www.elsevier.com/locate/jocn

Journal of Clinical Neuroscience 14 (2007) 189–200

Review

Care pathways for acute stroke care and stroke rehabilitation:From theory to evidence

Joseph Kwan *

Elderly Care Research Unit, University of Southampton, UK

Received 11 January 2006; accepted 25 January 2006

Abstract

Care pathways aim to promote evidence- and guideline-based care, improve the organisation and efficiency of care, and reduce cost.In the past decade, care pathways have been increasingly implemented as a tool in acute stroke care and stroke rehabilitation. In the mostrecent Cochrane systematic review, which included three randomised and 12 non-randomised studies, patient management with strokecare pathways was found to have no significant benefit on functional outcome, and patient satisfaction and quality of life might actuallybe worse. On the other hand, it was associated with a higher proportion of patients receiving investigations and a lower risk of developingcertain complications such as infections and readmissions. Overall, the evidence supports the use of care pathways in acute stroke but notstroke rehabilitation. Future developments, including electronic care pathways, patient pathways, and pre-hospital care pathways forhyperacute stroke, will be discussed.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Systematic review; Critical pathway; Hospitalization; Cerebrovascular accident; Randomised trials

1. Care pathways: The theory

Care pathways are increasingly being implementedacross many countries to improve the care of stroke pa-tients, but there is relatively little debate about what theyare and how they affect patient care and outcome.1,2 Onereason could be that care pathways are generally regardedas harmless, and policy makers are keen to adopt newhealth service interventions that are intuitively beneficial,even though they have not yet been thoroughly tested. Thispaper describes the definition and origin of care pathways,the theoretical basis of using them, and the best availableevidence of their benefits and risks.

0967-5868/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jocn.2006.01.026

* Present address: Elderly Care Research Unit, Level E (Mailpoint 807),Southampton General Hospital, Tremona Road, Southampton SO166YD, UK. Tel./fax: +44 2380 796128.

E-mail address: drjkwan@aol.com.

1.1. Definition of a care pathway

Care pathways are organisational interventions that aimto promote evidence- and guideline-based care, improvethe organisation and efficiency of care, and reduce cost.The medical literature is abundant with articles that praisethis tool, describing the many potential benefits associatedwith their use.1 However, the multitude of benefits ascribedto care pathways may be partly due to a lack of conceptualclarity surrounding the definition of the intervention.3

There is currently no universally accepted definition for acare pathway.4,5 Other terms used to denote a care path-way include clinical pathways, critical pathways, care pathmethod, anticipatory recovery pathways, and CareM-aps�.1,6 Although these terms are used to describe a com-mon concept, they may in fact be describing differentinterventions, used within different settings for differentconditions, and implemented according to different objec-tives. From examining the literature, the three essentialelements of a care pathway include: (i) it is a plan of care;(ii) it is developed and used by the multidisciplinary team

190 J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200

(doctors, nurses, therapists); and (iii) it is applicable to sev-eral aspects of care (assessment, diagnosis, investigation,treatment). It is indisputable that other elements, such asvariance reporting, are vital for the function of certain carepathways, but these elements are absent in many publishedcare pathways and are therefore not regarded by many asessential in defining a basic care pathway – although someexperts may disagree.7

1.2. The origin of care pathways

The concept of care pathways was acquired from theindustrial world. In 1958, the US Navy planned to buildthe Polaris submarine. Due to the enormous complexityof the project and the huge number (over 3000) of contrac-tors involved, a new method called ‘‘Program Evaluationand Review Technique’’ (PERT) was developed to assistwith the planning and scheduling of the project.8 PERThelped to define the essential tasks and the length of timeneeded to accomplish them.9 Around the same time, Du-Pont company and Remington Rand Corporation devel-oped a similar tool, called the ‘‘Critical Path Method’’, toassist with the scheduling of the shutdown of DuPontchemical plants.10 The term ‘critical’ referred to the stepsthat took the longest time, so that any delay in completingthese steps would delay the entire project. It was quicklyrealised that identifying the optimal pathway in the pro-duction of individual items was beneficial both in termsof cost and productivity.

Similar tools were applied to healthcare in the USA in the1980s when case management was first introduced in re-sponse to escalating healthcare costs and increasing con-sumer demands. Care pathways were used as part of casemanagement to promote high-quality patient care that isdelivered in a timely and cost-effective manner.11 Early carepathways were designed for medical conditions or surgicalprocedures that were regarded as common, ‘simple’ (that isusually single pathology with little variation in practice),and costly.12,13 Later, care pathways were applied to managemore complex conditions such as stroke, diabetes, psychoticillness, and palliative care. In some countries (including theUK), the primary objective of using care pathways is perhapsless to do with cost-containment, but more to do with thepromotion of evidence- and guideline-based practice.14–16

1.3. How care pathways operate

The four major aims of using care pathways are summa-rised in Table 1. Care pathways are designed to be used as a

Table 1Major aims of using care pathways

� To assist healthcare professionals in making clinical decisions according to� To improve the quality of patient care by reducing variation in clinical pr� To reduce length of stay and hospitalisation costs� To improve communication between disciplines, and between patient and� To improve quality of documentation and facilitate data collection for aud

structured clinical record by every member of the multidis-ciplinary team.17 The format of care pathways varies frombeing a small-scaled paper-based document to elaboratecomputer programs that guide the healthcare professionalthrough every step of the patient’s management.18

On the whole, care pathways are different from otherforms of information provision systems that aim to assisthealthcare professionals with clinical decision-making,such as guidelines and protocols. Furthermore, since carepathways are usually applicable to several aspects of care,they are also different from other mono-faceted tools suchas diagnostic algorithm (for example, stroke diagnosticprotocol for paramedical staff), or screening tools for spe-cific interventions (for example, thrombolysis for acuteischaemic stroke).19

In the acute setting or rehabilitation, care pathways areusually initiated at the time of admission and terminatedwhen the patient is discharged or transferred to anothersetting. The interventions are mapped out in terms of daysof inpatient care. For emergency or intensive care path-ways, the interventions can be mapped out in terms ofhours or even minutes of care, whereas care pathways inthe primary care or community setting often define theweeks or even months of care. The essential elements ofhow to design and implement care pathways are summa-rised in Table 2.

