assessment of all three risk factors was much more common in the nurse and GP recall groups (85%, 76%) than the audit group (52%). The advantage in the nurse recall compared with the audit group was 33% (95% confidence interval 19% to 46%); in the GP recall group compared with the audit group 23% (10% to 36%), and in the nurse recall group compared with the GP recall group 9% ( - 3% to 22%).

Authors conclusions

Setting up a register and recall system improved patient assessment at 18 months’ follow-up, but was not consistently better than audit alone in improving treatment or risk factor levels. Understanding the reasons for this is the key next step in improving the quality of care of patients with coronary heart disease.

243


Table 50: Moher et al. (2001) (continued) Reviewers notes The study assessed methods of promoting secondary prevention of coronary heart

disease in primary care. The ‘Ask’ was through the baseline audit of notes from patients' records at baseline and then patients were invited to participate and complete an enclosed questionnaire of two quality of life instruments. Subject disposition for GPs and practice nurse was not clearly identified in the study, but by assessing Table 1 in the paper, an estimation was made: 10 Practices with < 4 GP (1.0 FTE) were recruited; so number of GPs would be from (1x10) to (3x10) i.e. 10-30. 7 Practices with 4- GP (1.0 FTE) were recruited; so number of GPs would be from (4x7) to (5x7) i.e. 28-35. 4 Practices with ≥ 6 GP (1.0 FTE) were recruited; so number of GPs would be (≥ 6x4) i.e. ≥24. This gives a total of 62 to ≥ 89 GPs. 3 Practices with <1 practice nurse (1.0 FTE) were recruited; so number of practice nurses would be (0x3) i.e. 0. 8 Practices with 1- practice nurse (1.0 FTE) were recruited; so number of practice nurses would be (1x8) i.e. 8. 10 Practices with ≥ 2 practice nurses (1.0 FTE) were recruited; so number of practice nurses would be (≥ 2x10) i.e. 20. This gives to a total of ≥28 practice nurses.


Yes.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; BMI = body mass index; CBA = controlled before-after trial; CCT = controlled clinical trials; CFPC = College of Family Physicians of Canada; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FFS = fee for service; FP = family practitioner; FTE = full time equivalent; GEEs = generalised estimating equations; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; eLinkS = electronic linkage system; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; ORP = Outcomes Recognition Programme; PAR = practice assessment report; PCTs = Primary Care Trusts; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; THIN = The Health Improvement Network; TS = time-series design; UC = usual care; UMMS = University of Massachusetts Medical School; V.A. = Veterans Affairs clinic setting.

244


Table 51: Morgan et al. (1996) Citation Morgan, G. D., Noll, E. L., Orleans, C. T., Rimer, B. K., Amfoh, K., & Bonney, G.

(1996). Reaching midlife and older smokers: Tailored interventions for routine medical care. Preventive Medicine, 25(3), 346-354.


To assess the effectiveness of an office-based smoking cessation program (physician-delivered brief quit-smoking advice and counselling) tailored to midlife and older smokers (ages 50-74 years).

Study type/design RCT, duration = 6 months (a two year RCT), smoking specific Participant group Participants: Of 49 eligible primary care practices 39 practices accruing five or more

patients per practice. Inclusion criteria: Willingness to participate in a brief office-based training program, absence of a formalised smoking intervention program, and provider projections of ability to accrue 25 age-eligible patients within three months. Patients were eligible if they were aged 50-74, seeing the physician for non-crisis visit, were smokers. Status as a smoker was defined as having had a cigarette in the previous 7-days. Exclusion criteria: Practices that accrued less than five patients were excluded. Subject disposition: Entire practices (physicians and key non-physician office staff) were randomized to Immediate Intervention (n = 23) or Delayed Intervention (n = 26) conditions. Following exclusions (see sample size calculation) 18 Immediate Intervention (Intervention group) and 21 Delayed Intervention (control) practices were left. Specialties included family medicine (n = 20), internal medicine (n = 17), cardiology (n = 1). There were no differences in specialty across conditions. 659 patients.

Intervention Immediate intervention (brief quit-smoking advice and counselling). Intervention practices received on-site training to implement a modified National Cancer Institute (NCI) smoking cessation intervention (Ask about smoking at every opportunity, Advice all smokers to stop, Assist the patient to stop smoking, and Arrange for follow-up support). The training program was designed for brief office-based presentation using the principles of pharmaceutical and academic detailing.

Training involves praising patients for previous quit efforts, providing personalised feedback linking smoking to symptoms, discussing the health benefits of quitting for older smokers, and giving a clear message to stop smoking.

Comparator Delayed intervention (usual care). Control practices were instructed to provide usual care to their older smokers over the accrual and follow-up period.

Outcome definitions Physicians’ perceptions of effectiveness giving quitting advice, and program’s effectiveness and feasibility (Physicians completed questionnaire). Provider adherence (included eight areas, two of them correspond to ‘B’ (brief advice) and ‘C’ (cessation support): Patients in intervention group telephoned by program staff between two to four weeks after their office visit for brief follow-up counselling and to check on provider adherence to the treatment protocol. Follow-up telephone interviews were conducted by professional interviewers six months after enrolment.


Analyses: Physician practice group was the unit of randomisation. Tests: Chi-squared (X2) test (categorical variables), Mantel–Haenszel X2 (ordinal variables), and t tests (interval variables). Sample size calculation: Based on original estimates of effect size and expected variance among practices, a minimum number of five patients required per practice was calculated using power contour plots. Therefore, practices that accrued < 5 patients were excluded (after randomisation) in order to allow for sufficient power to detect hypothesized differences in quit rates.

245


Table 51: Morgan et al. (1996) (continued) A. (1) Inadequately Reported: Method of randomisation was not clearly stated, but stated that all practices were randomised (physician and key non-physician office staff) to either intervention or control groups. B. (0) Not reported C. (2) Adequate/Reported: Immediate and Delayed Intervention practices did not differ significantly in the mean number of patients enrolled, gender of patients enrolled, or reporting of quit attempts lasting 24 hours or more in the previous year. However, patients in the two conditions did differ in age, average number of cigarettes smoked daily, time elapsed until first cigarette of the morning, and contemplation status. D. (2) Adequate/Reported E. (1) Inadequately reported: Baseline characteristics were given for both groups. F. (0) Not reported G. (0) Unknown/Not reported



Results from patient reports of interaction with physician at two to four week follow-up (for the intervention group) (number of patients 259) showed that 88% of patients reported getting clear physician quitting advice since enrolment (B) and 95% reported receiving the Clear Horizons guide (a copy of a smoking cessation guide tailored to older smokers). Only 35%, however, reported having received a smoking follow-up letter. Results from patient report of physician interaction and quit rates at six-month follow-up by study group showed that there were significant differences between Immediate and Delayed practices with respect to receiving advice to quit (B) and discussions regarding smoking, setting a quit date and nicotine gum use (C) (this is shown in Fig 2 in the original paper). Statistical significance was at .001< p < .01; and p < .001. Self-reported quit rates (Q) at six-month follow-up were 15.41% for the Immediate Intervention (intervention) group versus 8.16% of subjects in the Delayed Intervention group (control) (p < .005).

Authors conclusions

Physician training was effective, judging by high provider adherence levels reported by patients and based on favourable physician ratings. The majority of intervention group physicians continued to address smoking and cessation with patients during the six months following the initial visit, further suggesting the intervention was well integrated with routine care. Smoking abstinence was significantly increased by training physicians and key office and clinical staff to intervene with older smokers. Brief interventions tailored to this age cohort can be successfully and efficaciously integrated into routine care.

Reviewers notes It was not clear how many physicians within each practice. Patients were only aged 50-74 years.


Yes, relevant as despite the study looked at the effectiveness of the intervention (in terms of abstinence rates), data were collected to reflect issues relevant to the ABC programme. Patients in the Immediate Intervention (intervention group) were telephoned by programme staff between two and four weeks after their office visit for brief follow-up counselling and to check on provider adherence to the treatment protocol.



Abbreviations: CI = confidence interval; FP = family practitioner; GP = General Practitioner; EHR = electronic health record; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; TS = time-series design; UC = usual care; UMMS = University of Massachusetts Medical School; V.A. = Veterans Affairs clinic setting.

246


Table 52: Ockene et al. (1994) Citation Ockene, J., Adams, A., Pbert, L., Luippold, R., Hebert, J., Quirk, M., et al. (1994). The

physician-delivered smoking intervention project: Factors that determine how much the physician intervenes with smokers. J Gen Intern Med 1994;9:379-384.


To determine factors that affect how much physicians trained to use a patient-centred smoking intervention intervene with their smoking patients.

Study type/design Randomised controlled trial, duration = 12 months, smoking specific. Participant group Participants: Internal medicine residents and general internal medicine attending

physicians receiving training; 540 smokers were eligible for the study. Inclusion criteria: All internal medicine residents trained in the University of Massachusetts Medical School’s postgraduate-year one and year two (PGY-1 and PGY-2) residency program from July 1989 through June 1991, and general internal medicine attending physicians, receiving training. Exclusion criteria: PGY-3 as they would not be available for one-year follow-up. Subject disposition: Residents (40), attending physicians (10), 60% men, 56% PGY-1 when initially trained, 24% PGY-2, 20% were attending physicians. No significant difference in baseline characteristics or pos-training assessments between the physicians randomised to the algorithm and no-algorithm conditions (age, gender, years of training, and post-training knowledge and skill).

Intervention Algorithm: instructing the physician in the ten smoking intervention steps each time he or she saw a smoker who was enrolled in the study.

Comparator No algorithm. Medical charts in both groups were labelled with a sticker identifying the patient as a smoker; all the physicians received the same training in smoking intervention skills which included the use of the intervention algorithm.

Outcome definitions Physician performance measure through Patient Exit Interview (PEI) scores (scores ranged from 0 (no intervention steps used) to 10 (all intervention steps used). Initial PEI scores compared for algorithm vs. no-algorithm conditions. PEI steps: 1 discussed smoking ‘C’, 2 advised to stop smoking (B), 3 discussed reasons ‘C’, 4 asked about past experiences ‘C’, 5 asked about problems ‘C’, 6 discussed specific strategies ‘C’, 7 asked about dependency ‘C’, 8 patient agreed to stop smoking or if not, patient agreed to cut down, 9 set date to stop/cut down ‘C’, 10 put plan in writing ‘C’, 11 discussed other behavioural changes ‘C’, 12 given written materials ‘C’, 13 set future contact or if not planned to discuss later ‘C’. Other outcome measures were Patient and Physician Variables Affecting Overall PEI Score.


Analyses: Initial PEI scores were compared for algorithm versus no-algorithm conditions. Individual PEI steps were evaluated separately to assess percentage of physicians who implemented each step, differences between two groups of physicians for each PEI step taken. Analyses consisted of chi-square test and student t-test for categorical and continuous variables, respectively. All physician and patient baseline factors significantly (p ≤ .05) associated with PEI score in univariate analyses were considered for multivariable analysis. Patient-physician concordance for the number of steps used by physicians was determined in a randomly selected set of patients and physicians; 13 physicians completed exit interviews similar to the PEI immediately after the intervention had been completed for 38 patients. Both physician and patient index scores were computed; the resulting Spearman’s rank order correlation coefficient was 0.773 (p = .0001). Percentage of observed agreement between physicians and patients on specific intervention steps ranged from 60.5% (“assessed reasons to stop smoking”) to 86.6% (“provided written materials”), with an overall agreement of 77%.

