Report - Department of Health · Web viewGiven that bariatric surgery is a significant procedure with potential complications, the provision of bariatric surgery should be targeted

Draft report for reviewing existing MBS items

Review name: Items for the surgical treatment of obesity

Department of Health and Ageing

24 May 2011


ContentsGlossary i

Types of bariatric surgeryiii

Foreword from the Department of Health and Ageing i

Executive Summary iv

1 Introduction to MBS Demonstration Reviews 11.1 Principles to guide MBS reviews 11.2 Purpose of this document 2

2 Background on the MBS surgical items for obesity 32.1 Description of the current services 32.2 Context 92.3 Justification for review 12

3 Clinical/research questions 143.1 Population 143.2 Intervention 153.3 Comparator 163.4 Outcomes 173.5 Research questions 17

4 Key stakeholders 194.1 Clinical Working Group 194.2 Clinical craft groups and others 194.3 Consumers and the general public 194.4 Consultants 204.5 The Department of Health and Ageing 20

5 Review methodology 215.1 Literature review 215.2 MBS data 225.3 Economic evaluation23

6 Data analysis 256.1 Breakdown of MBS claims for the treatment of obesity 266.2 Breakdown of items by broad age group 306.3 Breakdown of services claimed by ten-year age groups 316.4 Breakdown by gender 326.5 MBS linked data 33

7 Systematic review of meta-analyses, systematic reviews and evidence-based clinical guidelines 46

7.1 Methods for the systematic literature search 467.2 Review of meta-analyses and systematic reviews 527.3 Review of clinical guidelines62

Deloitte Access Economics


7.4 Review of registry and other relevant data 687.5 Conclusions of the clinical literature review 68

8 Systematic review of economic evaluations 708.1 Methods for the systematic literature search 708.2 Results of the systematic literature search 748.3 Systematic review of the primary economic studies 778.4 Conclusions from the economic literature review84

9 Conclusions 859.1 Conclusions regarding the MBS items 859.2 Conclusion regarding the adolescent patient population 889.3 General/other recommendations 88

References 90

Appendix A : Final literature review protocol100

Appendix B : Overview of the studies identified for the clinical literature review 104

Appendix C : Overview of the primary studies identified for the economic literature review168

ChartsChart 2.1 : Weight-for-age percentiles for boys 6

Chart 2.2 : Weight-for-age percentiles for girls 7

Chart 2.3 : Trends in obesity prevalence for adults, 1980 to 2007 11

Chart 6.1 : Total number of MBS services claimed for the treatment of obesity by year 26

Chart 6.2 : Total number of services claimed in the last five years by MBS items for the treatment of obesity 27

Chart 6.3 : Total number of MBS surgical obesity procedures claimed, by year 27

Chart 6.4 : Total number of services claimed relating to maintenance of AGB, by year 29

Chart 6.5 : Total number of MBS surgical reversal procedures claimed, by year 30

Chart 6.6 : Proportion of bariatric surgery items claimed by item number and broad age group31

Chart 6.7 : Service utilisation by MBS Item number and ten-year age groups 32

Chart 6.8 : Top 10 MBS item numbers claimed with item 30511 34








TablesTable 2.1 : MBS services being reviewed 3

Table 2.2 : Age and gender thresholds for overweight and obesity in children and adolescents5

Table 2.3 : Waist circumference cut-offs correlated to disease risk 8

Table 2.4 : Diseases and conditions associated with obesity 8

Table 2.5 : Prevalence rates for obesity by age and gender 10

Table 2.6 : Percentage of urban Australian adults aged 25-64 years in each obesity category in 1980 and 2000 11

Table 2.7 : MBS expenditure by item 12

Table 6.2 : Growth in the number of MBS items claimed under primary bariatric services* in the past five years 26

Table 6.3 : Total number of MBS surgical treatment of obesity claims, by year 28

Table 6.4 : Total number of services claimed relating to maintenance of AGB, by year 29

Table 6.5 : Total number of MBS surgical reversal procedures claimed, by year 30

Table 6.6 : Total numbers of MBS items claimed by gender 33

Table 6.7 : Proportion of people receiving primary surgical procedures 33

Table 6.8 : Descriptors of the top 10 MBS item numbers claimed with item 30511 34






Table 7.1 : Numbers of clinical studies identified and included in the literature review 47

Table 7.2 : Reasons for exclusion 48

Table 7.3 : Systematic reviews and meta-analyses included in the clinical literature review49

Table 7.4 : Guidelines reviewed 51

Table 8.1 : Numbers of economic studies identified and included in the literature review 71

Table 8.2 : Reasons for exclusion 72

Table 8.3 : Total numbers of studies included in the literature review 72

Table 8.4 : Primary economic studies included in the literature review 73

Table A.1 – Embase.com search, <1966 to 1 September 2010 (*) 100

Table A.2 – The Cochrane Library search, 2010 Issue 8 (*) 102



Table A.3 – Breakdown of database retrieval from The Cochrane Library, 2010 Issue 8 103

Table B.1 : Evidence hierarchy 104

Table B.2 : Body of evidence matrix for assessing guidelines 105

Table B.3 : Assessment of evidence base on meta-analyses 105

Table B.4 : Assessment of evidence base on systematic reviews 108

Table B.5 : Assessment of evidence base on clinical guidelines 114

Table B.6 : Data extracted from meta-analyses 116

Table B.7 : Data extracted from systematic reviews122

Table B.8 : Data extracted from clinical guidelines 148

Table C.1 : Data extracted from economic studies 168

FiguresFigure 3.1 : Clinical decision pathway for the management of the overweight or obese person

15

Figure 3.2 : Flowchart of patient pathway through MBS items under review 16



GlossaryABS Australian Bureau of Statistics

AIHW Australian Institute of Health and Welfare

AGB adjustable gastric banding

ASGB adjustable silicone gastric banding

BOLD Bariatric Outcomes Longitudinal Database

BPD(-DS) biliopancreatic diversion (with duodenal switch)

BMI body mass index (in kg/m2)

CEA cost effectiveness analysis

CHD coronary heart disease

CHF Consumer Health Forum of Australia

CI confidence interval

CM conventional management

CORE Center for Outcomes Research

CUA cost utility analysis

CWG Clinical Working Group

CVD cardiovascular disease

DALY disability adjusted life year

DEALE Declining Exponential Approximation of Life Expectancy

DOHA Department of Health and Ageing

DS duodenal switch

EWL excess weight loss

GB gastric bypass

GDP gross domestic product

GORD gastro-oesophageal reflux disease

HTA Health Technology Assessment

HRQOL health-related quality of life

ICER incremental cost effectiveness ratio

ICUR incremental cost utility ratio

ITT Intention to treat

LABS Longitudinal Assessment of Bariatric Surgery

(L)AGB (laparoscopic)* adjustable gastric banding

LY(G) life year (gained)


MBS Medicare Benefits Schedule

MSAC Medical Services Advisory Committee

NHMRC National Health and Medical Research Council

NICE National Institute for Health and Clinical Excellence

NIH National Institutes of Health

PBAC Pharmaceutical Benefits Advisory Committee

PSA probabilistic sensitivity analysis

QALY quality-adjusted life year

RCT randomised controlled trial

QOL quality of life

RR relative risk

RYGB Roux-en-Y gastric bypass

SAND Supplementary Analysis of Nominated Data

SD standard deviation

SG sleeve gastrectomy ( a type of gastrectomy)

SOS Swedish Obese Subjects (study)

SPANS Schools Physical Activity and Nutrition Survey

T2DM type 2 diabetes mellitus

UK UK

US United States

VAS visual analogue scale

VBG vertical banded gastroplasty (a type of gastroplasty)

WHO World Health Organization* Other surgical techniques when preceded by ‘L’ also indicate laparoscopy.