1.4. Variance reporting and analysis

Many (but not all) care pathways have an in-built sys-tem to assess clinical outcomes and monitor practice var-iation, called variance reporting or tracking.20,21 Variancesrepresent the discrepancies between planned and actualevents. They can also represent outcomes that differ fromthose anticipated or deviations from the projected time-line.22 Some experts argue that the most important differ-ence between care pathways and other forms ofinformation provision systems is that care pathways havea mechanism for variance reporting and analysis.7 There isno doubt that variance reporting is feasible and useful forless complex conditions such as fractured neck of femur.23

However, for more complex conditions such as acutestroke, where the progress of recovery and prognosis arehighly unpredictable, many centres have abandoned vari-ance reporting simply because variances occur too fre-quently and the process is too time-consuming. If thecare pathway also includes strategies to deal with commonvariances, the care pathway might become too lengthy andimpractical to use, leading to poor compliance or even

the best-available evidence, local policies, and national guidelinesactice and improving efficiency

healthcare professionalsit and research projects

Table 2Implementing care pathways: how to improve the chance of success

1. Laying the foundation� Collective willingness to embrace change and clear commitment from clinical and managerial leaders� Clear explanation to the staff of how the care pathway may influence patient care� Be realistic about what the care pathway can and cannot do� Frank discussions to explore the staff’s views and concerns� Define the current usual practice and avoid being distracted by exceptions� Identify the practical barriers to improve care, and methods of overcoming these barriers� Avoid ‘re-inventing the wheel’ – learn from what others have already done

2. Designing the care pathway� Involve all the disciplines in the design team, avoid leaving anyone out� Literature searches for the most up-to-date evidence and clinical guidelines� Define the variances of interest - keep them simple, few, and easy to document� The care pathway document should be simple, clear, concise, and user-friendly� Clarify information and avoid assuming staff’s knowledge about the condition� Avoid using non-standard abbreviations� Pilot the care pathway and refine it according to feedback and comments

3. Implementing the care pathway� Pre-launch advertising to generate interest and enthusiasm� Training sessions for staff to learn how to use the care pathway� Post-launch support and encouragement, and monitoring of staff’s compliance� Continual feedback from the staff and refining of the care pathway� Variance tracking and analysis, and regular feedback of variance reports� Commitment for continual resources and support from the clinical and managerial leaders� Continual updating of information in the care pathway according to the latest evidence and guidelines

J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200 191

abandonment. It is therefore wise to keep the variancedata set as minimal as possible, limiting them only to asmall number of critical indicators of process and out-come.24 In the future, electronic care pathways may re-solve this problem by automatic variance tracking andanalysis.

1.5. Potential benefits of using care pathways

There are many potential benefits of using care path-ways.15,25 Whether these potential benefits become realitydepends not only on the design and content of the carepathway, but also on the way that it is applied and imple-mented within the local setting.1 Like clinical guidelines,the potential benefits of using care pathways can be dividedinto three main categories: (i) benefits to the patient;(ii) benefits to the healthcare professional; and (iii) benefitsto the healthcare system.26 These are summarised inTable 3.

1.6. Potential concerns of using care pathways

The most commonly voiced concern about using carepathways is that they may reduce the level of autonomythat professionals can exert over patient care.25,27 Somehave used the term ‘cookbook medicine’ to describe theuse of care pathways,28–30 while others disagree.31 Whilecare pathways aim to standardize patient care, they areonly meant to be templates to assist clinical decision-mak-ing. The final decision of whether or not to use specifictreatment strategies should rest with the professionals

themselves. Care pathways may also raise the expectationof the patients and relatives (especially if they have accessto the document); this may be problematic if there is anydeviation from the recommended care plan.

Although care pathways aim to streamline patientcare, reduce paperwork, and lower costs, the reversecan potentially occur. Care pathways may in fact turn aroutine and simple task into a complicated and labour-intensive plan, with more paperwork to complete.6,32 Thismay in turn leave little time for other (and possibly moreimportant) aspects of patient care. Alternatively, carepathways may over-simplify patient care, and healthcareprofessionals may stop trying to solve certain clinicalproblems by themselves, especially if those items are notincluded in the pathway. With time, healthcare staffmay become de-skilled and less satisfied,33 and hospitali-sation costs could rise.34 Lastly, the process of design,implementation, and local evaluation of care pathwaysrequire a great deal of time, commitment and resource,much of which is scarce in this current climate ofhealthcare.

2. Different forms of stroke care pathways

Since care pathways were first introduced for stroke carein the mid-1980s,35 many different forms of ‘pathways’have been developed, such as those used to aid the diagno-sis of acute stroke and screening for thrombolytic ther-apy.36,37 Others have been developed to guide patientcare in the community after hospital discharge.38,39 In gen-eral, stroke care pathways can be broadly divided into

Table 3Potential benefits of using care pathways

Benefits to the patient� Promote evidence- and guideline-based treatments� Improve the thoroughness, efficiency and organisation of patient care� Reduce unnecessary investigations and errors� Improve communication with the patient and family� Enhance patient satisfaction and reduce the number of complaints and litigation claims

Benefits to the healthcare professional� Standardise patient care� Improve quality of documentation� Improve communication between disciplines� Facilitate the introduction of new or agency staff� Facilitate ‘hand-over’ between healthcare professionals (e.g. between shifts)� Enhance accountability for each task or intervention� Be used as a training tool for junior doctors, nurses, and therapists

Benefits to the healthcare system� Promote adherence to the latest evidence and guidelines� Part of the continuous quality improvement strategy� Facilitate audit and research� Process of designing the care pathway can encourage collaboration and team work� Variance reporting can improve quality of care� Reduce hospital cost

192 J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200

three categories: (i) those used for acute stroke manage-ment only; (ii) those used for stroke rehabilitation only;and (iii) those used for both acute stroke managementand stroke rehabilitation.

2.1. Care pathways for acute stroke management

Management of acute stroke is complex. It consists of alarge number of components such as consideration forthrombolysis, correction of deranged physiological param-eters, prompt establishment of the pathology and aetiologyusing appropriate investigations, and early implementationof secondary preventive measures. Effective acute stroketreatment therefore requires a highly organised and trainedteam of motivated healthcare professionals, working effi-ciently within a well-equipped and dedicated stroke servicestructure.37,40 Care pathways, with their many potentialadvantages, could help to improve acute stroke care andstandardise practice,41 and many such care pathways havebeen reported.35,42–56

In the hyperacute phase of stroke (in the first few hours),a care pathway could assist with the process of decidingwhether the patient is eligible for intravenous thrombolysis,making sure that the patient satisfies the criteria for itsadministration.40 Within the acute stroke unit, the carepathway can facilitate the process of general medical andnursing care during the first few days, as well as guidingearly multidisciplinary assessments and therapy. The treat-ment strategy for each stroke patient should be tailored tothe individual because of the unpredictable course of illnessand the frequent occurrence of complications such as infec-tions, dehydration, stroke progression, deep vein thrombo-sis, and seizures. In the first few days, the care pathwaycould ensure that all the important physiological parame-

ters are closely monitored and corrected if found to be out-side the agreed range (for example raised blood glucose andtemperature, and low oxygen saturation).57 Nurses can fol-low the pre-determined protocol in the care pathway forthe essential nursing procedures such as continence care,pressure area care, nutritional assessment, nasogastric feed-ing, and positioning of the patient. The choice of investiga-tions into the aetiology of the cerebral event, as well assecondary preventive measures, can also follow evidence-and guideline-based algorithms within the care pathway.This could make sure the right groups of patients receivethe right investigations and treatments at the right time,and avoid the inappropriate use of certain investigationsand treatments (according to research evidence and lo-cally-agreed guidelines). Not only may this lead to im-proved patient outcome, it may also reduce length of stayand other elements of hospitalisation cost.