247


Table 52: Ockene et al. (1994) (continued) Sample size calculation: Not reported.

A. (1) Inadequate: Method of assigning participants to the two groups was not detailed, although it was stated that physicians were randomised to one of the two groups. B. (2) Adequate/Reported: Educational evaluators blinded to intervention condition and postgraduate year of resident or status as an attending physician. C. (2) Reported: It was reported that there was no significant difference in baseline characteristics or post-training assessments between the two groups including age, gender, years of training, and post-training knowledge and skill level. D. (1) Inadequate: Inclusion/exclusion criteria were reasonably defined. E. (2) Adequate/Reported F. (0) Inadequate: Not stated. G. (0) Unknown/Not reported: Not stated.



PEI scores: No statistically significant difference in scores between UMMS physicians in algorithm group (4.88) compared to no-algorithm group (4.94). Counselling steps most frequently reported by patients to be taken by all trained physicians include: discussed smoking ‘C’, advised patient to stop smoking ‘B’, discussed reasons for stopping, asked about past experiences in stopping, and planned to continue to discuss smoking at a later visit if no further contact was scheduled to specifically discuss smoking ‘C’. Patient and Physician Variables Affecting Overall PEI Score: Of the ten factors found to be predictive in the univariate analyses, only two patient baseline factors (patient report of thinking of stopping smoking within six months and higher Fagerstrom Tolerance Score) and one physician factor (older age) were independently predictive of PEI score (p ≤ .05). The interaction of Fagerstrom Tolerance Score and thinking of stopping within the next six months was not significant.

Authors conclusions

Findings from the study indicate that trained physicians implement more intervention steps with smokers who expect and desire to stop smoking. However, provision of an intervention algorithm each time a physician sees a patient does not increase the likelihood that physicians who are trained to counsel smokers and cued by stickers on charts will be more likely to perform the intervention steps taught. Therefore, these more time-consuming additional efforts may offer little benefit. Findings from the study showed that either related to overall PEI index score or for each step considered separately did not support the hypothesis that algorithm would serve as a stimulus for trained physicians to use counselling steps.

Reviewers notes Relevance to study question

Yes, relevant. Examined the effectiveness of training and algorithm as an intervention to promote ABC.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; FP = family practitioner; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; TS = time-series design; UC = usual care; UMMS = University of Massachusetts Medical School; V.A. = Veterans Affairs clinic setting.

248


Table 53: Ockene et al. (1997) Citation Ockene, J. K., Lindsay, E. A., Hymowitz, N., Giffen, C., Purcell, T., Pomrehn, P., et al.

(1997). Tobacco control activities of primary-care physicians in the Community Intervention Trial for Smoking Cessation. COMMIT Research Group. Tobacco Control, 6 Suppl 2(-), S49-S56.


To determine the effect of physician training activities on physician behaviour by comparing the smoking intervention practices of physicians and their office staff in the Intervention and Comparison communities of the COMMIT RCT.

Study type/design RCT, duration = 4-years, smoking specific. Participant group Participants: Physicians in the intervention communities participating in the continuous

medical education (CME) Inclusion criteria: Not clearly stated. Exclusion criteria: Not clearly stated. Subject disposition: 869 physicians (465 Intervention, 404 control).

Intervention Three levels of training activities for physicians (basic, comprehensive, and “train the trainers”), and one type of training for staff in physicians’ offices. These are defined below: Basic training: a minimum of 45-minute didactic presentations on tobacco and use of the National Cancer Institute (NCI) “4As” (Ask, Advise, Assist, Arrange) counselling model. Comprehensive training: didactic presentations on tobacco, use of the 4A counselling model, office systems, and skill building exercises (role play) Training for office staff focused on tobacco control and office intervention “systems”. The COMMIT provided a standard protocol that had necessary flexibility to accommodate local variations in the Intervention communities. There were 58 specified activities to be implemented in the four channels for intervention: worksites and other organisations, cessation services, public education, and healthcare providers. The overall goal for the healthcare providers was to increase the quality and quantity of their intervention with smokers.

Comparator Was not clearly stated, probably usual care or no intervention. Outcome definitions Reported smoking intervention practices of physicians. Smoking control attitudes and

practices reported by primary-care physicians in the 22 communities, smoking policies and practices of 30 randomly selected medical offices in each community, and patient reports of physician intervention activities. To assess reported smoking intervention practices of physicians in the Intervention and Comparison communities, a 14-item survey was mailed to each targeted physician (physicians practising family, general, or internal medicine, cardiology, pulmonology, obstetrics/gynaecology, or osteopathy). It was projected that 25% of physicians would report setting “stop smoking” dates with patients most of the time and 25% of smokers would report being asked to set a “stop smoking” date. Another impact objective was that 60% of patients who smoke would report that their physicians advised them to stop smoking.


Analyses: The unit of analysis is the community (as the community was the unit of randomisation in the COMMIT trial). Sample size calculation: Not stated.

249


Table 53: Ockene et al. (1997) (continued) A. (2) Adequate/Reported: office practices randomly selected from lists compiled from telephone book for both intervention and comparison. B. (1) Inadequate C. (2) Adequate/Reported D. (0) Not reported E. (2) Reported: tobacco control activities for physicians and physician office practices were provided at baseline and post-intervention through surveys. F. (0) Unknown/Not reported G. (0) Unknown/Not reported



Physician survey results: Based on the complete sample of physicians identified as practising in the community, response rate for the 1993 physician survey for Intervention and Comparison communities were: 45% (range 19.8-64.8%) and 42% (range 10.6-61.7%). Telephone interviews of office staff: Response rates of 84% in both conditions. Physicians attended training: Intervention group more than those in comparison (53% vs. 26% p < .0005). In both conditions, training attendees perceived themselves as being better prepared to counsel smokers than non-attendees (p ≤ .01) and reported more activity in smoking intervention. Intervention communities carried out more office-based tobacco control activities (p = .002). Smokers in intervention were more likely to report receiving reading material about smoking from their physicians (p = .026). No other differences in physician intervention activities were reported by smokers between the two communities. 96% of the participating physicians in both conditions reported that they routinely ask new patients about smoking, and 79% asked established patients. ‘A’ but there were no significant differences between the Intervention and the control communities. Physicians in the Intervention communities were significantly more likely to: Report advising patients to stop smoking (p = .009), ‘B’ Encourage patients to set a quit date (p = .004), ‘C’ Recommend nicotine replacement therapy (p = .05) ‘C’ For each of these reported practices, with the exception of setting a quit date, physicians in the Intervention communities demonstrated significant increases in the post-intervention survey compared with the baseline survey. Endpoint cohort survey (1993): 66% of a possible 20 347 people provided smoking outcome data and 57% provided information on physician intervention. Over 70% of all smokers reported being asked about their smoking ‘A’, although less than half of light-to-moderate smokers had been advised to stop.

250


Table 53: Ockene et al. (1997) (continued) Significantly higher proportion of offices in the Intervention communities reported that

they posted no-smoking signs (p = .03) and anti-smoking posters (p = .02), and provided stop-smoking pamphlets (p = .02), local cessation resource guides (p = .02), and materials for the Great American Smokeout (p = .002). There was a significant improvement in all of the office system activities when the 1990 office practice survey is compared with the 1993 survey for the Intervention communities, lending support to the likelihood that these are actual changes in office practices as a result of education. Endpoint cohort survey (1993) results: 66% of a possible 20,347 people provided smoking outcome data and 57% provided information on physician intervention. They responded to questions about whether their physicians provided assistance with cessation ‘C’. Only one significant finding was found: More smokers from the Intervention communities reported receiving pamphlets about smoking than those in the comparison communities (p = .026). Across both conditions, all of the physician intervention activities were reported more frequently by heavy smokers (p < .001).

Authors conclusions

The COMMIT intervention had a significant effect on some reported physician behaviours, office practices and policies. However, most physicians still did not use state-of-the-art smoking intervention practices with their patients and there was little, or no, difference between patient reports of intervention activities of physicians in the intervention and comparison communities. Better systems and incentives are needed to attract physicians and their staff to CME (continuous medical educations) and to encourage them to follow through on what they learn. The recently-released Agency for Health Care Policy and Research clinical practice guideline for smoking cessation and other standards and policies outline these systems and offer suggestions for incentives to facilitate adoption of these practices by physicians.

Reviewers notes The comparison (control) was not defined. Relevance to study question

Yes, relevant as survey questions asked smokers series of questions to assess intervention and programme awareness. They were asked their smoking status, whether their physicians asked them whether or not they smoked ‘A’, advised them to stop smoking ‘B’, asked them to set a quit date ‘B’, gave them any pamphlets about smoking ‘C’, or prescribed nicotine replacement therapy ‘C’.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; COMMIT = community intervention trial for smoking cessation; FP = family practitioner; GEEs = generalised estimating equations; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; ORP = Outcomes Recognition Programme; PAR = practice assessment report; PCTs = Primary Care Trusts; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; UC = usual care; V.A. = Veterans Affairs clinic setting.

251


Table 54: Ozer et al. (2001) Citation Ozer, E. M., Adams, S. H., Lustig, J. L., Millstein, S. G., Camfield, K., El Diwany, S., et

al. (2001). Can it be done? Implementing adolescent clinical preventive services. Health services research, 36(6 Pt 2), 150-165.


The study hypotheses are: 1) Clinician screening and counselling with adolescents during routine well visits would increase following a clinic-wide intervention involving provider training, customized screening and charting tools, and the resources of a health educator; and 2) the significant increases in clinician screening and counselling would remain over an 18-month follow-up period.

Study type/design Pre-test/post-test study design, duration = 18 months, general prevention. Participant group Participants: Three outpatient paediatrics clinics within a large managed care system

throughout Northern California. Inclusion criteria: Not clearly stated, but it was noted that clinics were selected based on their provision of care to large numbers of adolescents and their agreement to participate in a longitudinal study of provision of clinical preventive services to adolescents. Exclusion criteria: Not stated. Subject disposition: 89 clinicians in the three sites agreed to participate in the study and were trained in the intervention (86% were physicians and 14% were nurse practitioners; 58% were female, and median year of graduation was 1979).

Intervention The intervention had three components: (1) 89 clinicians from three outpatient paediatric clinics attended a training to increase the delivery of preventive services; (2) customized adolescent screening and provider charting forms were integrated into the clinics; and (3) the resources of a health educator were provided to the clinics. The goal of the intervention was to increase clinicians’ screening and brief counselling of adolescents in the targeted health risk areas of tobacco, alcohol, sexual behaviour, and safety (seatbelt and helmet use). All clinicians received training in adolescent preventive services, and clinicians used screening and charting forms customized for this study. Study adolescents met with their primary care providers for a well visit that lasted 20 to 30 minutes. The overall intervention focused on adolescents taking responsibility for their health and learning ways to keep themselves healthy.