Types of bariatric surgerySurgical

ProcedureCurrent

MBS item*

Description of surgical procedure

Illustration of surgical procedure

Adjustable gastric banding (AGB)/Laparoscopic adjustable gastric banding (LAGB)

30511 Is a surgical procedure in which a small silicone band is placed around the top of the stomach to produce a small pouch about the size of a thumb, thereby limiting food intake.

Biliopancreatic diversion (BPD)

Biliopancreatic diversion with duodenal switch (BPD-DS)

30512 The first two segments of the small intestine, the duodenum and jejunum, are bypassed and the stomach pouch is attached to the ileum.

BPD in conjunction with DS is an additional adaptation where a proportion of the duodenum remains attached to the stomach.

Roux-en-Y gastric bypass (RYGB)

30512 A small stomach pouch is created to restrict food intake. Next, a Y-shaped section of the small intestine is attached to the pouch to allow food to bypass the lower stomach, the duodenum (the first segment of the small intestine), and the first portion of the jejunum (the second segment of the small intestine).



Sleeve gastrectomy (SG)

Also known as:gastric sleeve, tube gastrectomy

Various MBS items 30511 or 30518

Is the first component of the duodenal switch operation and involves removing the lateral 2/3rds of the stomach with a stapling device. It leaves a stomach tube instead of a stomach sack.

Vertical banded gastroplasty (VBG)

Also known as:stomach stapling

30511 The upper stomach near the oesophagus is stapled vertically to create a small pouch along the inner curve of the stomach. The outlet from the pouch to the rest of the stomach is restricted by a band.

*See section 2.1.1 of the Report for item descriptions. Source: Clinical Working Group and MSAC 2003.




Foreword from the Department of Health and AgeingBackground and purpose of review

The purpose of this review is to evaluate current evidence and clinical best practice relating to Medicare Benefits Schedule (MBS) items for the surgical interventions for the treatment of obesity. The procedures considered in the review include: adjustable gastric banding; vertical banded gastroplasty; sleeve gastrectomy; Roux-en-Y gastric bypass; and, biliopancreatic diversion with or without duodenal switch. Non-surgical interventions were deemed out of scope and only procedures currently practised in Australia are included. A diagram with details of each type of surgery is included in the report.

The review was carried out by Deloitte Access Economics with the support of a clinical working group (CWG). The role of the CWG was to provide clinical input to the consultants, to ensure the review reflects current clinical practice in Australia and draws valid conclusions from the evidence. The Department would like to thank the CWG members for the vital assistance provided in undertaking this review. The CWG consists of specialists nominated by: Obesity Surgery Society of Australia and New Zealand; and Royal Australasian College of Physicians.

A review protocol outlining the key research questions for the review was developed and modified through broad public consultation. The final review protocol was made public in January 2011.

Obesity is a disease in which fat has accumulated to the point where health is impaired, defined here in terms of Body Mass Index (BMI) over 30 for adults and a set of age-gender specific BMI thresholds for children and adolescents aged 2 to 18 years.

Clinically severe obesity is a condition generally defined as BMI ≥ 40 or between 35 and 40 where there are other major medical conditions such as high blood pressure and diabetes. It should be noted that BMI values in different populations may not correspond to the same degree of percentage of body fat or body fat distribution. As a consequence, in some ethnic groups, health risks are higher at BMIs lower than the existing defined BMI cut-off points.

i



The research questions addressed by the review were as follows: What is the safety of LAGB, VBG, SG and RYGB compared to non-surgical treatment of

obesity? What is the safety of VBG, SG and RYGB compared to LAGB in the treatment of obesity? What is the effectiveness of LAGB, VBG, SG and RYGB compared to non-surgical

treatment of obesity? What is the effectiveness of VBG, SG and RYGB compared to LAGB in the treatment of

obesity? What is the safety of each intervention compared to other relevant comparators in the

treatment of obesity? What is the effectiveness of each intervention compared to other relevant comparators

in the treatment of obesity?

Key Conclusions MBS items for obesity surgery should be split into separate items for each specific

obesity surgery procedure. Consideration should also be given to splitting obesity surgery MBS items into separate

items for laparoscopic and open procedures. The MBS items for surgical treatment of obesity should include adolescents under

special conditions (e.g. where the patient is aged over 15 years (or 14 years of age, where it can be demonstrated that exceptional circumstances exist).

The indication ‘morbid obesity’ in the current descriptors for MBS items for the surgical treatment of obesity should be replaced with ‘clinically severe obesity’ and identify the general BMI limits relating to the term.

There should be periodic reviews of the long-term safety and efficacy of emerging surgical techniques such as sleeve gastrectomy, and the long-term efficacy and cost effectiveness of gastric banding (including an analysis of reoperation and band adjustment rates).

All of the current surgical procedures in Australia were found to be safe and cost effective but some procedures were found to be safer than other procedures and some procedures were more cost effective than others. However, the review does not recommend restricting MBS items on this basis because different procedures will better suit different patients depending on their clinical condition.

If the conclusions are adopted, changes may be made to clarify which MBS item should be claimed for a specific procedure and provide an indication of when obesity surgery should be used. The proposed splitting of MBS item numbers into separate obesity procedures could support clinical best practice by providing more explicit information to surgeons and patients.

The proposed splitting could also prevent the claiming of MBS benefits for future, innovative procedures until they have been assessed as safe and cost effective by the Medical Services Advisory Committee (MSAC).

Clinicians will retain the capacity and responsibility to recommend to patients the most appropriate procedure for their circumstances.

ii



If conclusions arising from the review are accepted they will be discussed with stakeholders, including craft groups and consumers before they are implemented.

Next steps

Once public consultation has been undertaken the Review Report will be refined, as necessary, and will then be considered by MSAC. Any item amendments arising from MSAC’s advice regarding the findings of the review will need to be considered by Government.

iii



Executive SummaryA system for reviewing existing MBS items has been initiated of which this ‘demonstration review’ is one of the first. Its purpose is to evaluate the current evidence and clinical best practice relating to surgical treatment of obesity (bariatric surgery) to inform ongoing Government decisions in relation to aligning the existing MBS obesity items with contemporary clinical evidence. Non-surgical interventions are out of scope.

In line with the review protocol, which was developed with broad based public consultation, and which was revised in light of that consultation and a final protocol made public in January 2011, the review process comprised: background on surgical obesity items and development of research questions; consultation with key stakeholders; Medicare data analysis; clinical and economic literature review; and conclusions from the evidence.

Background and research questions

Obesity is a disease in which fat has accumulated to the point where health is impaired, defined here in terms of Body Mass Index (BMI) over 30 for adults and, for children and adolescents aged 2 to 18 years, a set of age-gender specific BMI thresholds. Clinically severe obesity is a condition generally defined as BMI ≥ 40 kg/m2, or between 35 kg/m2 and 40 kg/m2 where there are other major medical conditions such as high blood pressure and diabetes. It should be noted that BMI values in different populations may not correspond to the same degree of percentage body fat or body fat distribution. As a consequence, in some ethnic groups (e.g., Asian people), health risks are higher at BMIs lower than the existing defined BMI cut-off points1.

Epidemiological data on prevalence and demographic data from the Australian Bureau of Statistics (ABS) indicate that in 2008, 3.71 million Australians (17.5% of the population) were obese and, by 2025, this is projected to increase to 4.6 million Australians (18.3% of the population). MBS expenditure on the six items considered as part of the review has increased from $6.3 million in 2005 to $19.3 million in 2009. There are downstream effects of obesity (from associated diabetes, cardiovascular disease, cancers and osteoarthritis) on other MBS items, health system expenditures, productivity and other impacts, with the financial costs of obesity totaling around $8.3 billion in 2008. Hiatus hernia is frequently encountered during bariatric surgery and repair is essential to ensure good and durable long term outcome.