For common co-morbidities or complications afteracute stroke such as incontinence, dysphagia or pressuresores, some centres have also designed ‘co-pathways’ to as-sist with their management in a more structured and organ-ised manner.58 These might be useful in reducing theamount of information in the core care pathway, makingit simpler, more streamlined and user-friendly. Since manyof these complications or co-morbidities are not unique tostroke, the co-pathways could be shared across depart-ments within the same hospital.

2.2. Care pathways for stroke rehabilitation

During stroke rehabilitation, the patient is usually med-ically more stable. However, neurological problems (suchas communication problems, dysphagia, incontinence),and general complications (including chest infections,

J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200 193

depression, painful shoulder) can add significantly to thepatient’s impairment, disability and handicap. Meticulousattention to the prevention and early treatment of compli-cations is important, which may in turn lead to improvedoutcome. The more efficiently such interventions are ap-plied, the more patients are likely to benefit.

Care pathways for stroke rehabilitation are potentiallyuseful in enhancing the process of stroke rehabilitation;for example, by improving organisation, enhancing multi-disciplinary communication, and facilitating the applica-tion of evidence and guideline recommendations.41 Thereare several reports of the use of care pathways for strokerehabilitation in the literature.17,38,59–62 One practical prob-lem with developing care pathways for stroke rehabilita-tion is that many aspects of stroke rehabilitation are notsupported by robust research evidence. For example, it re-mains unclear which physiotherapy (or speech therapy)treatment strategy should be employed for which subgroupof patients, for how long, and with what level of intensity.If there is disagreement about the content of the pathwaybetween the different disciplines involved, it might lead toa lack of enthusiasm, non-compliance or even abandon-ment of the pathway.

Care pathways can also improve the efficiency of dis-charge planning, which can be a complex process withmany parties involved. Care pathways can assist with co-ordinating the communication between doctors, nurses,therapists, social services and the primary care team, ensur-ing a smooth discharge with clear instructions for follow-up activities such as continuing physiotherapy at home,or repeat swallowing assessment.

2.3. Care pathways for acute stroke management and stroke

rehabilitation

Some care pathways are designed for both acute strokemanagement and stroke rehabilitation.34,39,63–68 This typeof care pathway may be particularly useful in hospitals thathave a combined acute and rehabilitation stroke unit. Formany of these ‘combined’ care pathways, the main compo-nent is the acute stroke pathway (starting in the emergencydepartment), and after the first few days, the documenta-tion simply carries on with daily ‘continuation’ sheets forthe rehabilitation phase of hospitalisation. The design ofsuch continuation sheets differ between units, but manyadopt the ‘unitary patient record’ format with blank spacesfor the different disciplines to write and communicate.69 It

Table 4Reasons why randomised trials may be difficult to perform when evaluating c

� Care pathway care is a complex intervention, leading to difficulty in interp� The control intervention (e.g. ‘standard’ medical care) may also be comple� Large numbers of patients are needed to assess a modest size of effect and m

suming, and very expensive� Healthcare professionals may not want to take part because of pre-conceiv� Patients and carers may not give consent to be randomised especially when� Blinding of the patient may be difficult and blinding of the healthcare pro� Cross-contamination of treatment effects if individual patients are random

is then arguable whether such a documentation tool actu-ally satisfies the criteria for being a care pathway.

3. Care pathways for stroke: The evidence

Compared with trials of pharmacological agents or sur-gical procedures, the evaluation of health service interven-tions such as care pathways is more complex. In practice,care pathways are highly variable in their nature and theiroperation depends greatly on the local circumstances. Con-sequently, the results generated by studies of care pathwaysmay have questionable internal validity (that is, the extentto which differences between the study groups are realrather than a product of bias) and external validity (thatis, the extent to which trial results can be generalised to awider population).70,71

It is widely accepted that well-designed, large scale ran-domised controlled trials (RCTs) represent the most effec-tive method to evaluate a therapeutic interventionbecause they minimise the effects of bias and provide themost accurate estimate of effect for an intervention.72,73

However, robust RCTs are often difficult to perform whenevaluating care pathways, and the main reasons for this aresummarised in Table 4.

Randomisation of individual patients to care pathwayor standard medical care would only be meaningful if thetwo types of care are comparable in every way except forthe use of the care pathway. If individual patients are ran-domised within a single unit, the healthcare staff wouldhave to ‘switch’ between using a care pathway and notusing one. This is difficult to achieve in practice without‘cross-contamination’ of treatment effects. Moreover, if pa-tients are randomised to be managed in different wards(with one ward using the care pathway and another notusing the care pathway), then patient care is likely to be sig-nificantly different, which could also confound the trial re-sults. The consequence of cross-contamination may be therejection of an effective intervention as ineffective becausethe observed effect size is statistically non-significant (atype II error).74

When RCTs are not feasible, one might have to gatherevidence using multiple approaches including both quanti-tative and qualitative studies.75 Evidence from each ap-proach can be considered as one piece of the jigsawpuzzle, and when they are interpreted together, they mayprovide a more complete picture about an intervention;this multidimensional approach is sometimes known as

are pathways

reting trial resultsx in natureaximise the internal validity of study; this could be difficult, time-con-

ed (positive or negative) opinion about the interventioncare pathways are intuitively beneficial and are not believed as harmfulfessional is almost impossible - this could lead to biasised within a single unit

194 J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200

‘triangulation’.76 Triangulation has been used in nursingand sociological research where many of the interventionsand issues are complex.77–80 Triangulation should be re-garded more as a way of maximising comprehensivenessand encouraging a more flexible analysis of the data, ratherthan as a method of obtaining highly accurate data.76

4. Evidence from the Cochrane review

Using Cochrane methodology, a recent systematic re-view evaluated the effects of care pathways, as comparedto standard medical care, among patients with acute strokewho had been admitted to hospital.81 The major databaseswere searched including the Cochrane Controlled TrialsRegister, MEDLINE, EMBASE, and CINAHL. The re-view considered randomised controlled trials and non-ran-domised studies, which included quasi-randomised trials,comparative studies, before-and-after studies, and inter-rupted time series.

The primary outcome of interest was the proportion ofpatients who were dead or dependent (or requiring long-term institutional care) at the end of follow-up. Other out-comes of interest included: complication rates, process ofcare, readmission rates, patient and carer satisfaction,quality of life, duration of hospital stay, and hospitalisa-tion costs.

4.1. Characteristics of the included clinical studies

The review found three randomised controlled trialswith a total of 340 patients, and 12 non-randomised studieswith a total of 4081 patients, that compared care pathwaycare with standard medical care. Of the three randomisedcontrolled trials, two assessed care pathways that were de-signed for stroke rehabilitation,60,61 and in one study,67 thecare pathway was designed for acute stroke managementand rehabilitation.