Comparator Post-implementing the intervention. Outcome definitions Independent adolescent reports of clinician screening and counselling practices were

obtained from three separate groups of adolescents (mean age of 14 years) who attended well visits with their primary care providers: 1) One group prior to the preventive services intervention (pre-implementation); 2) the second group after the intervention had been initiated in the clinics (post-implementation); and 3) the third group during an ongoing follow-up phase (follow-up). The Adolescent Report of the Visit, a 39-item patient-report measure, included questions on whether clinicians screened and offered brief counselling messages for each of the five target risk areas and possesses adequate construct validity. An example of a screening question is: “Did your doctor ask if you smoke or chew tobacco?” ‘A’. An example of a counselling question for adolescents who were not engaging is: “Did your doctor encourage you to remain a non-smoker or non-tobacco user?” ‘B’. The effect of implementing the preventive services intervention was tested by examining changes in screening and counselling practices from the pre- to post-implementation period. The sustainability of the intervention over a longer period was evaluated by examining screening and counselling rates obtained during the follow-up period.

252


Table 54: Ozer et al. (2001) (continued) Data analyses & statistics

Analyses: Following a visit, adolescents completed surveys reporting on clinician screening and counselling for each of the target risk areas. Pre-implementation (three months), 104 adolescents completed surveys. Post-implementation of the training, tools, and health educator intervention, 211 adolescents completed surveys (five months). For 18 months post-implementation clinicians delivered services and 998 adolescents completed surveys. The evaluation of the implementation included analysing differences in screening and counselling rates at three different periods. Chi square analyses were used to examine the changes from T1 (pre-implementation=3m) to T2 (after the preventive services intervention had been initiated and lasted an average of five months), from T1 to T3 (follow-up phase began following the post-implementation period and continued for 18 months), and from T2 to T3. Sample size calculation: Not stated. A. (2) Adequate/Reported: Not RCT. B. (2) Not applicable: (Pre-post design). C. (2) Adequately reported D. (0) Not reported E. (1) Inadequately reported: Baseline percentages for screening and counselling rates of the five risk areas including tobacco use. F. (0) Unknown/Not reported G. (1) Inadequate



Screening: Based on adolescent reports, clinician screening practices increased significantly from T1 to T2. Adolescents reported being screened a greater percentage of the time at T2 than at T1 in tobacco use (61% to 95%; p < .0001). Brief Counselling: Clinicians’ brief counselling increased significantly from T1 to T2 in tobacco use (55 percent to 96 percent; p < .0001). General results: Screening: Alcohol use (59% to 96%; p < .0001), sexual behaviour (47% to 82%; p < .0001), seatbelt use (36% to 98%; p < .0001), and helmet use (33% to 98%; p < .0001). The aggregate screening score across all five risk areas also increased significantly (47% to 94%; p < .0001). The increases from T1 to T3 were of a similar size and significance level. Although screening rate increases remained highly significant through the follow-up phase, there were small but significant decreases in rates between T2 and T3. Decreases in screening rates for the areas of tobacco, alcohol, seatbelt, and helmet use ranged from 4% to 7% from the post-implementation to the maintenance phase. Brief Counselling: Clinicians’ brief counselling increased significantly from T1 to T2 in alcohol use (46% to 91%; p < .0001), sexual behaviour (41% to 76%; p < .0001), seatbelt use (32% to 99%; p < .0001), and helmet use (28% to 92%; p < .0001). The aggregate counselling score across all five risk areas also increased significantly (39% to 89%; p < .0001). The increases from T1 to T3 were of a similar size and significance. Counselling rates in the areas of tobacco use, alcohol use, sexual behaviour, and helmet use did not change significantly between T2 and T3.

253


Table 54: Ozer et al. (2001) (continued) Authors conclusions

This study evaluated an intervention to increase clinicians’ delivery of preventive services to their adolescent patients in a group-model HMO. This intervention offers a promising approach to facilitate the implementation of national guidelines that recommend screening and counselling adolescents on a range of risky behaviours. Improving the delivery of preventive services is an essential step toward decreasing adolescent morbidity and mortality.

Reviewers notes The study adopted an internal comparison design with multiple assessment points. The lack of comparison sites presents the possibility that the changes in screening and counselling rates were a reflection of broader preventive care changes within the HMO.


Yes, the study looked at an intervention to increase the delivery of preventive services by clinicians (including tobacco use).




254


Table 55: Pbert et al. (2006) Citation Pbert, L., Fletcher, K. E., Flint, A. J., Young, M. H., Druker, S., & DiFranza, J. (2006).

Smoking prevention and cessation intervention delivery by pediatric providers, as assessed with patient exit interviews. Pediatrics, 118(3), e810-824.


To evaluate the degree to which a smoking prevention and cessation intervention was delivered by providers to adolescents in a paediatric office setting during medical care, based on adolescent reports in patient exit interviews (PEIs)

Study type/design RCT, Duration = 12 months, smoking specific. Participant group Participants: Eight paediatric clinics and 2710 patients’ participants.

Inclusion criteria: Not stated for providers, but stated for patients. Exclusion criteria: Not stated for providers. Subject disposition: (The eight paediatric clinics included 28 providers in UC group and 33 in SI group, majority of providers were women average age 42 years, 48 paediatricians, ten nurse practitioners, one physician assistant, two paediatric residents, three missing data). 2710 (1365 UC, 1337 SI) adolescents aged 13-17 years, smokers (smoked in the past 30 days) and former smokers.

Intervention Special Intervention (SI) = Training and reminders to deliver a brief smoking prevention and cessation intervention to adolescents in the context of routine paediatric primary care practice. The paediatric provider training program consisted of a one-hour group training tailored to the individual practice and held on-site in each paediatric office, followed in two to four weeks by an individual feedback session. The intervention algorithm was reviewed in detail and a demonstration was given.

Comparator Usual Care (UC). Providers did not receive any training. Outcome definitions Percentage of intervention steps completed by paediatric provider, as reported in

PEIs, according to condition, everyone was Asked if ever smoked or tried smoking, current smokers were asked if were interested in Quitting


Analyses: Sites were compared with respect to the characteristics of providers and adolescent participants with either x2 tests or Fisher’s exact test, when the analysis contained small cells, for categorical data (e.g. ethnicity) and t tests for dimensional data (e.g., age). Sample size calculation: Not stated. A. (1) Inadequately Reported B. (1) Inadequate C. (2) Adequate/Reported: Providers in the UC and SI conditions did not differ with respect to any characteristics assessed at baseline. D. (0) Not reported E. (2) Reported: Smoked in past 30 days Fisher’s exact p =.07, When provider asked if smoked X(2)= .48, p = .79. Only p values of < .003 are significant according to the Bonferroni correction. F. (0) Unknown/Not reported G. (1) Inadequate: No practices dropped out but, it is unclear if any physicians dropped out.



255


Table 55: Pbert et al. (2006) (continued) Results (within scope of this review)

Comparison of Interventions Delivered by Providers in the SI and UC Conditions results: When everyone was asked if ever smoked or tried smoking ‘A’, 77.0% (797/1035) of patients in the usual care reported they were asked if ever smoked or tried smoking by their provider as compared to 95.8% (1201/1253) of those in the special intervention group, X2 181.87, p < .0005 The percentage of providers engaging in the smoking interventions differed significantly between the special intervention and usual care conditions, according to adolescent reports in the exit interviews. For non-smokers/former smokers, overall PEI scores were 7.24 (SD: 1.55) for the special intervention condition and 4.95 (SD: 1.69) for the usual care condition [t(1900) = 30.40; p < .001]. For current smokers, overall PEI scores were 8.40 (SD: 1.93) and 6.24 (SD: 2.12) for the special intervention and usual care conditions, respectively [t(173) = 7.05; p < 0.001]. Intervention fidelity of special intervention providers was 72.2% and 84.0% for non-smokers/former smokers and current smokers, respectively.

Authors conclusions

Paediatric providers who receive training and reminders to deliver a brief smoking prevention and cessation intervention to adolescents in the context of routine paediatric primary care practice can do so feasibly and with a high degree of fidelity to the intervention protocol.

Reviewers notes Limitations of the study: The accuracy of adolescent recall of provider intervention behaviour, particularly the recollection of specific topics discussed. A systematic bias for reporting receipt of intervention among SI participants, compared with UC participants, might have occurred because of differences in the level of social desirability demand among the adolescents. The small number of sites for randomization might have decreased the comparability of the two conditions.


Employed a randomized controlled design to examine the effectiveness of training providers and using reminders to deliver smoking prevention and cessation intervention.




256


Table 56: Piper et al. (2003) Citation Piper, M. E., Fiore, M. C., Smith, S. S., Jorenby, D. E., Wilson, J. R., Zehner, M. E., et

al. (2003). Use of the vital sign stamp as a systematic screening tool to promote smoking cessation. Mayo Clin Proc, 78(6), 716-722.


To test three hypotheses: (1) The expanded vital sign assessment tool will result in increased smoker identification ‘A’ (2) Using the expanded vital signs will increase the rate at which physicians advise ‘B’, assist, and arrange follow-up ‘C’ (3) The expanded vital signs will not only boost delivery of smoking interventions, but will also increase patients’ long-term abstinence from tobacco.

Study type/design Randomised controlled trial (used pre-post design), duration = 47 months (Feb 1995 to Dec 1998), analysed/eligible = 46/130 35.3%, smoking specific

Participant group Participants: Five primary health care clinics in Madison, 130 physicians saw patients at both vital signs clinics and control clinics, only 18 analysable physicians at the vital signs clinics and 28 analysable physicians at control clinics. Analysable physician defined as physician who had had at least ten patients interviewed during both the baseline and intervention phase. Inclusion criteria: Not clearly stated, but clinics participated in baseline phase (no criteria stated) were eligible to participate in the intervention phase. For participants any adults willing to participate were eligible. Exclusion criteria: Not stated. Subject disposition: Five of six outpatient health care clinics (one dropped out as it implemented a vital sign stamp in its visit protocol), four were part of one health maintenance organization in central Wisconsin, one was a satellite of the University of Wisconsin clinics system. 1611 participants (current smokers).

Intervention Vital sign stamp to remind clinicians to ask about smoking status. (three clinics with 18 analysable physicians).

Clinics assigned to the vital sign condition used progress note pages stamped with expanded vital signs that included smoking status, blood pressure, temperature, and pulse and respiratory rate. Staff members at intervention clinics who usually take vital signs (medical assistants and nurses) received in-service training on how to assess and document smoking status (e.g. currently smoking, smoked in the past, or never smoked) as a part of the regular collection of vital signs.

Comparator Standard vital signs that did not include smoking status, and staff did not receive training. (two clinics with 28 analysable physicians)

Outcome definitions Baseline phase: Whether the physician has asked the participant about cigarette smoking status ‘A’, if currently smoked cigarettes, if planned to quit in the next 12 months ‘C’. Follow-up questions on mean number of cigarettes per day, whether made a serious quit attempt within the past year, and whether smoked in the past 7-days.


Analyses: Unit of analysis was clinic, changes in physician advice, assistance, and arranging follow-up visit (at baseline, at intervention, and the change). Sample size calculation: Not stated.