1 Experts in bariatric surgery who attended the recent International Federation for the Surgery of Obesity and Metabolic Disorders, Asia Pacific Chapter, Japanese Society for the Surgery of Obesity and Metabolic Disorders (IFSO-APC & JSSO) Congress 2011 reached consensus that bariatric surgery should be considered for the treatment of obesity for acceptable Asian candidates with BMI > 35 with or without comorbidities (pers. comm., Dr Ken Loi, 16 March 2011).

iv



Currently the existing MBS descriptors for surgical treatment of obesity use the terminology of morbid obesity and do not define what morbid obesity is, nor place any restrictions on the use of the various procedures, potentially hindering compliance with clinical best practice by not distinguishing procedures which are more safe or efficacious, more cost effective, or only so in a particular target population. Given that bariatric surgery is a significant procedure with potential complications, the provision of bariatric surgery should be targeted to patients for whom the benefits of surgery clearly outweigh the risks. Decision on the type of surgery should be evaluated case by case based on the balance of safety and efficacy.

The ‘PICO’ criteria were used to develop the research questions for the review. ‘P’ - the target Population for the intervention. The 2003 Australian Clinical Practice

Guidelines (currently under review) recommend bariatric surgery for adults with a BMI > 40 or with a BMI > 35 and serious medical comorbidities who have instituted but failed adequate non-operative measures for weight loss with integrated components of a dietary regimen, appropriate exercise, and behaviour modification and support (DoHA 2003a). In a severely obese adolescent with obesity-related comorbidity, bariatric surgery may be considered as an additional therapy against the background of current conventional interventions such as a reduction in energy intake, an increase in energy expenditure and behaviour modifications within the context of an ongoing and coordinated multidisciplinary approach (Baur et al 2010, DoHA 2003b).

‘I’ - the Intervention being considered. The MBS items associated with the surgical management of clinically severe obesity are 14215, 30511, 30512, 30514, 30518 and 31441. Some of these MBS items include a range of different procedures, e.g. item 30511 includes adjustable gastric banding (LAGB), vertical banded gastroplasty (VBG) and sleeve gastrectomy (SG) all of which could be performed either laparoscopically or via the open technique; and item 30512 includes Roux-en-Y gastric bypass (RYGB), biliopancreatic diversion with or without duodenal switch (BPD or BPD-DS) and jejunoileal bypass. On the other hand, for some procedures, various item numbers may have been used. For example SG has likely to have been claimed against item 30511 or item 30518.

‘C’ – the Comparator for the MBS service. Safety and efficacy of bariatric procedures were evaluated against non-surgical intervention, LAGB, and other relevant comparators as identified in the literature review and data analysis. Non-surgical treatment comprises various combinations of behaviour modification (including exercise), pharmacotherapy and/or low-energy or very low-energy diets and in some instances, no treatment. LAGB is an intervention as well as a comparator, as it is by far the most widely used technique in Australia, making up the large majority of all procedures.

‘O’ – the clinical Outcomes most relevant to assess safety and effectiveness. Safety was assessed in terms of procedure-related complications and adverse events such as the rate of reoperation, post-operative length of hospital stay, morbidity and mortality (short term e.g. 90 days and long term e.g. 10 years). Procedural specific benefits were evaluated based on weight loss, maintenance of weight loss, quality of life, and improvement in and resolution of related comorbidities.

v



The clinical research questions determined for this review using the above PICO criteria were as follows: What is the safety of LAGB, VBG SG, and RYGB compared to non-surgical treatment of

obesity? What is the safety of VBG, SG and RYGB compared to LAGB in the treatment of obesity? What is the effectiveness of LAGB, VBG, SG and RYGB compared to non-surgical

treatment of obesity? What is the effectiveness of VBG, SG and RYGB compared to LAGB in the treatment of

obesity?

Consultation with key stakeholders

A Clinical Working Group (CWG) was established for the duration of the review, with five members, with an additional paediatric obesity expert on the Deloitte Access Economics review team. The draft protocol and draft review report were uploaded to the Department of Health and Ageing (DoHA) website and as such there are no exclusions from the consultation process. DoHA identified a number of organisations, including clinical craft groups, who were notified in writing of the review and asked to comment on the draft protocol and again on the draft review report. Consumers and the general public (including service providers, device manufacturers, sponsors of medical technologies, and the Medical Technology Association of Australia were also given opportunities to comment on the draft protocol and draft review report. In addition, two meetings were held with the Consumers Health Forum of Australia (CHF) to discuss the draft protocol and report. Deloitte Access Economics incorporated comments on the draft protocol in the final protocol, and similarly with the draft report.

Medicare data analysis

Over the past five years, the number of claims for MBS items associated with the surgical treatment of obesity increased from 55,000 services in 2005-06 to 147,000 in 2009-10. Growth averaged 38% per annum in 2007-08 and 2008-09, falling to 8% growth in 2009-10, representing a substantial increase in the number of services performed per 10,000 Australians. Over 97% of the total number of services were claimed under two items – 14215 (gastric band adjustments) with 87.0% of the total, and 30511 (which includes AGB, VBG and SG procedures) with 10.1%. In terms of primary surgical procedures for obesity, item 30511 (gastric reduction or gastroplasty) increased strongly in 2008-09, but then services fell by 14% in 2009-10, the reason for which is unclear since the item includes many different procedures. MBS item 30518 (partial gastrectomy) grew strongly and steadily thought this is not solely for surgical management of obesity, while MBS item 30512 (gastric bypass) first fell and then increased. Maintenance procedure items relating gastric reduction (item 14215 and 31441) have steadily increased over the past five years. The trend in the number of surgical reversals of bariatric surgery performed (item 30514) also increased steadily.

Gastric reduction or gastroplasty (item 30511) is the most common method of surgical treatment of obesity across all age groups. Among older Australians (65+), there is a higher proportion of item 30518 (partial gastrectomy) claimed, which may be for different non-obesity related reasons such as stomach cancer or benign conditions of stomach tumor.

vi



Most patients were aged 35-59 years; 77.7% of primary surgical procedures (item 30511, 30512 and 30518) and 83% of maintenance procedures (item 14125 and 31441) were for females. Women of reproductive age (15-49) comprise 55.5% of all bariatric surgeries.

The most common same-day MBS items claimed with item 30511 were assistance during the operation (item 51303), anaesthesia-related services (item 20791) and surgeon consultations (item 17610) (including blood pressure monitoring, repair of hernia, and division of adhesions to free relevant organs). Partial gastrectomy (item 30518) was associated mostly with consultations and diagnostic laparoscopy since it is not indicated solely for obesity, it was also associated with items involving organs not confined to the stomach area (e.g. suggesting this item used as part of other non-obesity related surgery). A more diverse combination of procedures was performed with item 30512 (e.g. biopsy, imaging of the chest and, interestingly, surgical removal of gallbladder). Band adjustments (item 14215) were associated primarily with specialist and general practitioner consultations, while repair/revision/ replacement (item 31441) was associated with consultations, laparoscopy, surgery assistance, anaesthesia-related claims and wound debridement. Reversals were associated primarily with surgery assistance and division of adhesions.

Clinical and Economic Evidence Review

EMBASE (Embase and Medline) and the Cochrane library were searched with the search terms and limits presented in Appendix A. The search strategy ensured all randomised controlled trials (RCTs) were included within the Grade I studies (meta-analyses and systematic reviews), more recent Grade II studies (RCTs) or evidence-based clinical guidelines, with systematic exclusion criteria to retain 62 clinical studies of which 22 were ‘high quality’ (again based on systematic criteria outlined in Chapter 7). Safety, efficacy and cost effectiveness (economic evaluation) were considered in the evidence review.

Gastric banding (mainly laparoscopic due to its recency and superiority over the open procedure) was found to be the most common weight loss operation in Australia, with the advantages of no alternation to gastrointestinal tract or anastomosis, reversibility, and lower operative mortality and morbidity compared with other bariatric procedures.

Gastroplasty (mainly VBG) is now almost completely replaced by LAGB as LAGB is less invasive, adjustable, reversible, has better outcomes, and patients tend to be more satisfied with it. The VBG operation involves a partitioning of the stomach using staples without resection of stomach. There is no malabsorption (a benefit) but the main complications are high reoperation rates, wound infection, gastric leaks, stomal stenosis and pouch dilatations.