Of the 12 non-randomised studies, one was a retrospec-tive comparative study53 and 11 were before-and-afterstudies.34,54–56,63,64,82–86 The care pathways were imple-mented for acute stroke in seven studies,53–56,82,84,86 strokerehabilitation in one study,83 and for acute stroke andrehabilitation in four studies.34,63,64,85 Three of the acutestroke care pathways began with treatment at the emer-gency department (including screening for thrombolytictherapy).53–55

4.2. Effects of introducing care pathways versus standard

medical care

A large number of outcomes were assessed by the stud-ies, which in part demonstrates the substantial variation inthe aims and reporting of these different studies. Results ofthe major outcomes have been summarised here, but theyshould be interpreted with caution because of the heteroge-neity of the methodology and findings between studies, thenon-randomised nature of the majority of the studies, and

the relatively small numbers of patients available for dataanalysis of each outcome. For the following outcomes,OR represents the odds ratio and CI represents the 95%confidence interval.

4.3. Death by the end of follow-up

Three studies (one randomised and two non-random-ised, total of 783 patients) reported this outcome. The ran-domised study showed a trend toward more deaths by theend of follow-up in the care pathway group (OR 1.77,CI = 0.61–5.140).61 The two non-randomised studiesshowed a trend toward fewer deaths by the end offollow-up in the care pathway group (OR 0.69,CI = 0.44–1.07).64,82 The aggregate result showed no sig-nificant difference (OR 0.88, CI = 0.49–1.57, p = 0.7).

4.4. Dependency at discharge

Two studies (one randomised and one non-randomised,total of 667 patients) reported this outcome. Dependencywas assessed using the Functional Independence Measure,with higher scores indicating higher level of independence.The randomised study showed a trend toward a lower levelof independence at discharge in the care pathway group(weighted mean difference [WMD] �4.9, CI = �14.6 to+4.8).60 The non-randomised study also showed a trend to-ward a lower level of independence at discharge in the carepathway group (WMD �3.6, CI = �7.4 to +0.2).83 Theaggregate result showed that patients in the care pathwaygroup had a significantly lower level of independence(WMD �3.8, CI = �7.3 to �0.2, p = 0.04).

4.5. Institutionalisation

Seven studies (one randomised and six non-randomised,total of 1613 patients) reported this outcome. The random-ised study showed a trend toward fewer patients dischargedto institutional care in the care pathway group (OR 0.57,CI = 0.24–1.35).61 The six non-randomised studies alsoshowed a trend toward fewer patients discharged to institu-tional care in the care pathway group (OR 0.82, CI =0.55–1.23).53,56,64,82–84 The aggregate result showed anon-significant trend toward fewer patients discharged toinstitutional care in the care pathway group (OR 0.79,CI = 0.55–1.13, p = 0.2).

4.6. Discharge home

Seven studies (one randomised and six non-randomised,total of 1613 patients) reported this outcome. The random-ised study showed no significant difference (OR 1.14,CI = 0.56–2.32).61 The six non-randomised studies alsoshowed no significant difference (OR 1.2, CI = 0.84–1.7).53,56,64,82–84 The aggregate result showed no significantdifference (OR 1.18, CI = 0.88–1.59, p = 0.3).

J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200 195

4.7. Pneumonia

Four non-randomised studies56,63,82,85 with 797 patientsshowed no significant difference (OR 0.89, CI = 0.53–1.5,p = 0.7).

4.8. Urinary tract infection

Six non-randomised studies55,56,63,82,84,85 with 1283 pa-tients showed that significantly fewer patients suffered uri-nary tract infections in the care pathway group (OR 0.51,CI = 0.34–0.79, p = 0.02).

4.9. Use of CT brain scan

Four non-randomised studies34,54,64,82 with 1315 pa-tients showed that significantly more patients received aCT brain scan in the care pathway group (OR 2.42,CI = 1.12–5.25, p = 0.02).

4.10. Use of carotid duplex study and echocardiography

Three non-randomised studies34,55,82 with 766 patientsshowed a trend toward more patients receiving a carotidduplex study in the care pathway group (OR 1.79, CI =0.76–4.2, p = 0.18). For echocardiography, two non-ran-domised studies 34,82 with 491 patients showed a non-significant trend toward more patients receiving anechocardiogram in care pathway group (OR 2.08, CI =0.94–4.58, p = 0.07).

4.11. Urinary catheterisation for patients with incontinence

One non-randomised study 82 with 351 patients showedno significant difference (OR 0.78, CI = 0.41–1.48,p = 0.4).

4.12. Therapy input

One randomised study showed no significant differencein the cumulative duration of physiotherapy or occupa-tional therapy at various follow-up time points.87 Onenon-randomised study found ‘‘no difference’’ in therapy in-put but no data were provided.55

4.13. Readmission or emergency department attendance

Two studies (one randomised and one non-randomised,total of 110 patients) reported this outcome. The random-ised study showed significantly fewer readmissions or emer-gency department attendances in the care pathway group(OR 0.15, CI = 0.04–0.59).67 The non-randomised studyalso showed significantly fewer readmissions or emergencydepartment attendances in the care pathway group (OR0.03, CI < 0.1–0.63).56 The aggregate result showedsignificantly fewer readmissions or emergency departmentattendances in the care pathway group (OR 0.11,CI = 0.03–0.39, p = 0.0006).

4.14. Patient and carer satisfaction

One randomised study60 with 121 patients showed thatpatients were significantly less satisfied with their hospitalcare in the care pathway group (WMD �1.1, CI = �1.91to �0.29, p = 0.008).

4.15. Quality of life

One randomised study61 with 152 patients reported thisoutcome, using Euroqol score as a measure of quality oflife. This study found no significant difference in the Euro-qol score at one or 3 months. However, at 6 months, themedian Euroqol score was found to be significantly lowerin the care pathway group, which suggests a lower qualityof life in the care pathway group. The study also found thatcontrols performed better in the Euroqol domain for socialfunctioning, but patients in the care pathway group per-formed better in the Euroqol domain for self-care.88

4.16. Duration of hospital stay

Six studies (two randomised and four non-randomised,total of 1915 patients) reported this outcome. The two ran-domised studies showed a trend toward longer mean lengthof hospital stay in care pathway group (WMD 3.99,CI = �0.29 to +8.27 days).60,61 The four non-randomisedstudies showed that mean length of hospital stay was sig-nificantly shorter in care pathway group (WMD �1.89,CI = �2.95 to �0.82 days).54,82–84 The aggregate resultshowed a non-significant trend toward shorter mean lengthof hospital stay in care pathway group (WMD �1.39,CI = �2.8 to +0.02 days, p = 0.14).