257


Table 56: Piper et al. (2003) (continued) A. (1) Inadequately reported: Stated that each clinic was randomly assigned to either the intervention or the control. B. (0) Not reported C. (2) Adequate. Baseline percentages for participants who were asked about smoking ‘asking’, identified smokers advised, current smokers given help by physician to set a quit date, current smokers prescribed, current smokers whose physician arranged follow-up. D. (1) Inadequate. Inclusion and exclusion criteria were not well defined. E. (2) Adequate. Baseline measure or estimates of changes in physician advice, assistance, and arranging follow-up visits 12 month pre-intervention (baseline) phase. F. (0) Unknown/Not reported G. (2) Adequate/Reported



First hypothesis: Proportion of all participants asked by any clinic staff person (physician, nurse, medical assistant) about smoking both before and after experimental manipulation of the vital sign stamp ‘A’. Results showed that there is an overall increase of 9.6% and 30.9% in the control and intervention clinics respectively. The mean increase in asking behaviour of physicians at the vital signs clinics was statistically significantly greater than that of physicians at control clinics (mean 13.76; SD 7.5 vs. mean 2.89; SD 12.9) t44 = -3.61; p = .002).

The second hypothesis: The expanded vital signs would increase the rates at which physicians advised identified smokers to quit ‘B’. Identified smokers were less likely to report being advised by their physician to quit during the post-intervention period compared with pre-intervention period (evident for four of the five clinics) whereas the fifth clinic showed no change.

No statistical difference between control and intervention clinics in the decline of physicians advising identified smokers to quit (t44=-0.33; p =.74). Overall, 58.7% of physicians who had seen at least ten patients during both the baseline and intervention phase decreased their rates of advising smokers to quit between the baseline survey and the intervention survey.

258


Table 56: Piper et al. (2003) (continued) Authors conclusions

A simple system-wide screening assessment tool, while effective in identifying more tobacco users, did not increase the rates at which physicians advised or assisted smokers to quit. Further system-wide changes may be needed to ensure that effective tobacco-dependence treatments are given to smokers.

Reviewers notes There was no change in the number of quit attempts and abstinence rates following the implementation of the vital sign stamp.


Relevant as the study examined the effectiveness of a system-wide screening assessment tool to increase rate of smoker identification ‘A’, physician advice to quit smoking ‘B’, and physician assistance in quitting and abstinence rates ‘C’.




259


Table 57: Richmond et al. (1998) Citation Richmond, R., Mendelsohn, C., & Kehoe, L. (1998). Family Physicians' Utilization of a

Brief Smoking Cessation Program Following Reinforcement Contact after Training: A Randomized Trial. Preventive Medicine, 27(1), 77-83.


Aim: to examine the effect of reinforcement contact following training, via follow-up phone calls.

Study type/design RCT; duration = 6 months, analysed/eligible = 5%, smoking specific. Participant group Participants: 198 family physicians (100control) – 98 (exp)).

Inclusion criteria: accepted invitations and participated in a workshop to be trained in smoking cessation techniques. Subject disposition: Sex (% male) 73 Working full time (%)87 In metropolitan areas (%)44 In solo practice (%)33 Age [mean (SD)] 42.5 (9.1) Years as a GP [mean (SD)] 13.0 (8.8) Estimated number of patients seen per week [mean (SD)] 143 (57).

Intervention A two hour workshop of training in a brief smoking cessation interventions and reinforcement contact from the medical practitioner trainer, consisting of three brief telephone calls at two weeks, two months, and four months after training, by a peer doctor.

Comparator No-intervention/ 2-hr workshop to be trained in a brief smoking cessation intervention but no reinforcement contact from the medical practitioner trainer.

Outcome definitions (1) Doctors completed self-report questionnaire, mailed at six months, which inquired about their current and future use of cessation interventions. (2) The number of smoking cessation information booklets given to smokers over the six-month study period.


Analyses: Pearson x2 test for proportions and the two-tailed t test for independent samples for continuous data; one-tailed t test for assessing differences between number of booklets distributed; α set at the 0.05. Sample size calculation: Sample size of 198; standardized effect size of 0.30 or greater; 80% power; alpha 0.05; projected loss to follow-up of 25%. A. (2) Adequate: Random numbers. B. (1) Inadequate: Unclear. Research assistant blinded to study group but unclear if the GPs were blinded to the intervention (i.e. the follow up or no follow up condition). C. (2) Adequate/Reported: Yes, participant characteristics tabulated with no significant differences between groups. D. (2) Adequate: Inclusion and exclusion criteria were defined. E. (0) Unknown/Not reported: No baseline measure or estimates of (‘untrained’) GPs’ smoking cessation intervention behaviour. F. (2) Adequate: Yes, not specified but apparent. G. (2) Adequate: The number of withdrawals and reason for withdrawal were stated.



(1) GP self-report of smoking cessation intervention/s delivered, at 6 months follow-up: 84%(control)/93%(exp); p = 0.06 (approached significance). (2) The mean number of booklets distributed to smokers by full-time physicians: 32.8; 95% CI 27.6 to 38.0 (control) vs. 40.1; 95% CI 35.4 to 44.8 (exp); p < 0.05 Note: follow-up abstinence rates also reported (but no comparative base-line data).

Authors conclusions

Three very brief phone calls in four months among full-time physicians were shown to significantly increase the distribution of smoking cessation information booklets given to patients. Follow-up phone calls after training increased GPs use of smoking cessation intervention/s, but statistical comparison only approached significance (p = .06).

260


Table 57: Richmond et al. (1998) (continued) Reviewers notes This study, conducted in Australia, demonstrates the potential of introducing

organisational or systems changes into prevention programmes in a general practice context similar to New Zealand.


Yes, relevant. Employed a controlled, randomized trial to examine organisational support mechanisms intended to facilitate the implementation of smoking cessation interventions by GPs to all tobacco users. However, the outcomes were not clearly delineated as ‘ABC’ but incorporated ‘ABC’ in essence.




261


Table 58: Roski et al. (2003) Citation Roski, J., Jeddeloh, R., An, L., Lando, H., Hannan, P., Hall, C., et al. (2003). The

impact of financial incentives and a patient registry on preventive care quality: increasing provider adherence to evidence-based smoking cessation practice guidelines. Preventive Medicine, 36(3), 291-299.


To test the effects of two organizational support processes, (1) the provision of financial incentives for superior clinical performance and (2) the availability of a patient (smoker) registry/cessation support and audit and feedback, on provider adherence to accepted practice guidelines and associated patient outcomes.

Study type/design RCT multi-centre; three group, 6 months, smoking specific. Participant group Participants: GPs in 40 clinics (10control) – 15 (exp1) – 15 (exp2)); (4813 baseline & 4734

patient exit surveys). Inclusion criteria: All 40 clinics sites providing primary care services (family practice, internal medicine, obstetrics/gynaecology) of a large Midwestern multi-specialty group practice. Exclusion criteria: Other clinics. Subject disposition: Not well defined: HCPs providing family practice, internal medicine and obstetrics/gynaecology services.

Intervention (1) “Incentive” group: Financial incentives for reaching preset clinical performance targets. (2) “Registry” group: Financial incentives and access to a centralised smoker registry and audit and weekly feedback on referral activity (performance feedback regarding their adherence to smoking cessation practice standards was provided in the same graph to all nine clinics in the registry experimental condition). Note: The threshold set to trigger a pay-out of financial incentives was: tobacco status clearly identified– 75%; advice to quit smoking – 65% (15 percentage points above the average baseline performance for these clinic practices). Incentives: clinics with one to seven providers could receive a $5,000 (US) award, and clinics with eight or more providers were eligible for a $10,000 (US) bonus per month.

Comparator No-intervention/distribution of printed versions of the smoking cessation guidelines only.

Outcome definitions (1) Clinical practice patterns: Tobacco user identification, providing advice to quit to current smokers, referral activity; measured in two rounds of brief, one-page cross-sectional patient pen-and pencil exit surveys. (2) Patient outcomes: Counselling, medication, quit rates.


Analyses: Analysis employed mixed-model analysis of covariance with clinics serving as the unit of analysis. Analyses of practice pattern changes (identification, offer of advice to quit) and patient outcomes (quitting) employed clustered logistic regression. Sample size calculation: Sample size estimation for the clustered design with 80% statistical power to detect differences of 2.6% in patient quit rates between the control and incentive conditions and a 6.5% difference between the control and registry conditions after allowance for multiple testing.

262


Table 58: Roski et al. (2003) (continued) A. (0) Unknown: No details of random sequence generation were provided. B. (2) Adequate: Unlikely that effective change mechanisms could have been instituted at control clinic sites simply listening to accounts of, or observation of, newly implemented change activities in intervention sites. C. (2) Adequate/Reported: The potential influence of several clinic-level covariates including the proportion of low-income patients, patients’ gender distribution, and clinic size (number of providers, average number of patient visits) was controlled by forming groups of similar clinics and their subsequent block randomisation. D. (0) Not Known: Inclusion and exclusion criteria were not well defined. E. (2) Adequate: Percentage of patients, tobacco use status identified at last visit; percentage of smokers who received advice to quit at last visit; percentage of smokers who were offered assistance to quit at last visit, and p values. F. (2) Adequate G. (2) Adequate: The number of withdrawals and reason for withdrawal were stated.


TOTAL: 10 points; Fair. Results (within scope of this review)

Identification of patients’ tobacco use status statistically significantly improved in the incentive (by 14.1%) and registry (by 8.1%) conditions over improvements in the control group (by 6.2%); p = .009. Clinical practice rates for advising smokers to quit and providing smokers with assistance to quit did not differ statistically significantly between the experimental conditions (p = .426; p = .451 respectively). Note: overall, tobacco user identification at last visit 49.8% at follow-up. Patients visiting registry clinics accessed counselling programs statistically significantly more often (p < .001) than patients receiving care in the control condition (patient outcome: ‘accessed’ not ‘provision of’ counselling).

Authors conclusions

This study was able to detect a modest positive impact of these processes on clinicians’ adherence to smoking cessation practice guidelines (identification of tobacco users). It is not clear whether significantly higher incentive payments would have been able to focus clinic sites’ attention more strongly on the need to provide smokers with advice and assistance to quit.

Reviewers notes The characteristics of the HCPs were not defined. It is assumed that the participants were GPs, other clinic staff may or may not have been able to influence clinic performance. Note that identification rates did not increase to the same extent in the registry clinics, even though these clinics were offered the same financial incentives (the attention of all physicians and staff in the participating clinic sites may have been diverted in multiple ways from implementing robust office systems)..


Yes, relevant. Employed a controlled, randomised trial to examine such organisational support mechanisms intended to facilitate health care improvement. Note: also reported patient tobacco use endpoints.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; BMI = body mass index; CBA = controlled before-after trial; CCT = controlled clinical trials; CFPC = College of Family Physicians of Canada; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FFS = fee for service; FP = family practitioner; GEEs = generalised estimating equations; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; eLinkS = electronic linkage system; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; ORP = Outcomes Recognition Programme; PAR = practice assessment report; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; TS = time-series design; UC = usual care; UMMS = University of Massachusetts Medical School; V.A. = Veterans Affairs clinic setting.

263


Table 59: Rothemich et al. (2008) Citation Rothemich, S. F., Woolf, S. H., Johnson, R. E., Burgett, A. E., Flores, S. K., Marsland,

D. W., et al. (2008). Effect on cessation counseling of documenting smoking status as a routine vital sign: An ACORN study. Annals of Family Medicine, 6(1), 60-68.