Gastric bypass (mainly RYGB) used to account for more than 80% of bariatric operations in the United States (US), however, with the advent of the LAGB procedures that percentage is changing. Gastric bypass was recommended in the US for the treatment of clinically severe obesity because it offers a balance between effectiveness and risks i.e., the procedure is considered to have acceptably low morbidity and mortality rates.

Biliopancreatic diversion (BPD) and biliopancreatic diversion with duodenal switch (BPD-DS) involves removing a portion of the stomach and connecting the remaining stomach to the distal part of the small intestine to induce malabsorption. In a BPD

vii



procedure the first two segments of the small intestine, the duodenum and jejunum, are by bypassed and the stomach pouch is attached to the ileum. BPD in conjunction with DS is an additional adaptation where a proportion of the duodenum is preserved and the pyloric valve stays intact. The duodenum is tolerant of stomach acid and therefore is much more resistant to ulceration while the pyloric valve regulates the release of stomach contents into the small intestine, facilitating greater nutritional uptake and reduces the occurrence of the complications such as dumping syndrome.

Sleeve Gastrectomy (SG) was developed as the first part of a two-part surgical procedure for patients who are at high risk from complications, being followed at a later date (after a 6 to 12 month period) by a gastric bypass or a duodenal switch. SG is a relatively new procedure that is short in duration, non-reversible and usually performed laparoscopically. The procedure involves dividing the stomach vertically to reduce its size to about 25%. Because laparoscopic SG is a restrictive procedure where normal digestive pathway is preserved, there are less nutritional risks compared to RYGB and other malabsorptive procedures.

Summary of Clinical Evidence Review

While the Review Report has answered the four questions presented in the final Research Protocol a broader view was taken in considering the evidence. BPD, not mentioned in the Research Protocol questions, has been included in the Review Report. The Department of Health and Ageing Foreword also includes two additional questions and they have been addressed in the discussion below.

What is the safety of LAGB, VBG, SG and RYGB compared to non-surgical treatment of obesity?

Bariatric surgery is generally considered to have acceptably low morbidity and mortality rates (Kelly et al 2005, Kelly et al 2009). However, bariatric surgery is associated with complications that may affect the patient’s quality of life, and benefits of weight loss from surgery should be balanced against the postoperative morbidity and mortality risks. Judicious patient selection and diligent peri-operative care are critical (Mechanick et al 2008).

In Australia, bariatric surgery is recommended for adults with a BMI > 40 or with a BMI > 35 and serious medical comorbidities who have instituted but failed adequate non-operative measures for weight loss with integrated components of a dietary regimen, appropriate exercise, and behaviour modification and support (DoHA 2003a). Surgery for adolescents (15 to 18 years old, and in exceptional circumstances at age 14) is only recommended in circumstances involving appropriate pre-operative education and post-operative follow-up, long-term multidisciplinary care, and adequate engagement of the young person and the family. There are certain patient subgroups for whom bariatric surgery is not recommended – for example, patients with significant cognitive disabilities and patients with untreated or untreatable psychiatric or psychological disorders.

viii



What is the safety of VBG, SG and RYGB compared to LAGB in the treatment of obesity?

Bariatric procedures are not uniformly low risk, and the common and unique risks with each type of surgery should be balanced against their benefits. Generally, potential for postoperative complications is related to surgical complexity, which is correlated with the effectiveness of the procedure.

Laparoscopic approaches to each of the bariatric procedures (gastric banding, gastric bypass, VBG, SG, and BPD-DS) have been demonstrated to achieve improved safety compared with open techniques of the same procedure. Rates of conversion to open procedure are low for LAGB and LRYGB. No conversions to open surgery have been reported for LSG, although data are limited. Laparoscopy is regarded as the routine approach to all bariatric procedures. This is because laparoscopic procedures result in the same gastrointestinal transformation as open procedures but minimise the problems related to open surgery such as wound infection, incisional hernia and general anaesthetic risk related to laparotomy. Laparoscopic procedures may not, however, be possible for all patients (e.g. the super obese, or patients who have had multiple abdominal surgeries).

LAGB is associated with lower initial operative mortality and morbidity than other surgical procedures for obesity and may therefore be preferred for people who want a safer operation with potentially lower weight loss. However, there is some concern about reoperation/revision rates despite its greater safety, faster recovery period, The relative safety of LAGB was preferred for adolescent patients despite somewhat lower excess weight loss (EWL) overall. LAGB is currently off-label for adolescents in the US; despite this, more recent trends have shown increasing utilisation of LABG in adolescents undergoing bariatric surgery.

Biliopancreatic diversion (BPD) and biliopancreatic diversion with duodenal switch (BPD-DS) is a technically demanding procedure and is associated with the highest morbidity and mortality rates of all bariatric surgery techniques. The high mortality, coupled with higher incidence of long-term nutritional and vitamin deficiencies, stomal ulceration, severe protein-energy malnutrition, and dumping has resulted in limited widespread acceptance of BPD. Risks outweigh benefits in adolescents.

RYGB has a lower morbidity and mortality than with truly malabsorptive BPD and its DS variant, but notably higher than for restrictive LAGB. Gastric bypass may result in both vitamin and mineral deficiency, but these can be overcome by appropriate supplementation and monitoring. Laparoscopic RYGB improves short-term recovery from surgery and has a lower incidence of incisional hernias, pulmonary function impairment and post-operative pain than open RYGB, while remaining equally effective. Current evidence on RYGB in adolescent patients shows safety risks and efficacy benefits. In summary compared to restrictive procedures, gastric bypass procedures have been shown to have higher complication rates including surgical mortality and long-term nutritional and surgical complications.

VBG has comparable operative safety and postoperative recovery, relative to RYGB. Like banding, complication rates are much lower and hospital stays shorter for

ix



laparoscopic compared to open VBG. Compared to LAGB, VBG is associated with increased peri- and postoperative complications, while patients who had undergone SG have greater severities of complications and are at risk for developing micronutrient deficiencies.

Morbidity following laparoscopic SG is relatively low compared with BPD-DS or RYGB with similar reduction in co-morbidities, but attracts higher early postoperative complication compared to LAGB. The lower risk of complications compared with gastric bypass means that SG may be preferred for patients at a higher risk of complications from bariatric surgery generally if LAGB is considered unsuitable. There are minor technique differences between bariatric centres and countries, until techniques are standardised and proof of longer-term efficacy becomes available, SG should only be offered to adolescents within the context of a controlled prospective study.

What is the effectiveness of LAGB, VBG, SG and RYGB compared to non-surgical treatment of obesity?

Surgery for obesity is generally only recommended for clinically severe obese patients for whom non-surgical treatments have failed to achieve and sustain adequate weight loss. Bariatric surgery for clinically severe obesity is universally reported to be more effective in inducing long-term, sustainable weight loss than non-surgical methods. Weight loss correlates with improvement in metabolic diseases such as type 2 diabetes and has a positive impact on quality of life.

Non-surgical treatment does not consistently lead to weight loss. Some follow-up studies have shown people in non-surgical groups to regain weight, despite initial weight loss. A study by Colquitt et al (2009) comparing LAGB with no surgery (i.e. non-surgical treatment) in people with a BMI 30-35 with co-morbidity found 98% and 35% of people respectively achieved satisfactory weight loss (EWL >25%) at two years. Another systematic review of long-term weight loss studies in obese adults, comparing bariatric surgery with non-surgical intervention found less than 5kg weight loss after two to seven years of dietary and lifestyle therapy, 5-10kg weight loss after one to two years of pharmacological therapy, and 25-75 kg weight loss after two to four years of undertaking bariatric surgery (Douketis et al, 2005). Large weight loss with surgery, however, may depend on ongoing dietary/lifestyle interventions as surgical therapy is usually combined with dietary therapy and/or a behaviour counselling/lifestyle modification program.