4.17. Hospitalisation cost

Five studies (two randomised and three non-random-ised) reported this outcome. One randomised study foundno significant difference in hospitalisation cost betweencare pathway and control groups60 and another random-ised study found a lower mean hospitalisation cost in thecare pathway group.67 Two non-randomised studies founda ‘‘fall’’ in the mean hospitalisation cost55,63 and one non-randomised study found a 14.6% fall in the mean hospital-isation cost, but no actual cost data given.84

4.18. Quality of documentation

Two studies (one randomised and one non-randomised)reported this outcome. Both found that quality of docu-mentation was significantly better (that is, more likely tobe recorded) in the care pathway group. The randomisedstudy found that patients in the care pathway group hadsignificantly more comprehensive documentation of:(i) certain aspects of neurological and nutritional assess-ments; and (ii) notification to the general practitionerregarding the patient’s discharge from hospital.87 The

196 J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200

non-randomised study found that introduction of the carepathway significantly improved the documentation of: (i)different aspects of neurological assessment; and (ii) ana-tomical site of the brain lesion and its pathological type.82

5. Interpretation of results

Non-randomised studies are particularly susceptible tobiases, and all clinical studies of care pathways are suscep-tible to confounding. One obvious bias is selection bias,that is stroke patients may have been selected to be man-aged using a care pathway and may have differed fromthose who were managed using standard medical care. Inone study, it was stated that patients were ‘‘selected for carepathway care using strict screening criteria’’53 and we sus-pect that this was also the common practice in other stud-ies. Consequently, the clinicians may have selected thestroke patients with better (or worse) prognosis and biasedtheir findings.

There are other potentially important biases in non-ran-domised studies. Most of the studies were retrospective andonly one study82 included consecutive cases. It is possiblethat some cases were missed or excluded, which may haveinfluenced outcome. The investigators who assessed theoutcomes were not reported to be blinded to the treatmentoption and this may have biased their assessment of non-fatal outcomes. Moreover, publication bias may have influ-enced the results of the non-randomised studies, such thatthose showing no benefit or worse outcome with care path-way care may not have been published.27 Finally, someauthors chose to write ‘‘no difference’’ rather than reportthe actual data, or may have omitted the negative resultsall together from their publication, making it more difficultto interpret the results.

The most important confounding factors include the dif-ference in patient care between comparison groups (on topof the introduction of the care pathway), variations in thedefinition and components of the intervention, and thesmall number of studies included in the data analysis.

6. Effect of care pathways on major outcomes

Many experts think that care pathways must be benefi-cial and could not possibly do any harm, but this view isnot wholly supported by this review. There is no conclusiveevidence that care pathways provide additional benefit overstandard care for the major clinical outcomes (death or dis-charge destination). In fact, there is some evidence fromone randomised60 and one non-randomised study83 thatpatients in the care pathway group might be significantlyless independent at discharge. Furthermore, evidence fromtwo randomised trials suggests that patient satisfaction andquality of life might be lower in the care pathwaygroup.60,61 The reasons for these observed effects are un-clear, but if the main objective of the care pathways in thesestudies was to shorten the duration of hospital stay, thenthere might have been pressure for the healthcare staff to

discharge the patients as quickly as possible, sometimes be-fore the patients or carers were ready for discharge. Thereis also weak evidence, chiefly from non-randomised studies,that care pathway care might be associated with better pro-cess of care, hence leading to fewer complications (includ-ing urinary tract infections, readmissions or emergencydepartment attendances) and more widespread use of cer-tain investigations (such as CT brain scans).

6.1. Effect of care pathways on hospitalisation cost

Although it is intuitive to believe that care pathwaysshould reduce hospitalisation cost, no firm conclusioncan be drawn from the included studies. The chief determi-nant of cost is length of hospital stay, but the analysis ofthis outcome is difficult to interpret because two random-ised studies suggested pathway care increased the lengthof stay, whereas four non-randomised studies showed areduction. Four studies55,63,67,84 reported a reduction inmean hospital cost, and two studies found no differencein cost.34,60

Only one study reported the items of costs (that is, whatitems were included in the final sum) and their individualvalues.60 Without knowing the cost of individual items,comparison between studies is virtually meaningless. Fur-thermore, using care pathways can be associated withmany indirect and opportunity costs such as the time andeffort invested in designing the pathway, promoting itsuse, educating the staff, printing the pathway documents,as well as maintaining staff enthusiasm and variance track-ing. All these costs are very difficult to estimate or quantify,and they could substantially increase the overall cost ofusing care pathways.

7. Care pathways and the basic structure of the stroke service

Care pathways are only likely to be effective in improv-ing stroke care if the basic structure of the stroke service isalready in place, such as the routine admission of patientsto a stroke unit where patient care is provided by a coordi-nated multidisciplinary team made up of dedicated, moti-vated and skilled healthcare professionals.89,90 If the basicstructure is not in place (for example, if patients managedby different medical teams in different general wards), thencare pathways are less likely to significantly influence theprocess of care since there would be too many practicalbarriers to the implementation and operation of thepathway.91

It is therefore common to find that care pathways areintroduced as part of an overall quality improvementscheme, rather than as a stand-alone organisational inter-vention.92 For example, acute stroke care pathways maybe introduced as part of a multi-faceted program consistingof a number of objectives such as to improve the accuracyof pre-hospital diagnosis of stroke, to speed up the transferof patients to hospital for emergency treatments such asthrombolysis, to improve the efficiency of triage and assess-

J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200 197

ment of patients with suspected stroke, including emer-gency neuroimaging for those who might be eligible forthrombolysis.

8. Care pathways for acute stroke management or stroke

rehabilitation?

Although only a small number of studies were includedin this systematic review, the results were consistently bet-ter for acute stroke pathways than stroke rehabilitationpathways. This can be explained by that fact that carepathways probably have the best potential to improve theprocess of care during the acute (or hyperacute) phase ofstroke where the process of assessment, investigation, diag-nosis can be highly complex (and often disorganised), andwhere speed is the essence when emergency treatments suchas thrombolysis are considered. However, in stroke rehabil-itation patients are often medically stable and managed bya coordinated multidisciplinary team within a well-struc-tured service, hence the care pathway is less likely to pro-vide additional benefit. The evidence so far supports theuse of care pathways for the management of acute strokebut not stroke rehabilitation.

9. Care pathways: The future

9.1. Electronic care pathways

Many hospital documents which have traditionally beenpaper-based (e.g. patient records) are increasingly becom-ing electronic. However, the introduction of any electronicpatient record would be too expensive and time-consumingif it is only used as a tool to facilitate documentation andretrieve data faster; it must also add value in supportingclinical decision-making, improve quality of care, and pos-sibly reduce hospitalisation costs.93 For this purpose, carepathways are increasingly being embedded into electronic

Table 5Advantages and disadvantages of using electronic care pathways

Advantages� Provision of ‘real-time’ documentation for an up-to-date record of pat� Compliance with documentation may improve (e.g. compulsory comple

next)� Electronic links to other parts of inpatient records (e.g. investigation re

nication and efficiency of patient care� Variance tracking and collection of outcome data can be automated� Access to up-to-date clinical information may improve (e.g. links to us� Confidentiality can be maintained or enhanced (e.g. by restricting acce� Automatic generation of personalised educational information and ‘pati

and family members� Automatic backing-up of electronic records may lower the risk of losin

Disadvantages� Lack of computer hardware or computer breakdowns may lead to pro� Healthcare professionals need extra training in order to use the system� Compatibility issues (e.g. between primary and secondary care) may be� Worries about security, confidentiality, and attacks by computer viruse� Costs of web site design and management can be very substantial

patient records. The potential advantages and disadvan-tages of using electronic care pathways are summarised inTable 5.