To assess whether the vital sign intervention influences patient-reported frequency and intensity of tobacco cessation counselling. The study was designed to examine not only the overall counselling rates (A2-5) but also the intensity of counselling (A2 vs A3-5). A1 Ask Identify and document tobacco use status for every patient at every visit. A2 Advise In a clear, strong, and personalized manner, urge every tobacco user to quit. A3 Assess Determine whether the tobacco user is willing to make a quit attempt at this time. A4 Assist For the patient willing to make a quit attempt, use counselling and pharmacotherapy to help him or her quit. A5 Arrange Schedule follow-up contact, in person or by telephone, preferably within the first week after the quit date.

Study type/design Cluster-randomised controlled trial, duration = 6 months, analysed/eligible = 8/84=21.4%, smoking specific.

Participant group Participants: 18 practices (73 clinicians) within a 25-mile radius of downtown Richmond, Virginia, 6,729 adult patients. Inclusion criteria: Practices with the full-time equivalent of at least two clinicians (family physicians, general internists, nurse-practitioners, or physician’s assistants) who specialized in primary care and who provided care to adult patients (wholly or in part) were eligible. Exclusion criteria: Exclusions were: the specialist clinician from the study, residency programs, clinics serving special populations (e.g., urgent care centres, indigent/free clinics, student health centres), practices not under the auspices of the overseeing institutional review boards, or practices with an existing systematic tobacco identification and reminder system. Subject disposition: N = 73 clinicians (14 family medicine practices, three internal medicine practices, and one practice featuring both specialties. Two practices were urban, two were rural, and 14 were suburban). Only one practice used electronic medical records). Control protocol allocated and received: nine practices with 40 clinicians (median = 4, range 2-7). Intervention protocol allocated and received: nine practices with 33 clinicians (median = 3, range 2-6)

Intervention Nurses and medical assistants were instructed to assess the tobacco use status of every adult patient and record it with the traditional vital signs.

Comparator Did not use any systematic tobacco screening or identification system. Outcome definitions Main outcome measure (A2-5): Proportion of smokers reporting clinician counselling

activity of any kind and the frequency of two counselling subcomponents: simple quit advice ‘B’ and more intensive discussion ‘C’. Proxy for intervention: Ask ‘A’

264


Table 59: Rothemich et al. (2008) (continued) Data analyses & statistics

Analyses: The unit of analysis was the practice, because practices rather than patients or clinicians were randomised. Control Clusters analysed: nine practices with 39 clinicians, (median = 4, range 2-7) and 3,848 exit surveys (median = 426, range 296-560). Intervention Clusters analysed: nine practices with 32 clinicians (median = 3, range 2-6) and 2,881 exit surveys (median = 333, range 241-430). Clusters excluded from analysis: zero practices, but one clinician not represented in both pre-intervention and comparison periods. Sample size calculation: Simulation predicted 80% power to detect a 12% effect size if enrolled 18 practices and surveyed 27 pre-intervention period smokers and 81 comparison period smokers per practice. A. (2) Adequate/Reported: Used Microsoft Excel and a random number generator to randomize practices within the matched groups to the intervention or control condition. B. (1) Inadequate: Practices were offered $500 for participation and an additional $1,000 if assigned to the intervention group. C. (2) Adequate/Reported: Baseline characteristics on the % of patients receiving cessation counselling (the 5A’s) Asked if smoking (A1), counselled to quit (A2-5), given simple advice (A2)d, and had more intensive discussion (A3-5)d. D. (2) Adequate/Reported E. (1) Unknown/Not reported: Across the 18 practices, 6,729 patients (1,149 smokers), including 3,848 patients (588 smokers) from control practices and 2,881 patients (561 smokers) from intervention practices, completed exit questionnaires during the comparison period. F. (2) Unknown/Not reported: Used an intention-to-treat analysis; at the practice that experienced a two-month delay in implementing the intervention. G. (2) Adequate/Reported:


TOTAL: 12 points; good Results (within scope of this review)

A total of 6,729 adult patients (1,149 smokers) at 18 primary care practices completed exit questionnaires during a 6-month comparison period. Ask ‘A’: "Did a nurse or doctor ask you today if you smoke?” For all patients: Control group 26, 3% (yes), intervention 66.0% (yes) (difference of 39.7, 95% CI 35.5 to α) p = .001. For smokers: 49.4% of the patients in the control group stated that their nurse or doctor asked them about their smoking status as compared to 79.5% of the intervention group (difference of 30.1, 95% CI 24.0 to α) p < .001 Counselling ‘B’ & ‘C’ either or both of the following: “If you smoke, did your doctor advise you today to stop smoking?” ‘B’ “If you smoke, did your doctor talk with you today to help you quit smoking?” ‘C’ Among 561 smokers at intervention practices, 61.9% reported receiving any counselling, compared with 53.4% of the 588 smokers at control practices, for a difference of 8.6% (p = .04) ‘B’. The effect was largely restricted to simple advice, which was reported by 59.9% of intervention patients and 51.5% of control patients (p = .04). There was no significant increase in more extensive discussion ‘C’, with 32.5% and 29.3% of patients at intervention and control practices, respectively, reporting this type of counselling (p = .18).

265


Table 59: Rothemich et al. (2008) (continued)

Authors conclusions

The vital sign intervention promotes tobacco counselling at primary care practices through a modest increase in simple advice to quit. When implemented as a stand-alone intervention, it does not appear to increase intensive counselling.

Reviewers notes Practices were offered $500 for participation and an additional $1,000 if assigned to the intervention group. Authors commented on randomisation: The randomisation procedure assigned nine practices to the vital sign intervention, and there was no attrition of intervention or control practices before or after randomisation. Relied on patient report. Funding support: The study was funded through two grants to Dr Rothemich from the Robert Wood Johnson Foundation, including a Generalist Physician Faculty Scholars Program award (036798) and a supplemental grant (043145).


Yes relevant as it assessed effect of vital sign intervention on delivery by clinicians of smoking counselling.




266


Table 60: Secker-Walker et al. (1992) Citation Secker-Walker, R. H., Solomon, L. J., Flynn, B. S., LePage, S. S., Crammond, J. E.,

Worden, J. K. & Mead, P. B. (1992) Training obstetric and family practice residents to give smoking cessation advice during prenatal care. American Journal of Obstetrics and Gynecology, 166, 1356–1363.


To evaluate the effectiveness of training obstetric and family practice residents to provide smoking cessation advice (in terms of the manner and consistency with which the smoking cessation protocol was followed, and the immediate impact on the women’s agreement to change their smoking behaviour).

Study type/design Pre-and-post design for the healthcare professionals, as part of an ongoing randomised controlled trial that assessed effectiveness of smoking cessation advice provided to pregnant smoking women, duration = 24 months, analysed/eligible = 20/35 (57.14%), smoking specific.

Participant group Participants: Residents (n = 35), 250 pregnant women (40-45% smokers at the onset of pregnancy). Inclusion criteria: Not stated for the residents. Exclusion criteria: Not stated for the residents. Subject disposition: 19 Obstetric residents (14 female), 16 family physician residents (11 female) at the University of Vermont, in Maternal-Infant Care clinic, provides prenatal care to women with little or no insurance. Pregnant women (smokers).

Intervention Single-component educational training for residents in obstetrics and family practices to provide smoking cessation advice during prenatal care. This is a one-hour workshop + refresher sessions and progress. Each year the residents trained by a physician and psychologist in small groups during one-hour workshop. Review of project, description of advice to be given to subjects in intervention condition, use of specially designed protocol prompt sheet to guide the advice, videotaped demonstration of the advice being delivered by a family physician and an obstetrician, a role play of the advice by pairs of residents with corrective feedback, and a description of the basic care advice to be given to control group subjects. Refresher sessions and progress (30-minute individual refresher session with the psychologist a few days before starting the first rotation on the Maternal-Infant Care clinic). Intervention and basic care protocols reviewed, intervention protocol role-played with simulated patient, and appropriate feedback provided to the resident. Each month the progress of the trial and residents’ adherence to the protocol were discussed by the smoking cessation counsellor during Maternal-Infant Care clinic conferences.

Comparator Before and after the training of residents providing both the intervention group (of pregnant women) and the control group (of pregnant women on the Basic care).

Outcome definitions Video rating (comparisons of pre-training and post-training videotapes), adherence to protocol. Comparing and rating (through seven criteria) videotapes of the residents providing their usual advice to a simulated patient before training, with videotapes of each resident delivering the intervention protocol to the same simulated patient at the end of their refresher session.

267


Table 60: Secker-Walker et al. (1992) (continued) Data analyses & statistics

Tapes scored independently by two trained raters blind to the training status of the resident. Analyses: Videotape scores and coded data from exit interviews of pregnant smoking women taking part in RCT of smoking cessation advice analysed using paired student test t to compare pre-and post training videotapes. X2 analyses used for comparing the unit of analysis (pregnant smoking women) from both groups. Proportion of patients advised by trainee to stop smoking or stay quit during the second ante-partum visit following intervention reported at exit visit. Sample size calculation: Not stated. A. (0) Not Reported: Not stated for the residents, but for pregnant women. B. (1) Inadequate: Treatment allocation for pregnant women but not for the residents, however, assessment was blind (tapes scored by two people who were blind to the training status of the resident). C. (1) Inadequate: No information was provided for the characteristics of the residents apart from the gender. Information for pregnant women was detailed on demographics, pregnancy status, and smoking history. D. (0) Not reported: Not stated for the residents, but for pregnant women. E. (2) Adequately reported: Reasonably well reported. F. (0) Unknown/Not reported G. (2) Study was ongoing when published


TOTAL: 6 points; fair Results (within scope of this review)

Evaluation of resident training was in three ways: Immediate impact of training on content and style of each resident’s smoking cessation advice with first group of residents trained in 1988 (12 obstetric residents, all eight first and second year family practice residents), this was by comparison of the videotapes. Each trained resident’s adherence to the protocols was assessed from exit interviews with all women enrolled in the trial immediately after the first and second ante-partum visits with the physicians and before the women in the intervention groups received counselling, were conducted by a research assistant and took < 2 minutes to complete five questions about what woman agreed to do about her smoking behaviour. Comparing responses of women in intervention and basic groups to the two questions asking what they had agreed to do about changing their smoking behaviour. Comparisons of the first 20 residents showed: ‘A’ Ask, ‘B’ Brief advice, and Cessation ‘C’ (all included in protocol prompt sheet): Significant effect for training on content and style of smoking cessation advice but not on time taken to deliver the advice. Average score increased from 2.15 (SD .88) to 6.05 (SD .69), t = - 16.28, p = .0001. ‘A’ Ask about smoking: 98% and 96% (first and second ante-partum visit) the intervention group vs. 95% and 66% (first and second ante-partum visit) the basic care group respectively (p = .0001). ‘B’ Brief advice: 96% and 91% (first and second ante-partum visit) the intervention group vs. 92% and 52% (first and second ante-partum visit) the basic care group respectively (p = .0001). Long-term adherence to intervention and basic care protocols was assessed by comparing the responses to the exit interviews conducted during the first year of the trial with those conducted in the second year. There were no significant differences within the groups found.

268


Table 60: Secker-Walker et al. (1992) (continued) Authors conclusions

The structured advice consistently provided by the trained residents was effective in gaining commitments from pregnant smokers to change their smoking behaviour.