What is the effectiveness of VBG, SG and RYGB compared to LAGB in the treatment of obesity?

Biliopancreatic diversion (BPD) and biliopancreatic diversion with duodenal switch (BPD-DS) is the most effective procedure of those compared in inducing weight loss, particularly in ‘super-obese’ patients. BPD with or without DS is reported to lead to greater long-term weight loss and comorbidity resolution compared to other bariatric surgery such as gastric bypass, which in turn is superior to LAGB. Many studies found

x



LAGB to be less effective than most other procedures. VBG and LAGB are considered to achieve similar weight loss results, but weight regain is more common with VBG.

There are few relatively long-term study data and quality of life post surgery studies available for newer surgical techniques such as SG, compared with LAGB, gastric bypass and VBG. Evidence to date suggests that SG leads to greater early EWL than LAGB. When used as a first-stage procedure, laparoscopic SG has been shown to reduce weight, comorbidities, and operative risk compared to immediate DS/gastric bypass. However, for some, effective weight loss is achieved with SG alone and the procedure is now increasingly being seen as an effective stand-alone restrictive procedure for weight loss. However evidence on long-term weight loss is limited. Overall conversion rates to other procedures from SG due to insufficient weight loss are also unknown.

Three year data shows that the extent of weight loss, for each procedure is: BPD-DS (70% EWL and 35% BMI reduction), followed by RYGB (65%-70% EWL and 35% BMI reduction), SG (50%-65% EWL and 25% BMI reduction) VBG (50%-60% EWL and 25%-30% BMI reduction) and LAGB (50% EWL and 25% BMI reduction).

Laparoscopic procedures are generally associated with similar efficacy as open procedures. No significant difference in weight loss has been reported between open AGB and LAGB procedures at 12 months and both were associated with a statistically significant reduction in weight compared with baseline. Likewise, when open VBG was compared against laparoscopic VBG, similar EWL at 12 months was reported (open 55% vs. laparoscopic 47%). Weight loss achieved with open RYGB and laparoscopic RYGB have both been shown to be substantial and not significantly different between procedures. This is because the final anatomic reconfiguration is the same for laparoscopic and open RYGB, and therefore, weight loss and comorbidity outcomes are expected to be identical.

There are limited outcome data for subgroups of the morbidly obese, including adolescents and patients with particular comorbidities.

In the first 3 years data shows that weight loss appears to be greatest with BPD-DS (70% EWL and 35% BMI reduction), followed by RYGB (65%-70% EWL and 35% BMI reduction), SG (50%-65% EWL and 25% BMI reduction) VBG (50%-60% EWL and 25%-30% BMI reduction) and LAGB (50% EWL and 25% BMI reduction). Weight loss correlates with improvement in or resolution of comorbidities, and is related to surgical complexity and the potential for postoperative complications.

Summary of Economic Literature Review

It was not in the scope of this report to generate a ‘ground up’ economic evaluation. Instead, a systematic review was conducted of existing economic literature relating to the surgical management of obesity. Where international economic literature has been included in the analysis, it should be noted that the applicability of such literature to the

xi



Australian health care context is somewhat limited. The key points from the review of economic literature are summarized below:

Most economic evaluations of bariatric surgery for obesity have been published since 2005 and evaluate LAGB and gastric bypass. Importantly, there is a lack of well-performed Australian studies and cost utility analysis (CUAs).

VBG is used less frequently today than other bariatric surgery procedures, and there is limited economic evidence for VBG; however, published data suggest more recently developed surgical techniques are cost effective when compared with VBG.

Bariatric surgery for obesity is universally reported to be cost effective compared with no surgery even across extensive deterministic and probabilistic sensitivity analyses (PSAs). The likelihood of publication bias should be considered; that is, researchers finding surgery less cost effective may be less likely to seek publication. However, Picot et al (2009) performed arguably the most comprehensive economic evaluation for a NICE review, which is less likely to have vested interests, and found surgeries to be cost effective.

Many studies show surgery to be cost saving to the health care payer after several years, although the scope of costs in these studies should be carefully considered.

Based on the published literature, surgery appears to be more cost effective in women and younger people (due to greater life expectancy over which benefits accrue), and people with higher BMIs and comorbidities such as diabetes (in whom surgery makes the greatest clinical difference). Surgery is also more cost effective in people with newly diagnosed diabetes compared with established diabetes (at least two years since diagnosis).

Generally, when compared with no surgery, lower incremental cost effectiveness ratios (ICERs) are reported for LAGB than for gastric bypass. However, these results should be interpreted with caution since the cost effectiveness of one procedure versus another should only be compared using the incremental costs and benefits for one procedure versus another procedure, and within the same study to control for other factors.

In the three studies directly comparing the cost effectiveness of gastric bypass and banding (Campbell et al 2008, Campbell et al 2010, Clegg et al 2002), bypass appears cost effective relative to banding, with a relatively favourable cost for the additional clinical benefits (i.e. using the conventional threshold of $50,000 per QALY gained). However, this outcome may in part be driven by the underlying data since there are (a) a lack of head-to-head study data, and (b) a lack of evidence on long term outcomes with banding.

Laparoscopic gastric bypass appears to be cost effective compared with open gastric bypass, assuming similar outcomes, since savings in complication costs outweigh any additional procedure costs. From an economic viewpoint, laparoscopic RYGB should potentially be used over open gastric bypass unless laparoscopic procedures are contra-indicated in the patient or conversion is required during surgery.

xii



Conclusions

Conclusions in relation to MBS items

Based on the evidence in this review and to ensure further clarity around existing MBS items, that consideration be given to: Splitting MBS item 30511 (gastric reduction or gastroplasty by any method) into

several items, each describing a specific procedure. Splitting MBS item 30512 into separate items for gastric bypass and BPD with or

without DS. Specifying in MBS item 30518 the type of gastrectomy operation and the surgical

indication for this. Splitting all relevant MBS items for obesity surgery (currently items 30511 and 30512)

into separate items for laparoscopic and open procedures. Include in the MBS items for surgical treatment of obesity special conditions that

need to be met before surgery is performed in adolescents 2. The indication of ‘morbid obesity’ should be redefined for MBS items relating to

bariatric surgery for obesity, and the terminology changed to prevent ‘indication creep’.

Conclusions in relation to the adolescent patient population Bariatric surgery in severely obese adolescents should only be considered within the

context of an ongoing and coordinated multidisciplinary approach. Surgeons performing bariatric surgery on adolescents should be experienced,

credentialed for bariatric surgery and affiliated with a team experienced in the long-term follow-up and management of the metabolic and psychosocial needs of the adolescent bariatric patient and family. The institution should be one that is either participating in a study of the outcomes of bariatric surgery, or sharing data.

LAGB is the bariatric surgery of choice for adolescent patients because of its relative safety and its potential reversibility.

SG should only be considered investigational and BPD and DS procedures should not be recommended in adolescents.

Adolescent patients should be followed-up on a 4-6 weekly basis post-surgery with long-term follow-up extended beyond 10 years, and ideally for the whole of life.

Other issues for further consideration There should be periodic reviews of the long-term efficacy of emerging surgical

techniques such as SG, and the long-term efficacy and cost effectiveness of gastric banding (including an analysis of reoperation and band adjustment rates).

Bariatric surgery in adolescents is not recommended for:

2 For example: Bariatric surgery is recommended to adolescents only if they fulfil all of the following criteria: 1. minimum age of 15 years (14 years in exceptional circumstances); 2. attainment of Tanner stage 4 or 5 pubertal development; 3. attainment of final or near-final adult height; 4. severe obesity (BMI>40kg/m 2 or >35kg/m2 with severe obesity-associated complications); 5. persistence of the level of obesity despite involvement in a minimum of 6 months supervised multidisciplinary therapy involving lifestyle modification and pharmacotherapy; 6. both the adolescent and the family engaged to participate in the ongoing treatment, lifestyle change and review following surgery; and 7. The adolescent is able to provide informed consent for the surgery (Baur et al 2010).

xiii



children under the age of 14 years; pregnant or breastfeeding adolescents; patients with significant cognitive disabilities; patients with untreated or untreatable psychiatric or psychological disorder; or patients with Prader-Willi syndrome or other similar hyperphagic conditions.