Implementation of an effective electronic care pathwaysystem requires a massive cultural shift in the local organi-sation, to move from a traditional and trusted practiceusing pen and paper to the unknown of computerised sys-tems, where the technology is often untrustworthy andexpensive to install. Nevertheless, the evolution is happen-ing and care pathways are likely to become an integral partof this change, especially in a clinical environmentwhere ‘continuous quality improvement’ is an essentialcomponent.

9.2. Patient pathways

Patient pathways are care pathways that are designedfor the patient and/or family to read and keep. There aretwo main types of patient pathways. The more commontype is the generic patient pathway that is given to allstroke patients and they usually contain educational infor-mation about stroke and the main strategies of manage-ment. In many cases, this type of patient pathway is nomore than a fact sheet to inform them of what to expectin hospital.

The other, and more useful, type of patient pathway isthe individualised pathway that contains informationwhich is tailored towards the patient.24 For instance, ifinvestigations revealed a high cholesterol level and signifi-cant carotid artery stenosis, then the information given tothe patient could be specific to these areas (such as adviceon low fat diet, use of statins, and information regardingcarotid endarterectomy). The pathway can also generatea ‘progress report’ on the patient’s recovery and rehabilita-tion goals that have been set by the multidisciplinary team.

The idea is that the patient and family members can taketheir time to read and digest the information and return

ient progresstion of all required ‘fields’ on one electronic page before going onto the

sults, outpatient visits) and primary care sector may improve commu-

eful web-based resources)ss or password protection)ent pathway’ (see below) may improve communication with the patient

g vital clinical information

blem with access and delays in patient care

problematics

198 J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200

with further questions. It is hoped that communication be-tween healthcare staff and patient (and family) can im-prove, which might lead to higher level of satisfactionand possibly fewer complaints and litigations. Moreover,this type of patient pathway can arm the patient with theinformation that is needed to accept responsibility for theirown care, their recovery and their discharge planning.22

The patient pathway has ramification for the healthcarestaff too – the patient will no longer be kept ‘in the dark’about his or her care. When there is discrepancy betweenthe suggested intervention and what was actually delivered,then the patient and family may want to know why.

An electronic care pathway could be designed to auto-matically generate these individualised patient pathwaysat regular intervals, making sure that the patient andthe family is kept fully informed at all times. Like all carepathways, the content of a patient pathway should bekept up-to-date according to the latest evidence andguidelines.

9.3. Pre-hospital care pathways for hyperacute stroke

Stroke management begins when the emergency medicalservices personnel (including paramedical staff) first arriveat the scene to assess and transport the patient to theappropriate hospital. In a recent systematic review of thebarriers to acute stroke therapy, many of them were foundto be in the pre-hospital stage:94 (i) patient or family doesnot recognise symptoms of stroke or seek urgent help;(ii) patient or family calls a primary care physician ratherthan an ambulance; (iii) paramedical staff do not recognisethe symptoms of stroke; and (iv) paramedical staff do nottriage stroke as an emergency and transport of the patientto hospital is non-urgent.

While the first two barriers could only be tackled bypublic educational campaigns, the latter two might beovercome by using well-designed acute stroke protocolsor care pathways. Pre-hospital care pathways already ex-ist in other emergency conditions such as acute myocar-dial infarction and hypoglycaemia. Pre-hospital carepathways are probably most useful for emergency condi-tions which must be diagnosed and treated immediatelyby the paramedical staff, or transferred to a specialist cen-tre without delay once the initial diagnosis is made. Re-cent advances in the pre-hospital treatment of acutemyocardial infarction include ‘telemedicine’ (electronictransfer of the electrocardiogram [ECG] to the nearestspecialist centre) and the administration of a thrombolyticagent by paramedical staff once the patient is confirmedas having a myocardial infarction by the specialist centre.Care pathways for such a condition may improve thediagnostic accuracy and assist in the emergency manage-ment, including immediate life support, pain relief, andinterpretation of the ECG.95 For those who might be eli-gible for immediate thrombolysis, a checklist or algorithmto screen for contraindications is paramount to avoid pro-tocol violation and unnecessary adverse events.

Accuracy of pre-hospital diagnosis of stroke is alsobecoming more important as new acute treatments arebeing developed and tested. In the future, pre-hospitalstroke care might involve the immediate administrationof a neuro-protective agent by paramedical staff.96 Forexample, a clinical trial is underway to evaluatethe effectiveness of pre-hospital administration ofmagnesium sulphate (FAST-MAG Trial) for patients withhyperacute stroke (http://clinicaltrials.gov/show/NCT000-59332). While the aim is to speed up the transfer ofappropriate patients to specialist centres for emergencytreatments, it is also important to quickly diagnose andtreat non-stroke conditions such as epileptic seizures andhypoglycaemia. Various stroke assessment tools and train-ing programs have been shown to have a positive impacton the accuracy of pre-hospital diagnosis of stroke as wellas the speed of transfer to hospital.

Paramedical staff are in a privileged position – they ar-rive at the scene where the suspected stroke occurs andmuch of the circumstantial information can provide valu-able clues about the diagnosis and possible cause of theproblem. A thorough history from the bystander can oftenhelp to distinguish a stroke from a stroke-mimic; for exam-ple, did the patient suffer a seizure prior to collapsing witha hemiparesis? A well-designed pre-hospital care pathwaycan assist with the following: (i) clinical assessment of a pa-tient with suspected stroke; (ii) immediate life supportincluding airway, breathing and circulation; (iii) immediateexclusion and treatment of important non-stroke condi-tions such as hypoglycaemia; and (iv) urgent transfer ofthe appropriate patients to the hospital for emergencytreatments, including notifying the emergency departmentof their imminent arrival. As new treatments are developedand licensed to be used within the first few hours afterstroke onset, pre-hospital care pathways might become avaluable tool to improve the efficiency and effectivenessof emergency stroke care.

Acknowledgement

Financial support: The author’s salary was supported bythe Glaxo Wellcome Stroke Training Fellowship (2000–1)and Stroke Association Clinical Training Fellowship(2001–2).

References

1. de Luc K. Are different models of care pathways being developed? Int

J Health Care Qual Assur 2000;13:80–6.2. Norris AC. Care pathways and the new NHS. J Integrated Care

1998;2:78–83.3. Hale C. Issues in the evaluation of multidisciplinary pathways of care.

In: Wilson J, editor. Integrated Care Management: the Path to

Success? Oxford: Butterworth/Heinemann; 2002.4. Overill S. A practical guide to care pathways. J Integrated Care

1998;2:93–8.5. Field MJ, Lohr KN. Guidelines for Clinical Practice: from Develop-

ment to Use. Washington DC: National Academy Press; 1992.6. Hunter DJ, Fairfield G. Disease management. BMJ 1997;315:50–3.

J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200 199

7. National electronic Library for Health. About Integrated CarePathways (ICPs). NeLH 2004 Available from: URL: http://librar-ies.nelh.nhs.uk/Pathways/.

8. Woolf CR, Cass W, McElroy J. The use of ‘‘Program Evaluation andReview Technique’’ (PERT) in the design and control of a medicalresearch project. Comput Biomed Res 1968;2:176–86.