Reviewers notes The study was still progressing at the publication date, so involved first 20 residents in the analysis. Also the study was included in the systematic review by Anderson 2004. Although the study involved Obstetricians (not GP) they were those who provide antenatal care for pregnant women.


Yes, it is relevant as it assessed the effectiveness of an educational training for residents in obstetrics and family practices to provide smoking cessation advice during prenatal care.




269


Table 61: Solberg et al. (2000) Citation Solberg, L. I., Kottke, T. E., Brekke, M. L., Magnan, S., Davidson, G., Calomeni, C. A.,

et al. (2000). Failure of a continuous quality improvement intervention to increase the delivery of preventive services. A randomized trial. Effective Clinical Practice, 3(3), 105-115.


The hypothesis: that a managed care organization, working through an iterative process of change called continuous quality improvement, can stimulate its private primary care clinics to develop better delivery systems for eight clinical preventive services: blood pressure monitoring, Pap smear, cholesterol monitoring, tobacco use cessation, breast examination, mammography, influenza vaccine, and pneumococcal vaccine.

Study type/design Cluster randomised controlled trial, duration = baseline 5 weeks; 22 months intervention; 5 weeks post-intervention, analysed/eligible = 44/164 = 27%, general prevention including smoking ‘A’ (and seven other preventive services).

Participant group Participants: All Health Care Professionals from 44 primary care clinics in greater Minneapolis-St. Paul (including approximately 500 primary care Doctors). Inclusion criteria: Part of a medical group that contracted with one or both of the sponsoring HMOs; were located within 50 miles of the Twin Cities; management at each clinic agreed to establish clinic guidelines for the study's preventive services and to identify and support a management sponsor and an internal multidisciplinary improvement team for those services. Exclusion criteria: Already involved in a CQI or other prevention programme or research study. Subject disposition: On average, 12 primary M.D.s; 1.5 mid-level practitioners; 13 nurses and 34 other staff per clinic.

Intervention The IMPROVE project: Through leadership support, training, consulting, and networking, each intervention clinic was assisted to use CQI multidisciplinary teams to develop and implement system changes. Including: a leader and facilitator for each clinic's team, an initial six-hour conference overview of CQI methods and prevention systems, six four-hour workshops over six months for experiential training in an improvement process to build prevention systems (reinforced with modular manuals); consultation by telephone (every six weeks) and by visit (every three months); bimonthly breakfast meetings and a quarterly newsletters to clinic staffs.

Comparator No intervention/usual care. Outcome definitions The proportion of patients who were up-to-date for each service within three months of

an index primary care visit on the basis of chart audit and the proportion of patients who were offered a service if not up-to-date (according to Patient Recent Visit Survey).


Analyses: Patients selected from appointment lists for randomly identified days (patients 19 years of age and older completed surveys at baseline [n = 6830] and at follow-up [n = 6431]). Outcome measures were calculated as a grand mean of the rates at which each service was performed in the intervention and control groups. Adjusted for the potential clustering effects of clinics as well as for patient age, sex, health status, socioeconomic status, ethnicity, and purpose for visit using multilevel logistic regression analysis (adjusted rates were nearly identical to the unadjusted rates). Sample size calculation: 22 clinics in each group and 175 patients per clinic; to detect a minimum between-group difference of 15% for rates of delivering preventive services, with a power of 0.80 and a type 1 error of 0.05 (two-tailed) (then inflated sample size by 20% in anticipation of survey return rates of less than 100%, resulting in slightly greater power).

270


Table 61: Solberg et al. (2000) (continued) A. (2) Adequate/Reported: All 44 clinics grouped into one of eight unique cells and matched pairs of clinics then randomised (matched on size, readiness to improve preventive services, and existing quality-based organizational culture). B. (1) Inadequate: Treatment allocation not concealed (not possible), however, outcome assessment (i.e. chart audits and PEIs) concealed by study group. C. (2) Adequate/Reported: The 22 intervention clinics and the 22 control clinics were identical in terms of randomisation variables and other clinic characteristics. D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: Yes, baseline rates on smoking “Ask” ≈75% (and all other 7 preventive services were reported). F. (2) Adequate/Reported: Conducted an intention-to-treat analysis, using all 22 clinics in each group. G. (1) Inadequate: No clinics dropped out, however, individual HCPs not specifically accounted for.



Ask: 8.6% increase (control) vs. 15% increase (exp) [p = n.s.] Note: only one of the selected preventive services (pneumococcal vaccine) had a significantly greater increase in the intervention group (17.2% absolute increase from baseline to follow-up compared with a 0.3% absolute increase in the control group; p = .003).

Authors conclusions

In this trial, CQI methods did not result in clinically important increases in preventive service delivery rates; the actual differences observed were mostly so small that they would not be clinically important enough to warrant the amount of effort expended, even if they had been statistically significant. Generally, teams did not complete or repeat the improvement cycle, were slow to implement changes, and usually implemented incomplete changes. CQI may be an inappropriate mechanism for making preventive services improvements.

Reviewers notes At the time of the follow-up survey and audit, six clinics had not implemented any system changes. The (suggested) critical process of systematically prompting or reminding the clinician about missing prevention services during a visit were only in place for a few services, in a few of the intervention clinics. If office systems are an important vehicle for improvement in the delivery of clinical preventive services, then, as Solberg et al. suggest, the critical unanswered question is “How might these systems be established and maintained in typical practice settings without research-funded support”? Note that for “Ask”, the study clinics already performed at a relatively high level (≈75%), therefore, to some degree, this particular outcome may have been subject to a ceiling effect.


Yes, employs a large cluster randomised controlled trial of moderately long duration to test CQI vs. usual care for smoking “Ask” and seven other preventive services.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FFS = fee for service; FP = family practitioner; GEEs = generalised estimating equations; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; ORP = Outcomes Recognition Programme; PAR = practice assessment report; PCTs = Primary Care Trusts; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; TS = time-series design; UC = usual care; UMMS = University of Massachusetts Medical School; V.A. = Veterans Affairs clinic setting.

271


Table 62: Strecher et al. (1991) Citation Strecher, V. J., O'Malley, M. S., Villagra, V. G., Campbell, E. E., Gonzalez, J. J., Irons,

T. G., et al. (1991). Can residents be trained to counsel patients about quitting smoking? Results from a randomized trial. J Gen Intern Med, 6(1), 9-17.


To evaluate the effectiveness of training and prompting under realistic conditions, including: the use of simple generalisable interventions; training conducted by existing faculty; and evaluation at several sites with residents from three primary care specialties.

Study type/design Randomised factorial four group design (Tutorial & prompt [n = 66] vs. Tutorial only [n = 66] vs. Prompt only [n = 60] vs. control [n = 58]), duration = 6 months, analysed/eligible = 89%, smoking specific.

Participant group Participants: Physicians, n = 261. Inclusion criteria: All resident physicians from 11 primary care programmes in three university medical centres. Exclusion criteria: Not completing a pre-test questionnaire (pre-randomisation); not working at least half day in ambulatory care setting. Subject disposition: 157 medical, 52 family practice, 25 paediatrics, mean number per clinic = 22; 31% postgraduate year one (PGY1), 36% (PGY2), 33% (PGY3).

Intervention Tutorial: Two sessions in smoking cessation counselling including slide show/lecture, video, handouts/flowchart, group discussion delivered by the clinic director or another educator. Prompt: Chart-based reminders to assist physicians to provide counselling (flow-chart placed on patients’ records by clinic staff [nurses] at patent check-in).

Comparator Tutorial and prompt vs. tutorial only vs. prompt only vs. no intervention. Outcome definitions ‘B’ The percentage of smokers provided with brief advice (as measured by researcher

administered patient exit interview). ‘C’ The occurrence of five specific cessation support techniques used by physicians: setting a quit date, prescribing a quit date, prescribing NRT, giving self-help material, providing follow-up (as measured by blinded, researcher administered telephone interview). ‘Q’ Patient reported smoking status six months after exit interview (telephone follow-up).


Analyses: chi-square, t-tests ANOVA, ANCOVA. Sample size calculation: Not stated. A. (2) Adequate/Reported: Computer randomisation of physicians. B. (2) Inadequate: Blinding of assessors to group allocation. C. (1) Inadequate: Stated participation did not differ by study group or site (characteristics not specifically reported). D. (2) Adequate/Reported: Yes. E. (2) Adequate/Reported: Point estimates provided by physician pre-test questionnaire for cessation support (63-70%) and use of cessation support techniques/smoker (0.5-0.7) and corroborated by 937 patient exit interviews. F. (2) Adequate/Reported: Yes. G. (0) Unknown/Not reported: Not clear.



272


Table 62: Strecher et al. (1991)(continued) Results (within scope of this review)

The mean percentages of return patient smokers advised to quit increased by 16 % for the tutorial + prompt group and 14 % for the tutorial only group but only by 6% for the prompt only group and 2% for the control group. Patient exit interviews [n = 937], using ANCOVA and adjusting for physician speciality: ‘Brief Advice’ (tutorial) – mean frequency of brief advice given = 53 %(non-tutorial) vs. 62 % (tutorial); p < .05. ‘Brief Advice’ (prompt only) – mean frequency of brief advice given = 57 %(non-prompt) vs. 62 % (tutorial); p < .05. ‘Cessation’ (tutorial) - mean number of techniques used/patient = 0.3(non-tutorial) vs. 0.6 (tutorial); p < .05. ‘Cessation’ (prompt only) - mean number of techniques used/patient = 0.4(control) vs. 0.5 (prompt only); n.s. No interaction effects. ‘Quit’ – (as measured by biochemical test): tutorial & prompt [5.6%]; tutorial only [3.4%]; prompt only [5.7%]; control [1.7%]: all n.s.

*Outcomes Control Tutorial Prompt Tutorial & prompt

Ask ‘A’ Assumed 100%– clinic nurses identified patients who smoked

Not measured

Pre-test

67% 64% 70% 63% † Advice ‘B’ Post-

test 69% § 78% 76% § 79%

Pre-test

0.7 0.6 0.5 0.5 ‡ Cessation ‘C’ Post-

test 0.9 ∆1.5 0.9 ∆1.9

Quit rates ‘Q’ 1.7% 3.4% 5.7% 5.6% All non-significant

*While patient outcomes are outside the scope of this review, quit rates (only) have been reported here. However, aspects of the included trial that relate specifically to patient outcomes have not necessarily been appraised or synthesised further. Further details of patient outcomes are available in the original authors’ publication. † The percentage of smokers provided with brief advice, as measured by researcher-administered patient exit interview. ‡ The mean number of techniques used: the occurrence of five specific cessation support techniques used by physicians: setting a quit date, prescribing a quit date, prescribing NRT, giving self-help material, providing follow-up, as measured by blinded, researcher administered telephone interview. ‘Q’ Patient reported smoking status six months after exit interview (telephone follow-up, n = 937). § Statistically significant difference, compared to the control group, at the p < .05 level. ∆ Statistically significant difference, all Tutorial compared to non-Tutorial, at the p < .001 level.

Authors conclusions

In a large randomised trial involving 234 residents from 11 training programmes and three specialties, a simple, feasible teaching programme, the tutorial, significantly increased residents’ self-reported smoking-cessation counselling practices and attitudes. Whether used alone or in combination with the tutorial, the prompt, a second intervention using a chart based reminder, had less effect on residents counselling practices.