Adults over 60 years of age should be assessed on a case by case basis and the objective of surgery should be to improve their quality of life.

Bariatric surgery may be considered as a first-line option (instead of lifestyle interventions or pharmacotherapy) for adults with a BMI over 50 kg/m2 in whom surgical intervention is considered appropriate.

Morbidity and mortality rates are increased in patients with a pre-operative BMI >65 kg/m2 undergoing BPD-DS, in these patients, staged bariatric surgical procedures may be an option.

Ideally, bariatric surgery should be performed by a surgeon who has substantial experience, performs bariatric surgeries frequently (50-100 cases per year), operating in properly equipped, high volume weight loss centres (100 cases per year) with integrated and multidisciplinary treatment, as there is a steep learning curve associated with bariatric surgery and this reduces operative mortality.

AGB, RYGB and BPD are all effective in treating clinically severe obesity, but differ in the degree of weight loss and range of complications. The choice of procedure should be tailored to the individual’s situation weighing necessary outcomes versus tolerance or risk and lifestyle change.

Gastric bypass and BPD (with or without DS) should be reserved for heavier patients because of the potential for metabolic complications related to malabsorption.

To monitor safety and ensure good practice around obesity surgery, that consideration be given to the establishment of registries.

Patient selection criteria should be updated regularly to reflect new technologies and ongoing refinement in surgical techniques.

Recommendations should be developed on anaesthesia and intensive care for obese subjects.

Medical imaging, lifting and transport equipment as well as beds should be adapted for patients whose corpulence is incompatible with standard models.

Further consideration should be given to the merits of allowing suitable trained and qualified staff, such as practice nurses, nurse practitioners, physician assistants and residents, to adjust gastric bands under the supervision of a medical practitioner. The medical practitioner under whose supervision the adjustment is provided would retain responsibility for the health, safety and clinical outcomes of the patient.


xiv



1 Introduction to Quality Framework Demonstration Reviews

In the 2009-10 Budget, the Australian Government agreed to put in place a new evidence-based framework for managing the Medicare Benefits Schedule (MBS) into the future through the measure: ‘Medicare Benefits Schedule – A quality framework for reviewing services’ (MBS Quality Framework). A key component of the Framework is implementing a systematic approach to reviewing existing MBS items to ensure they reflect contemporary evidence, offer improved health outcomes for patients and represent value for money. A primary objective is identifying and evaluating current MBS services that present potential safety and quality issues, and identifying opportunities to encourage more appropriate clinical practice.

Deloitte Access Economics, as part of its contract with the DoHA, has undertaken a review of the evidence relating to surgical intervention for the treatment of obesity, with the aim of aligning the existing MBS obesity items with current evidence and clinical best practice.

The objective of this review is to carry out an evidence-based assessment of the MBS surgical obesity items to inform ongoing Government decisions in relation to MBS funding for these services3.

1.1 Principles to guide MBS reviewsMBS Quality Framework demonstration reviews are underpinned by the following key principles: reviews have a primary focus on improving health outcomes and the financial

sustainability of the MBS, through consideration of areas potentially representing: patient safety risk; limited health benefit; and/or inappropriate use (under or over use).

reviews are evidence-based, fit-for-purpose and consider all relevant data sources; reviews are conducted in consultation with key stakeholders including, but not limited

to, the medical profession and consumers; review topics are made public, with identified opportunities for public submission and

outcomes of reviews published;

3 This review focuses on surgical interventions for the treatment of obesity only. Evidence-based assessment of non-surgical treatment of obesity including diet and exercise, behavioural treatments and pharmacotherapy are beyond the scope of this review. However, non- surgical interventions for the management of obesity are commonly considered conventional therapy and thus are included as a comparator in the economic evaluation.

1



reviews are independent of Government financing decisions and may result in recommendations representing costs or savings to the MBS, as appropriate, based on the evidence;

secondary investment strategies to facilitate evidence-based changes in clinical practice are considered; and

review activity represents efficient use of Government resources.

1.2 Purpose of this documentThis report is intended to outline the methodology and findings of the review of the MBS surgical obesity items 14215, 30511, 30512, 30514 and 30518. To this original list, item 31441 has subsequently been added.

This report follows from the protocol which aimed to: define the relevant clinical questions that are the focus of the review; clarify the role of existing MBS service/items in current clinical practice; clarify the mechanisms for identifying evidence and provide an opportunity for

discussion of clinical and methodological issues; clarify timelines associated with this project; and clarify roles and responsibilities of key stakeholders.

The remainder of the report is structured as follows – the initial chapters mirroring those in the protocol. Chapter 2 provides background on the MBS surgical obesity items. Chapter 3 outlines the clinical/research questions. Chapter 4 describes the roles, responsibilities and engagement of key stakeholders in

the review. Chapter 5 summarises the review methodology and timeframe. Chapter 6 presents the data analysis. Chapter 7 presents the findings from the clinical literature review. Chapter 8 presents the findings from the economic literature review. Chapter 9 provides conclusions from the review. The report concludes with references and appendices.

2



2 Background on the MBS surgical items for obesity

2.1 Description of the current servicesThis section provides an outline of: the MBS services being reviewed, including item numbers and a description of the

service; the conditions/diseases the services are used to monitor/treat/diagnose; service delivery setting; type of service provider or providers; and year of adoption into the Australian health system where available (e.g. approval by

relevant bodies, year of introduction to Medicare, any major amendments to items).

2.1.1 Service item descriptors

The MBS services being reviewed are summarised in Table 1.1.

Table 1.1: MBS services being reviewed

MBS item Item description14215 LONG-TERM IMPLANTED RESERVOIR associated with the adjustable gastric band,

accessing of to add or remove fluid (for adding or removing fluid via the implanted reservoir to adjust the tightness of the gastric band)

30511 MORBID OBESITY, gastric reduction or gastroplasty for, by any method (Anaes.) (Assist.)

30512 MORBID OBESITY, gastric bypass for, by any method including anastomosis (Anaes.) (Assist.)

30514 MORBID OBESITY, surgical reversal, by any method, of procedure to which item 30511 or 30512 applies (Anaes.) (Assist.) MBS Explanatory Note T.8.19*

30518 PARTIAL GASTRECTOMY (Anaes.) (Assist.)31441 LONG-TERM IMPLANTED RESERVOIR associated with the adjustable gastric band,

repair, revision or replacement of (Anaes.)*MBS Explanatory Note T.8.19: Revision of gastric procedure, for example to correct misplacement of the gastric band or other adverse effects of the initial surgery, involves complete reversal of the initial surgery immediately followed by another reduction, gastroplasty or bypass procedure. For revision item 30514 can be claimed with either item 30511 or 30512, whichever is relevant. For cases where division of adhesions exceeds 45 minutes either item 30378 (laparotomy) or item 30393 (laparoscopy) can also be claimed. Source: MBS.

3



2.1.2 The indication – clinically severe obesity

Obesity is a disease in which fat has accumulated to the point where health is impaired.

Obesity is the accumulation of excess fat in the body, defined here in terms of Body Mass Index (BMI) over 30 for adults and, for children and adolescents aged 2 to 18 years, a set of age-gender specific BMI thresholds.

BMI is the most commonly used measure of obesity and is calculated as the ratio of weight in kilograms to the square of height in metres.

BMI = weight (kg) / [height (m)]2

For adults, weight classifications based on BMI are as follows, used in the Australian Bureau of Statistics (ABS) National Health Survey. Underweight <18.5 Normal range 18.5 to < 20.0 and 20.0 to < 25.0 Overweight 25.0 to < 30.0 Obese ≥ 30.0

The World Health Organization (WHO) categorises obesity into three categories: Class I obesity 30.0 to <35.0; Class II obesity 35.0<40.0; and Class III obesity ≥ 40.0.