9. Modell ME. A Professional’s Guide to Systems Analysis. 2nded. London: McGraw-Hill; 1996.

10. Landry JA. Project Management: Comparing PERT/CPM, Gantt

Charts, Function Points, Metrics. Lecture Notes for Agricultural and

Biosystem Engineering. Canada: McGill University; 2001.11. Crummer MB, Carter V. Critical pathways - the pivotal tool. J

Cardiovasc Nurs 1993;7:30–7.12. Nelson S. Following pathways in pursuit of excellence. Int J Health

Care Qual Assur 1995;8:19–22.13. Bailey R, Weingarten S, Lewis M, et al. Impact of clinical pathways

and practice guidelines on the management of acute exacerbations ofbronchial asthma. Chest 1998;113:28–33.

14. Kitchiner D, Bundred P. Clinical pathways. A practical tool forspecifying, evaluating and improving the quality of clinical practice.Med J Aust 1999;170:54–5.

15. Campbell H, Hotchkiss R, Bradshaw N, et al. Integrated carepathways. BMJ 1998;316:133–7.

16. Riley K. Paving the way. Care pathways are a tool to standardise care.Health Serv J 1998;26:30–1.

17. Quigley PA, Wallace Smith S, Strugar J. Successful experience withclinical pathways in rehabilitation. J Rehabil 1998;64:29–32.

18. Yandell B. Critical paths at Alliant Health System. Qual Manag

Health Care 1995;3:55–64.19. Harbison J, Massey A, Barnett L, et al. Rapid ambulance protocol

for acute stroke. Lancet 1999;353:1935.20. Brown SW, Nemeth LS. Developing a variance reporting system to

facilitate quality improvement. Outcomes Manag Nurs Pract 1998;2:10–5.21. Schriefer J. Managing critical pathway variances. Qual Manag Health

Care 1995;3:30–42.22. Dykes PC. Designing and implementing critical pathways: an over-

view. In: Dykes PC, Wheeler K, editors. Planning, Implementing and

Evaluating Critical Oathways. New York: Springer Publishing Com-pany; 1997. p. 8–25.

23. Roberts HC, Pickering RM, Onslow E, et al. The effectiveness ofimplementing a care pathway for femoral neck fracture in olderpeople: a prospective controlled before and after study. Age Ageing

2004;33:178–84.24. Zander K. Integrated care pathways: eleven international trends. J

Integr Care Pathways 2002;6:101–7.25. Pearson SD, Goulart-Fisher D, Lee TH. Critical pathways as a

strategy for improving care: problems and potential. Ann Intern Med

1995;123:941–8.26. Thomson R, Lavender M, Madhok R. How to ensure that guidelines

are effective. BMJ 1995;311:237–42.27. Sutton AJ, Duval SJ, Tweedie RL, et al. Empirical assessment of

effect of publication bias on meta-analyses. BMJ 2000;320:1574–7.28. Bertram DA, Thompson MC, Giordano D, et al. Implementation of

an inpatient case management program in rural hospitals. J Rural

Health 1996;12:54–66.29. Harding J. Practice guidelines. Cookbook medicine. Physician Exec

1994;20:3–6.30. Wilson J. Best practice guidelines. Br J Nurs 1999;8:293–4.31. O’Malley S. Pathways: improving outcomes, not just ‘cookbook

medicine’. Qual Lett Healthc Lead 1997;9:12–4.32. Wigfield A, Boon E. Critical care pathway development: the way

forward. Br J Nurs 1996;5:732–5.33. Miller LM, Nugent KP. Surgical integrated care pathway develop-

ment: compliance and staff satisfaction. J Integr Care Pathways

2003;7:36–46.34. Wee AS, Cooper WB, Chatham RK, et al. The development of a

stroke clinical pathway: an experience in a medium-sized communityhospital. J Miss State Med Assoc 2000;41:648–53.

35. Zander K. Nursing case management: strategic management of costand quality outcomes. J Nurs Adm 1988;18:23–30.

36. Hickman JL. Outcomes management for stroke patients usingthrombolytics. Crit Care Nurs Clin N Am 1998;10:101–15.

37. Broderick JP. Practical considerations in the early treatment ofischemic stroke. Am Fam Physician 1998;57:73–80.

38. Schmidt SM, Guo L, Scheer S, et al. Epidemiologic determination ofcommunity-based nursing case management for stroke. J Nurs Adm

1999;29:40–7.39. Dignan MB, Howard G, Toole JF, et al. Evaluation of the North

Carolina stroke care program. Stroke 1986;17:382–6.40. Adams Jr HP, Brott TG, Furlan AJ, et al. Guidelines for thrombo-

lytic therapy for acute stroke: a supplement to the guidelines for themanagement of patients with acute ischemic stroke. A statement forhealthcare professionals from a special writing group of the StrokeCouncil, American Heart Association. Stroke 1996;27:1711–8.

41. Langhorne P, Dennis M. Stroke Units: an Evidence Based

Approach. London: BMJ Books; 1998.42. Abissi CJ, Sepe E, Patlak C, et al. Cerebral infarction : comparison of

a care plan with case-management to traditional care [abstract].Neurology 1995;45 (Suppl 4):A240. (280P).

43. Barch C, Spilker J, Bratina P, et al. Nursing management of acutecomplications following rt-PA in acute ischemic stroke. The NINDSrt-PA Stroke Study Group. J Neurosci Nurs 1997;29:367–72.

44. Duryee R, Calancjini P, Miller R. The impact of a stroke pathway:measuring effectiveness [abstract]. Neurology 1996;46 (Suppl):A428.(S63.007).

45. Hainsworth DS, Lockwood-Cook E, Pond M, et al. Developmentand implementation of clinical pathways for stroke on a multihospitalbasis. J Neurosci Nurs 1997;29:156–62.

46. Jungkind K, Corish C. Critical path network. Pilot acute ischaemicstroke program saves $9756 per case. Hosp Case Manag 1999;7:87–90.

47. Lagoe RJ, Aspling DL. Benchmarking and clinical pathway imple-mentation on a multihospital basis. Nurs Econ 1997;15:131–7.

48. Odderson IR, Keaton JC, McKenna BS. Swallow management inpatients on an acute stroke pathway: quality is cost effective. Arch

Phys Med Rehabil 1995;76:1130–3.49. Ramachandran TS, Culebras A, Hainworth D. Development and

implementation of a clinical pathway for stroke in acute care[abstract]. Neurology 1996;46 (Suppl):A319. (P04.119).