Reviewers notes The ‘Ask’ was performed by non-physician clinic staff (nurses). This can be considered an intervention in itself. However, there was no measure or report of the percentage of all clinic patients asked their smoking status (assumed to be 100% but unlikely). Note: the largest changes occurred in setting a quit date and prescribing NRT but little change in actual physician counselling per se.

273


Table 62: Strecher et al. (1991) (continued) Relevance to study question

Yes. The pragmatic four group RCT design used existing staff for training (not smoking cessation experts) and existing nursing staff to administer the prompt with little opportunity for physicians to reinforce counselling.




274


Table 63: Tremblay et al. (2001) Citation Tremblay, M., Gervais, A., Lacroix, C., O'Loughlin, J., Makni, H., & Paradis, G. (2001).

Physicians taking action against smoking: An intervention program to optimize smoking cessation counselling by Montreal general practitioners. Canadian Medical Association Journal, 165(5), 601-607.

Level of evidence * Level IV Country Canada Research question/aims

The paper describes the theoretical model underlying the Physician Taking Action Against Smoking programme, as well as selected components of the programmed implemented during its first three years. The results of two cross-sectional surveys of independent random samples of GPs in 1998 and 2000, to monitor cessation counselling practices, are also reported.

Study type/design Two cross-sectional surveys, duration = two surveys, first four months (1998 from April to July), second six months (2000 from April to September), analysed/eligible = 76.6% (for 1998) and 69.6% (for 2000), smoking specific.

Participant group Participants: GPs practising in Montreal in 1997. Inclusion criteria: Although not clearly stated, authors noted that the target population for the programme comprises an 2130 GPs practising in Montreal in 1997: 165 GPs in 29 local community health centres, 1062 GPs in group practices and 903 GPs in solo practices. Exclusion criteria: Not stated. Subject disposition: Over the first three years of the programme, 180 GPs, 37 specialists, 27 residents and 373 other health care professionals participated in a total of 32 one-to-two-hour intensive training workshops designed to increase knowledge about and skills in smoking cessation counselling. Two 30-minute conferences on smoking cessation counselling provided in conjunction with continuing medical education activities (total of 239 GPs participated in these conferences).

Intervention The interventions were: Thirty-two one-to-two hour workshops about smoking cessation counselling. Two 30-minute conferences on smoking cessation counselling. Publication of 11 articles about smoking cessation counselling in journals for Quebec physicians, Publication of clinical guidelines for smoking cessation, Production and distribution of 2-sided plastic-coated card presenting a summary, in algorithm format, of smoking cessation interventions, Publication and distribution of a smoking cessation guide, presented in a cigarette pack format, Publication and distribution of Breath Free for Life, a list of smoking cessation in Montreal, Support of involvement of the College des medicines du Quebec in awareness to campaigns by Quebec Council on Tobacco and Health to encourage patients to stop smoking and to ask for help from their physicians, Collaboration in the publication of two newspaper articles, and two articles in a consumer magazine to reinforce the importance of physicians in smoking cessation and to encourage smokers to ask for help from their physicians.

Comparator The interventions in the survey conducted in 2000 compared to 1998.

275


Table 63: Tremblay et al. (2001) (continued) Outcome definitions Monitoring the smoking cessation counselling practices of GPS:

Two cross-sectional surveys (using self-administered questionnaires) conducted to monitor trends in the smoking cessation counselling practices of Montreal GPS and trends in psychosocial factors related to counselling. (One in the period April-July 1998, the second in April-September 2000).


Analyses: In 1998, of 440 GPs randomly selected from the 1997 College des medicines du Quebec database for Montreal and eligible to participate, 337 (76.6%) completed the questionnaire. The 2000 survey was completed by 316 (69.6%) of 454 randomly selected GPs. Sample size calculation: Not stated. A. (1) Inadequately reported: Reported but method of randomisation not stated. B. (0) Not reported: detailed descriptions of subject selection, sampling, data collection and study variables were referred to only; available from another paper written in French. C. (2) Adequate/Reported: Respondents to the 1998 survey younger than respondents in 2000 (mean age 45.3 (SD 9.5) vs. 47.8 (SD 10.4) years; p = 0.002). No other significant differences. D. (0) Not reported E. (1) Inadequately reported: Beliefs and attitudes of GPs toward cessation counselling, changes in psychosocial aspects of counselling over time. F. (1) Inadequate: Missing data were excluded from the denominators. G. (0) Unknown/Not reported


TOTAL: (5) points; Fair Results (within scope of this review)

Smoking cessation counselling practices of GPs in the two surveys: Usually uses a system to identify smokers ‘A’ (statistically significant difference in 1998 and 2000 between male and female physicians (p < .005). Cessation counselling coverage (offers cessation counselling to more than half, at least one in two visits, or devotes two minutes or more to cessation counselling) ‘B’ Offers support to more than half of patients at various stages of smoking cessation ‘C’; this includes the following GP activities: Offers written educational materials, discusses strategies for quitting and discusses setting a quit date, and refers patient to community resources (statistically significant difference in 2000 between male and female physicians (p < .005).

Authors conclusions

Our experience suggests that an integrated, theory-based programme to improve physicians’ counselling practices could be a key component of a comprehensive strategy to reduce tobacco use.

Reviewers notes The study is of evidence level IV as it compared results from two surveys over a time period.


Yes, it is relevant as it looked at using the intervention to improve counselling practices among GPs.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; CFPC = College of Family Physicians of Canada; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FFS = fee for service; FP = family practitioner; GP = General Practitioner; EHR = electronic health record; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; PEI = patient exit interview; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SSCS = specialist smoking cessation services; TS = time-series design; UC = usual care.

276


Table 64: Unrod et al. (2007) Citation Unrod, M., Smith, M., Spring, B., DePue, J., Redd, W., & Winkel, G. (2007).

Randomized controlled trial of a computer-based, tailored intervention to increase smoking cessation counselling by primary care physicians. Journal of general internal medicine : official journal of the Society for Research and Education in Primary Care Internal Medicine, 22(4), 478-484.


To test the effectiveness of a brief computer-tailored intervention designed to increase provision of smoking cessation counselling by primary care physicians and the quit rate among patients who smoke.

Study type/design Randomised controlled trial, duration = 6 months, smoking specific. Participant group Participants: 70 physicians (35 intervention, 35 control) and 518 patients.

Inclusion criteria: Internal or family medicine specialty, plans to continue practicing in current location for at least one year, at least 75 patient visits per week, primarily English speaking patients, and fewer than 25% geriatric patients. Exclusion criteria: Not stated. Subject disposition: 579 physicians contacted to participate, 75 (13%) not eligible, 262 (45%) could not be reached, 184 (32%) refused. Physician participants were: 83% Caucasians, 13% Asians, 1% African Americans; Average age 51.1 (SD = 8.1); 74% males; Average years post-medical school 24.3 (SD = 8.6); Physician specialty 68% Internal Medicine, 32% Family Medicine; “Current smokers” (4%), “never smokers” (67%), “former smokers” (29%).

Intervention Smoking Cessation Expert-System Intervention that integrates a brief, tailored expert-system report with face-to-face physician-delivered counselling in a primary care setting. The intervention had two objectives: (1) to bolster the rate at which physicians delivered smoking cessation services and (2) to increase patients’ quit rates. The training was conducted during a 40-minute visit to the physician’s office and followed an “academic detailing” approach.

Comparator Physicians in the control condition were not given any training and were instructed to continue their usual smoking cessation practices.

Outcome definitions Physician performance of the 5As was assessed using patient exit interviews (PEI) immediately after the visit. Quit rates and smoking behaviours were assessed six months post intervention through patient phone interviews.


Analyses: Physician performance of each of the 5As was analysed using a hierarchic generalized linear model analysis of variance, controlling for baseline variables identified as covariates. Sample size calculation: Assumed a 12.5 and 7.5% quit rate in the treatment and control groups, respectively. For power = .80 and alpha = .05, the N required to detect this low-moderate size group difference is 564.

277


Table 64: Unrod et al. (2007) (continued) A. (2) Adequate/Reported: A random number generator was used to assign physicians to intervention or usual care control. B. (1) Reported: Physicians learned their group assignment after signing the informed consent. C. (2) Adequate/Reported: No differences between physicians randomised to intervention and control groups on any measured demographic or medical practice factor. D. (2) Adequate/Reported E. (1) Inadequately reported: Not adequately reported for physician (physician performance of each of the 5As was analysed using a hierarchic generalised linear model analysis of variance, controlling for baseline variables identified as covariates). For patients authors stated that patient 7-day point-prevalence abstinence was analysed via a generalised linear model using Logit Link function, with physician as clustering variable. F. (1) Not reported: But analyses done immediately after the visit and no physician dropped G. (1) Inadequate



Physician 5A Performance (equated to ‘ABC’): Linear modelling indicated that: Intervention physicians exceeded controls on “Assess”, “Advise”, “Assist”, and “Arrange” (p < .0001). Specific results for (‘ABC’) are: More intervention than control physicians advised their patients to quit smoking (OR 2.79; 95% CI 1.70, 4.59) ‘B’. Per patient report, twice as many intervention as control physicians assessed patients’ readiness to quit smoking (OR 5.06; 95% CI 3.22, 7.95) ‘C’ Two to four times as many intervention as control physicians provided quit smoking assistance including: set goals (OR 4.31; 95% CI 2.59, 7.16), provided written materials (OR 5.14; 95% CI 2.60, 10.14), discussed smoking cessation medication (OR 4.72; 95% CI 2.90, 7.68), and referred to a quit-smoking program (OR 6.48; 95% CI 3.11, 13.49). Intervention physicians arranged a follow-up to discuss smoking at five times the rate of control physicians (OR 8.14; 95% CI 3.98, 16.68) ‘C’. The Ask ‘A’ results were not statistically significant Other results for quit rate and abstinence: Intervention patients were 1.77 (CI 0.94, 3.34, p = .078) times more likely than controls to be abstinent (12 versus 8%), a difference that approached, but did not reach statistical significance, and surpassed controls on number of days quit (18.4 versus 12.2, p < .05) but not on number of quit attempts.

Authors conclusions

The current study shows that a brief computer-tailored intervention can significantly increase primary care physicians’ implementation of the 5As and result in modest effects on smoking outcomes. This intervention holds promise in reaching the main objective of the Smoking Cessation Clinical Practice Guideline: to intervene with all smokers who visit their physician. Increased rates of guideline implementation can potentially have an enormous public health impact. Further research is needed to increase the potency of the intervention with regard to quit rates and to determine how best to integrate it into outpatient health care settings.

Reviewers notes Adherence to the Smoking Cessation Clinical Guideline (Delivery of the 5As) Used patient report. All physicians were paid $150, and those in the intervention group received an addition of $50 in recognition of greater time commitments expected of them. Smokers were paid $20 for completing initial assessments and $10 for the follow-up interview. Physicians learned their group assignment after signing the informed consent (so physicians were aware of their allocation to intervention (knowledge to exposure) which makes the study liable to differential bias.

278


Table 64: Unrod et al. (2007) (continued) Relevance to study question

Yes, relevant as the study assessed the effectiveness of an intervention to enhance adherence to smoking cessation guidelines by primary care physicians.