Class III obesity is further disaggregated into further categories. Clinically severe obesity (or morbid obesity4) ≥ 40, or between 35 and 40 where there

are other major medical conditions e.g. high blood pressure and diabetes (DoHA, 2003a) Super obese ≥ 45 or 50 (Sturm R, 2007)

These weight classifications are not necessarily suitable for all ethnic groups. For example, at the existing BMI cut-off point for overweight, there is a higher proportion of Asian people who are at high risk of type 2 diabetes and cardiovascular disease (WHO, 2010). Cut-off points vary between different Asian populations: overweight is between BMI of 22.0 and 25.0 while obese varies from BMI 26.0 to 31.0 (WHO, 2010). In addition, the Australian Institute of Health and Welfare (AIHW, 2006) notes that Polynesians would not be considered obese until they reached a higher BMI cut-off.

For children and adolescents (2-18 years), a set of age and gender specific BMI-thresholds have been developed for epidemiological purposes. These thresholds were developed by Cole et al (2000) and are based on data from Brazil, Great Britain, Hong Kong, the Netherlands, Singapore and the United States, aligning with the adult obesity threshold of 30 kg/m2 by age

4 The National Institutes of Health (NIH) in the United States (US) suggests that the term ‘clinically severe obesity’ be used in place of the older terminology ‘morbid obesity’. The current guidelines in Australia and the MBS still use ‘morbid obesity’. In this report, the term ‘clinically severe obesity’ is used from here on, in line with the NIH and CWG recommendation.

4



of 18 years. The thresholds have been accepted as the international reference standard for comparing obesity in children and adolescents for population and clinical research purposes (Table 1.2). In clinical practice, Australia adopts BMI-for-age charts from the US Centres for Disease Control and Prevention (Centers for Disease Control, 2000) to assess weight in relation to age for children and adolescents below the age of 18 years (Chart 1.1 and Chart 1.2). The percentile charts identify children with a BMI ranging between the 85 th and 95th percentile as ‘at risk of overweight’ and those children with BMI values above or equal to the 95 th percentile as obese (Denney-Wilson 2003, DoHA 2003b).

Table 1.2: Age and gender thresholds for overweight and obesity in children and adolescents

Source: Cole et al (2000)

5



Chart 1.1: Weight-for-age percentiles for boys

Source: Centers for Disease Control 2000.

6



Chart 1.2: Weight-for-age percentiles for girls

Source: Centers for Disease Control 2000.

While BMI is by far the most commonly reported measure of obesity, a number of measures which reflect the distribution of fat stored in the body are also popular. Centrally distributed fat has been found to be a better indicator of metabolic risks for a range of diseases (such as cardiovascular disease, type 2 diabetes, and cancer) than total body mass when BMI is below 35 (DoHA, 2003a). Measures of weight distribution are technically measures of a subset of obesity known as abdominal obesity or central adiposity. These measures include the waist-

7



to-hip ratio, and waist circumference (DoHA, 2003b). Cut-offs for waist circumference values associated with an increased risk of metabolic complications are provided in Table 1.3.

Table 1.3: Waist circumference cut-offs correlated to disease risk

Risk of metabolic complications Males FemalesIncreased 94.0-101.9 cm 80.0-87.9 cmSubstantially increased ≥102.0 cm ≥88.0 cmSource: DoHA 2003a

People who are obese have higher rates of mortality and morbidity than those who have a healthy body weight, with a high BMI responsible for 7.5% of the total burden of disease in 2003 (Begg et al, 2007). Being overweight increases the relative risk of a variety of conditions which affect health and quality of life (see Table 1.4). Obesity causes almost one-quarter of type 2 diabetes (23.8%) and osteoarthritis (24.5%), and around one-fifth of cardiovascular disease (21.3%) and colorectal, breast, uterine and kidney cancer (20.5%) (Preventative Health Taskforce, 2009).

Table 1.4: Diseases and conditions associated with obesity

Relative risk (RR) Associated with metabolic consequences

Associated with excess weight

Greatly increased (RR>3) Type 2 diabetes Sleep apnoeaGall bladder disease BreathlessnessHypertension AsthmaDyslipidaemia Social isolation and depressionInsulin resistance Daytime sleepiness and fatigueNon-alcoholic fatty liver disease

Moderately increased (RR 2-3) Coronary heart disease OsteoarthritisStroke Respiratory diseaseGout/hyperuricaemia Hernia

Psychological problemsSlightly increased (RR 1-2) Cancer (breast, endometrial,

colon and others)Varicose veins

Reproductive abnormalities/impaired fertility

Musculoskeletal problems

Polycystic ovaries Bad backSkin complications Stress incontinenceCataract Oedema/cellulitis

Source: DoHA (2003a).

Weight loss can reduce the severity of some of these conditions, as well as improving cholesterol levels, blood pressure and glycaemic control and decreasing the symptoms of osteoarthritis.

8



2.1.3 The service delivery setting, providers and health system adoption

While weight management through diet and exercise is first line therapy, with a role also for pharmacological management, there is evidence that bariatric surgery plays a part in treating people who are clinically severely obese and have not been able to achieve long-term weight loss through more conservative means (Access Economics, 2008).

The current Clinical Practice Guidelines for the Management of Overweight and Obesity in Adults highlight the concerns associated with the epidemic proportion of overweight and obesity throughout Australia. The Guidelines note that bariatric surgery combined with permanent lifestyle changes is the most effect weight-loss treatment and has clear effects on some of the morbidities associated with obesity. Despite bariatric surgery is increasingly recognised as the treatment of choice, there is not widespread availability in the public hospital to meet the demand (DoHA, 2003a).

The current Clinical Practice Guidelines for the Management of Overweight and Obesity in Children and Adolescents recommend bariatric surgery as a management strategy in clinically severe obese adolescents, and recommend that it should only be pursued in tertiary institutions with specialist obesity services, where appropriate assessment, therapy planning and multi-disciplinary support are available (DoHA, 2003b). These are echoed, with more specific detail, in a recent set of recommendations for bariatric surgery in adolescents from the Royal Australasian College of Physicians, the Australia and New Zealand Association of Paediatric Surgery, and the Obesity Surgery Society of Australia and New Zealand (Baur et al 2010).

Increased prevalence of clinically severe obesity (DoHA, 2003a) and community awareness of the effectiveness of surgery in producing weight loss have led to growing demand for bariatric surgery in the public hospital system.

In the community setting, gastric reduction or gastroplasty (item 30511), gastric bypass (item 30512), surgical reversal (item 30514) and partial gastrectomy (item 30518) have been subsidised under the MBS since 1992, while the reservoir items (item 14215 and 31441) were added in 1999.

2.2 ContextThis section provides an outline of; incidence and prevalence of the diseases or conditions for which the services are

provided; MBS usage and expenditure; alternate MBS funded services/comparator services; other potentially impacted services; and any other relevant information.

9



2.2.1 Prevalence of obesity

Obesity is considered one of the greatest public health challenges confronting Australia and many other industrialised countries. Even among developed nations, Australia is one of the most overweight, with over 60% of adults and one in four children overweight or obese, according to the Preventative Health Taskforce (2009).

Access Economics (2008) estimated the prevalence of obesity in Australia based on Australian measured anthropomorphic data from the 2000 AusDiab dataset (International Diabetes Institute, 2001) and from the NSW Schools Physical Activity and Nutrition Survey (SPANS, 2004) study for children (Booth et al, 2006, 2003), together with and self-reported data from the 2006-07 Supplementary Analysis of Nominated Data (SAND) from the general practice study, Bettering the Evaluation and Care of Health.5 Due to a lack of definitive data, obesity rates were conservatively assumed to be 0% for children under the age of four years, increasing with age thereafter to peak in the 55-64 year age group, thereafter falling ( Table1.5).