50. Summers D, Soper PA. Implementation and evaluation of strokeclinical pathways and the impact on cost of stroke care. J Cardiovasc

Nurs 1998;13:69–87.51. Wentworth DA, Atkinson RP. Implementation of an acute stroke

program decreases hospitalization costs and length of stay. Stroke

1996;27:1040–3.52. Wilkinson G, Parcell M, MacDonald A. Cerebrovascular accident

clinical pathway. J Qual Clin Pract 2000;20:109–12.53. Baker CM, Miller I, Sitterding M, et al. Acute stroke patients

comparing outcomes with and without case management. Nurs Case

Manag 1998;3:196–203.54. Ross G, Johnson D, Kobernick M. Evaluation of a critical pathway

for stroke. J Am Osteopath Assoc 1997;97:269–72, 275–6.55. Bowen J, Yaste C. Effect of a stroke protocol on hospital costs of

stroke patients. Neurology 1994;44:1961–4.56. Pasquarello MA. Measuring the impact of an acute stroke program

on patient outcomes. J Neurosci Nurs 1990;22:76–82.57. Hacke W, Kaste M, Olsen TS, et al. European Stroke Initiative (EUSI)

recommendations for stroke management. Eur J Neurol 2000;7:607–24.58. Dykes PC. Working out comorbidities and variances with copathways

and algorithms. In: Dykes PC, Wheeler K, editors. Planning,

Implementing and Evaluating Critical Pathways. New York: SpringerPublishing Company; 1997. p. 26–32.

59. Romito D. A critical path for CVA patients. Rehabil Nurs

1990;15:153–6.60. Falconer JA, Roth EJ, Sutin JA, et al. The critical path method in

stroke rehabilitation: lessons from an experiment in cost containmentand outcome improvement. Qual Rev Bull 1993;19:8–16.

200 J. Kwan / Journal of Clinical Neuroscience 14 (2007) 189–200

61. Sulch D, Perez I, Melbourn A, et al. Randomized controlled trial ofintegrated (managed) care pathway for stroke rehabilitation. Stroke

2000;31:1929–34.62. Goldberg G, Segal ME, Berk SN, et al. Stroke transition after

inpatient rehabilitation. Top Stroke Rehabil 1997;4:64–79.63. Crawley WD. Case management: improving outcomes of care for

ischaemic stroke patients. Medsurg Nurs 1996;5:239–44.64. Hamrin EKF, Lindmark B. The effect of systematic care planning

after acute stroke in general hospital medical wards. J Advanced Nurs

1990;15:1146–53.65. Moloney A, Critchlow B, Jones K. A multi-disciplinary care

pathway in stroke - does it improve care? [Abstract]. Age Ageing

1999;28 (Suppl 1):42–3.66. Rosenberger JM, Wiemers NE. CareMaps in medical rehabilitation. J

Care Manag 1999;5:23–37.67. Schull DE, Tosch P, Wood M. Clinical nurse specialists as collabo-

rative care managers. Nurs Manage 1992;23:30–3.68. Jones S, Sulch D, Starke I. Effect of the introduction of a care

pathway and educational programme on quality of care and outcomein stroke [Abstract]. Age Ageing 2001;30 (Suppl 2):54.

69. Morrison LG, Lam S, Sutherland M, et al. A unitary patient recordimproves admission documentation in a medical assessment unit in amajor teaching hospital. Health Bull (Edinburgh) 2001;59:218–23.

70. Juni P, Altman DG, Egger M. Systematic reviews in health care:assessing the quality of controlled clinical trials. BMJ 2001;323:42–6.

71. Medical Research Council. Framework for Design and Evaluation of

RCTs for Complex Interventions to Improve Health. London: Med-ical Research Council; 2000.

72. Peto R, Baigent C. Trials: the next 50 years. Large scale randomisedevidence of moderate benefits. BMJ 1998;317:1170–1.

73. Pocock SJ, Elbourne DR. Randomised trials or observationaltribulations? N Engl J Med 2000;342:1907–9.

74. Torgerson DJ. Contamination in trials: is cluster randomisation theanswer? BMJ 2001;322:355–7.

75. Kwan J. Integrated care pathways for acute stroke: an evaluation oftheir effects using multiple approaches [MD Thesis]. Edinburgh:Edinburgh University; 2002.

76. Mays N, Pope C. Qualitative research in health care: Assessing qualityin qualitative research. BMJ 2000;320:50–2.

77. Rees CE, Bath PA. The use of between-methods triangulation in cancernursing research: a case study examining information sources forpartners of women with breast cancer. Cancer Nurs 2001;24:104–11.

78. Murray JS. Methodological triangulation in a study of social supportfor siblings of children with cancer. J Pediatr Oncol Nurs 1999;16:194–200.

79. Connelly LM, Bott M, Hoffart N, et al. Methodological triangulationin a study of nurse retention. Nurs Res 1997;46:299–302.

80. Friedemann ML, Smith AA. A triangulation approach to testing afamily instrument. West J Nurs Res 1997;19:364–78.

81. Kwan J, Sandercock P. In-hospital care pathways for stroke. The

Cochrane Database of Systematic Reviews 2004 (Issue 4).82. Kwan J, Hand P, Dennis M, et al. Effects of introducing an integrated

care pathway in an acute stroke unit. Age Ageing 2004;33:362–7.83. Mosimaneotsile B, Raun K, Tokishi C. Stroke patient outcomes: does

an integrated delivery model of care make a difference? Phys Occu

Ther Geriatr 2000;17:67–82.84. Odderson IR, McKenna BS. A model for management of patients

with stroke during the acute phase. Outcome and economic implica-tions. Stroke 1993;24:1823–7.

85. Widjaja LS, Chan BP, Chen H, et al. Variance analysis applied to astroke pathway: how this can improve efficiency of healthcaredelivery. Ann Acad Med Singapore 2002;31:425–30.

86. Wilkinson G, Parcell M, McDonald A. Finalist: ACHS QualityImprovement Award. Cerebrovascular accident clinical pathway. J

Qual Clin Pract 2000;20:109–12.87. Sulch D, Evans A, Melbourn A, et al. Does an integrated care

pathway improve processes of care in stroke rehabilitation? Arandomized controlled trial. Age Ageing 2002;31:175–9.

88. Sulch D, Melbourn A, Perez I, et al. Integrated care pathwaysand quality of life on a stroke rehabilitation unit. Stroke 2002;33:1600–4.

89. Lanska DJ. The role of clinical pathways in reducing the economicburden of stroke. Pharmacoeconomics 1998;14:151–8.

90. Gainer FE. Developing a clinical pathway for stroke. OT Pract

1996;1:30–3.91. Luttman RJ. The critical path method alone does nothing to improve

performance. Qual Rev Bull 1993;19:142–3.92. Ringel SP, Hughes RL. Evidence-based medicine, critical pathways,

practice guidelines, and managed care. Reflections on the preventionand care of stroke. Arch Neurol 1996;53:867–71.

93. Sado AS. Electronic medical record in the intensive care unit. Crit

Care Clin 1999;15:499–522.94. Kwan J, Hand P, Sandercock P. A systematic review of barriers to

delivery of thrombolysis for acute stroke. Age Ageing 2004;33:116–121.

95. McNamara NS, Wharton TP. Critical pathways for primary angio-plasty in acute myocardial infarction at community hospitals withoutcardiac surgery. Crit Path Cardiol 2002;1:74–102.

96. Suyama J, Crocco T. Prehospital care of the stroke patient. Emerg

Med Clin North Am 2002;3:537–52.