279


Table 65: Wadland et al. (2007) Citation Wadland, W. C., Holtrop, J. S., Weismantel, D., Pathak, P. K., Fadel, H., & Powell, J.

(2007). Practice-based referrals to a tobacco cessation quit line: assessing the impact of comparative feedback vs. general reminders. Annals of Family Medicine, 5, 135-142.


To assess the impact of comparative feedback or general reminders on practice-based referrals to a tobacco cessation quit line and to estimate the costs for the projected quit responses.

Study type/design Cluster-RCT, duration = 18 months, assessing impact of quarterly comparative feedback vs. general reminders to primary care clinicians within practices on their performance in generating referrals to a telephone quit-line service over an 18-month period between January 1, 2003, and June 30, 2004, smoking specific.

Participant group Participants: Primary care clinicians within practices in Michigan. Inclusion criteria: Not clearly stated, but the initial recruitment was from a health plan list (3,777) primary care practices in which only 858 practices seemed to fit the definition for usual primary care. Exclusion criteria: Not stated. Subject disposition: 87 primary care practices (43 intervention, 44 control) with 308 clinicians (163 Intervention, 145 control): 55% Family medicine (n = 171), 29% Internal medicine (n = 88), 16% Obstetrics/Gynaecology (n = 49), 1992 patients.

Intervention Quarterly comparative feedback report: A one-page report showing two graphs. First graph displays the quarter benchmark for referrals, individual clinician, his/her practice group, and performance of study group in numbers of referrals and percentage relative to the benchmark; the second graph shows the actual number of referrals made by the individual clinician per quarter.

Comparator General reminders: are simple postcards informing of the Quit Line services. They were sent to control practices and their clinicians on the same quarterly basis as the feedback reports.

Outcome definitions Changes in clinician referrals from baseline to post-intervention (18 months) for both groups, ‘C’; estimated quit responses of referred smokers, ‘Q’; estimated incremental costs of the feedback intervention. Clinicians in the intervention group were given quarterly comparative feedback on their individual and group referrals and how their referral activity compared with the mean referrals based on the overall top 10% of clinicians. Estimated quit responses and the associated incremental costs of the intervention were calculated based on the level of service obtained referred smokers. Primary care clinicians’ performance in generating referrals to a telephone quit line service. Achievable Benchmark Care (ABC) feedback defined as the mean number of referrals per quarter for the top 10% of the clinicians


Analyses: (number of referrals over time was the unit of analysis) X2 tests and t-test for baseline differences between practice, clinician and patient characteristics. The Mantel-Haenszel X2 test for repeated measures to assess the difference in the primary measure (number of referrals) between the two groups, over time. Resource costs were used to calculate the costs of services in both groups. The mean annual salary and benefit cost were used for nurse ($54,600), analyst ($76,602), clerk ($26,000), and statistician ($103,125) staff. Sample size calculation: The study required nearly 80 practices and 300 clinicians to detect a 25% difference. Because of a high intra-class correlation (ICC = .92) and a lack of adequate power to detect significant differences, individual clinician-and practice-level analyses were not completed.

280


Table 65: Wadland et al. (2007) (continued) A. (1) Reported: Practices were randomly assigned to intervention or control (so that clinicians within the same practice have the same allocation), method of randomisation was not stated. B. (0) Not reported: Not stated. C. (2) Adequate/Reported: No significant differences in prevalence of smoking or delivering of the 5As between intervention and control practices at baseline. D. (1) Inadequately reported E. (2) Reported: Yes. Baseline percentages of clinician asking, advising, assessing, assisting, and arranging follow-up (delivering of the 5As) for the two groups. F. (0) Unknown/Not reported G. (1) Inadequate



Absolute Quit Line referrals from baseline to 18 months showed a significant difference between intervention and control groups over time ‘C’. There were more referrals from the intervention practices (484) than from the control practices (220); p < .001 More estimated quits occurred in the intervention practices than in the control practices (66 vs. 36 quits) ‘Q’. Type of referral whether fax or telephone was also looked at. Results showed that referral through faxes exceeded those through telephone (84% vs. 16%), but telephone referrals resulted in greater likelihood of enrolment (77% telephone vs. 44% fax, p < .001) Cost of intervention per quit was higher than the cost of the control per quit ($174.04 vs. 108.93).

Authors conclusions

Providing simple, clear, and comparative feedback on referrals to a quit-line service had a modest impact on the referral performance of clinicians with limited increased costs; it may require a major integration of practice and system supports to become truly effective.

Reviewers notes Authors highlighted that they did not analyse individual clinical or practice differences due to the high intra-class correlation and lack of statistical power. They also noted that the exit surveys assess only two half-days per practice and may not accurately reflect longitudinal changes for individual clinicians, but they are valid for the entire group.


Yes, relevant as it compared the referrals of primary care clinicians to a tobacco quit-line service ‘C’ while using feedback report or general reminders.



Abbreviations: AHRQ = Agency for Healthcare Research and Quality; BMI = body mass index; CBA = controlled before-after trial; CCT = controlled clinical trials; CFPC = College of Family Physicians of Canada; CHD = coronary heart disease; CI = confidence interval; CME = continuous medical education; CMEP = continuing medical educational plan; FFS = fee for service; FP = family practitioner; GEEs = generalised estimating equations; GP = General Practitioner; GIMC = general internal medicine clinic; EHR = electronic health record; eLinkS = electronic linkage system; EMR = electronic medical record; HCP = health care professional; HMO = Health Maintenance Organisation; HSO = Health Services Organisation; ICC = inter-correlation coefficient; ICCs = intra-cluster correlation coefficients; NCI = National Cancer Institute; NRP = nicotine replacement therapy; OR = odds ratio; ORP = Outcomes Recognition Programme; PAR = practice assessment report; PCTs = Primary Care Trusts; PEI = patient exit interview; PGY = post graduate year; RCT = randomised controlled trial; RR = rate ratio; SD = standard deviation; SI = special intervention; SSCS = specialist smoking cessation services; TS = time-series design; UC = usual care; V.A. = Veterans Affairs clinic setting.

281


Table 66: Young et al. (2002) Citation Young, J. M., D'Este, C., & Ward, J. E. (2002). Improving family physicians' use of

evidence-based smoking cessation strategies: A cluster randomization trial. Preventive Medicine, 35(6), 572-583.


To assess a multifaceted, practice-based intervention involving audit, feedback, and academic detailing to improve family physician smoking cessation advice.

Study type/design Cluster randomised controlled trial, duration = 6 months, analysed/eligible = 35%, smoking specific.

Participant group Participants: 39 practices (20 intervention, 19 control), 60 family physicians Inclusion criteria: Exclusion criteria: Worked fewer than two days per week, planning to leave the practice within six months, were on extended or maternity leave, did not employ a receptionist, or were already participating in a clinical audit or if more than 50% of their patients spoke a language other than English at home. Subject disposition: Of 184 eligible FPs, 65 (35%) consented to participate. five FPs were unable to complete baseline data collection due to personal illness (n = 2), unexpected absence (n = 2), or inability of practice staff to comply with patient recruitment requirements (n = 1). Of 2,332 eligible patients attending FPs at baseline, 1,788 (77%) consented to participate. Overall, 1,241 completed follow-up questionnaires were received (70% response rate among consenters).

Intervention Multifaceted Intervention in Smoking Cessation Advice phased across three academic detail visits. The intervention is a combination of audit and feedback, resources for FPs, resources for practices, and resources for patients.

Comparator Control group involved an intervention of identical format and intensity but about cervical screening.

Outcome definitions Authors assessed change in FP behaviour by using two objective methods, namely patient recall of smoking cessation advice during their most recent consultation (referred to as the “index” consultation) and medical record audit. They also measured FP self-report. Outcomes were: Smokers’ recall of FP advice to stop smoking. The components of smoking cessation advice included all elements of ‘ABC’ these are: Ask smoking status, ‘A’, Advise smoker to quit ‘B’, provide practical advice, set a “quit date”, provide written materials, recommend nicotine gum, recommend nicotine patches, arrange a follow-up appointment, refer to a smoking clinic (all are ‘C’).


Analyses: Using generalized estimating equations (GEEs), Logistic regression analysis using GEEs also was conducted to assess the impact of the intervention to improve documentation of smoking status in patients’ medical records. At baseline: Crude rates of recall by smokers of advice to stop smoking during the index consultation were significantly higher among those attending FPs in the intervention group than in the control group. Sample size calculation: Authors stated that since no intra-cluster correlation coefficients (ICCs) for provision of smoking cessation advice in family practice were reported in the existing literature, samples size calculations were performed for a range of plausible ICCs assuming an average cluster size of 16 and for a range of baseline rates for provision of smoking cessation advice from 10 to 30%. Calculation showed that it is needed to recruit 18 practices in each group (36 practices in total).

282


Table 66: Young et al. (2002) (continued) A. (2) Adequate: Used random numbers generated by SAS® statistical software. B. (1) Inadequate: No information provided for allocation of treatment, although post-test medical record audit was conducted blind to group allocation. But for data entry, it was stated that data were entered into a database created by research assistants who were blind to group allocation of practices. C. (2) Adequate/Reported D. (2) Adequate/Reported E. (2) Unknown/Not reported: For each primary outcome, univariate associations with patient and FP variables were assessed. All variables found to be significant or nearly significant (P < 0.25) in univariate analysis were included into a generalized linear model with practice as the classification (“cluster”) variable. F. (0) Unknown/Not reported G. (2) Adequate/Reported



Smokers’ recall of FP advice to stop smoking (‘ABC’): 54% of smokers attending FPs in the intervention group recalled being given advice to stop smoking (adjusted 95% CI: 39–68%) ‘B’. Between baseline and post-test, there was a significant increase in recall of advice among smokers attending FPs in the intervention group and a non significant increase among those attending FPs in the control group (Table 4 of the original paper). Rates of documentation of patients’ smoking status ‘A’ improved in both groups from baseline to post-test, but the improvement in the intervention group was not significantly greater than in the control group. From baseline to post-test, documentation of any form of smoking cessation advice within the index consultation with a smoker improved significantly in the intervention group. However, there also was a non-significant improvement in the control group. Again, improvement in the intervention group was not significantly greater than in the control group For all outcomes except discussing passive smoking and referral to a smoking cessation clinic, there was significant improvement from baseline to post-test in the intervention group. In summary improvements between baseline and post-test in patient recall of FP advice about nicotine replacement patches and gum were significantly greater in the intervention than in the control group (p = .0056 and p - .0002, respectively). There were substantial increases in patient recall of assessment of smoking status and FP use of “quit dates”, behavioural advice, and written materials in the intervention group, these changes however, were not significantly greater than those in the control group. Notation of patients’ smoking status and smoking cessation advice in medical records remained suboptimal in both groups.

Authors conclusions This multifaceted intervention was successful only in promoting FPs’ use of nicotine replacement therapy. While the use of other effective cessation strategies appeared to increase, a larger trial is needed for further evaluation.

283


Table 66: Young et al. (2002) (continued) Reviewers notes FPs completed self-administered questionnaires before and after participating in the

intervention. At post-test, FPs were also asked to indicate their views about the usefulness of specific components.


Yes, relevant, as it assessed an intervention to improve delivery of FP smoking cessation advice.




Documents

A systematic review of the literature - Health Services Assessment