Table 1.5: Prevalence rates for obesity by age and gender

Age Group Males (%) Females (%) Males (‘000) Females (‘000)

Total (‘000)

0-4 0% 0% 0 0 05-19 7.8% 6.2% 165.4 124.9 290.320-24 11.1% 9.3% 84.7 68.2 152.925-34 19.4% 13.5% 281.8 193.0 474.835-44 19.9% 21.2% 301.5 324.6 626.145-54 23.2% 29.2% 338.6 430.8 769.455-64 28.5% 35.6% 344.9 431.7 776.665-74 22.2% 31.9% 164.4 244.2 408.675+ 14.2% 16.9% 79.6 134.3 213.9Total 16.5% 18.5% 1,760.8 1,951.8 3,712.5

Source: Access Economics (2008).

The table shows that, combining these prevalence rates with demographic data from the ABS, in 2008, 3.71 million Australians (17.5%) were estimated to be obese. Historically, rates of obesity have been increasing (Chart 1.1). By 2025, a total of 4.6 million Australians (18.3% of the population) are projected to be obese on current trends (Access Economics, 2008).

5 Other data sources were investigated but rejected for various reasons (e.g. age), including the National Nutrition Study (1995) and the Child and Adolescent Physical Activity and Nutrition Survey in Western Australia.

10



Chart 1.1: Trends in obesity prevalence for adults, 1980 to 2007

0%

5%

10%

15%

20%

25%

30%Males, measured BMIFemales, measured BMIMales, self-reported BMI (BEACH)Females, self-reported BMI (BEACH)

Source: Access Economics (2008).

A recent study by Walls et al (2010) reveals that there was not a uniform increase in BMI between 1980 and 2000: there was a substantially higher increase in the more severe categories of obesity. Table 1.6 shows rapid growth in the percentage of obese adults in obesity class II and class III in Australia. Both trends in the prevalence of total obesity and shifts in population BMI distribution need to be considered when measuring the burden of obesity in Australia and the effectiveness of obesity prevention and treatment campaigns.

Table 1.6: Percentage of urban Australian adults aged 25-64 years in each obesity category in 1980 and 2000

0%

20%

40%

60%

80%

100%

Obese class I Obese class II Obese class III

Males

1980 2000

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

Obese class I Obese class II Obese class III

Females

1980 2000

Source: Walls et al (2010).

11



2.2.2 MBS expenditure on the review itemsMBS expenditure by item is shown in Table 1.7 with growth rates in the six items averaging well over 30% per annum, more than three times the growth rates in total MBS expenditures. In current dollar terms, expenditure has increased from $6.3 million in 2005 on the six items to $19.3 million in 2009. (Data in Chapter 6 for the review are up to the 2009-10 year.)

Table 1.7: MBS expenditure by item

2005 2006 2007 2008 200914215 ($’000) 2,754 3,872 5,251 7,808 9,78530511 ($’000) 2,686 3,324 4,806 6,990 6,73230512 ($’000) 127 170 143 136 15630514 ($’000) 538 747 1,093 1,403 1,86130518 ($’000) 90 141 213 449 57931441 ($’000) 75 90 106 133 170Total items ($ million) 6.3 8.3 11.6 16.9 19.3Growth (%) 41.4 33.1 39.2 45.7 14.0MBS Total ($ billion) 10.6 11.3 12.4 13.7 15.1Growth in MBS (%) 14.5 6.4 9.9 10.7 10.1Source: MBS online data. Note: Item number 30518 may be claimed for other purpose not related to bariatric surgery.

2.2.3 Alternate and other impacted MBS services

The downstream impacts of obesity (diabetes, cardiovascular disease, cancers and osteoarthritis), affect other MBS items and health system expenditures. The total financial costs of obesity in Australia in 2008 were estimated as $8.3 billion (Access Economics, 2008). Of these costs, the Australian Government bears over one-third (34.3% or $2.8 billion per annum), and state governments 5.1%.

In addition, while performing bariatric surgery it is important for surgeons to find and repair hiatus hernia as this greatly reduces the rate of complications (e.g. reflux) post surgery. Clinical experts suggested that hernia is located in around one quarter to one half of patients undergoing surgery. As such, this review also examines items claimed on the same day as the bariatric surgery with a patient and surgeon match, revealing other linked items.

2.3 Justification for reviewThis section provides an outline of: the review identification mechanism; and key issues raised to date.

2.3.1 The review identification mechanism

The systematic approach to reviewing existing MBS items, in the context of the MBS Quality Framework, has commenced with several demonstration reviews to test the proposed review

12



process. This review will serve as one of these demonstration reviews. Review outcomes will be considered by Government in the second half of 2011.

The demonstration reviews were identified, having regard to current literature, as presenting potential quality and safety concerns or the opportunity to encourage more appropriate clinical use.

The review of the surgical items for the treatment of obesity will inform recommendations aimed at strengthening the evidence-base of the items and their use.

2.3.2 Key issues to date

Currently the existing MBS item descriptors for surgical treatment of obesity do not: define morbid obesity; or place any restrictions on the use of procedures, potentially hindering compliance with

clinical best practice by not distinguishing procedures which are: more safe or efficacious; more cost effective; or only so in a particular target population.

Given that bariatric surgery is a significant procedure with potential complications, the provision of bariatric surgery should be targeted to patients for whom the benefits of surgery clearly outweigh the risks. By strengthening the evidence base for the safety, effectiveness and cost-effectiveness of each of the surgical procedures for the treatment of bariatric surgery, rational, consistent and equitable access to bariatric surgery can be promoted in both the public and private sectors. It is expected that by offering only high quality bariatric surgery programs with multi-disciplinary approach to carefully selected patients, health outcomes can be improved and the financial sustainability of the MBS can be maintained.

13



3 Clinical/research questionsThis section outlines the process for defining the clinical questions for the review.

The PICO criteria are used to develop the clinical questions. The four elements of the PICO criteria are: the target population for the intervention; the intervention being considered; the comparator for the existing MBS service (where relevant); and the clinical outcomes that are most relevant to assess safety and effectiveness.

A key aim of the stakeholder consultations and literature review, described in later chapters, was to better define these clinical questions in terms of the specific populations, interventions, comparators, and outcomes relevant for Australian practice. This chapter therefore presents the clinical research questions in more general terms.

3.1 PopulationThe target population is identified by breaking down larger populations into those relevant for each intervention.

The overall population is Australians obtaining care under the MBS. The most recent Clinical Practice Guidelines for the Management of Overweight and Obesity in Adults (currently under review) indicate bariatric surgery for adults with a BMI greater than 40 or with a BMI greater than 35 and serious medical comorbidities who have instituted but failed adequate non-operative measures for weight loss with integrated components of a dietary regimen, appropriate exercise, and behaviour modification and support (see Figure 1.1) (DoHA 2003a).

In overweight and obese adolescents, bariatric surgery is considered as a non-conventional weight-management strategy. A more recent published position paper (Baur et al 2010) recommends that surgical treatment should only be considered, along with other criteria, in adolescents with severe obesity (BMI greater than 40 or a BMI greater than 35 with severe obesity associated complications) who are over the minimum age of 15 years. All candidates for surgery need to be carefully assessed by a multi-disciplinary team. The stakeholder consultations and literature review detailed in later chapters confirm the target population for each of the interventions.

14



Figure 1.1: Clinical decision pathway for the management of the overweight or obese person

Determine degree of overweight or obesity

- Measure height and weight; calculate BMI (kg/m2)- Measure waist circumference if BMI is <35 kg/m2

Clinical assessment of overweight and obesity

- weight related comorbidities- energy intake and physical activity levels- weight history- background- environment (family, work and social)

If BMI is > 25 kg/m2 or waistcircumstance is above cutoff point Assess readiness to change behaviours

and motivation

Assess and screen for depression, eating and mood disorders

Treat comorbidites and other health risks if present

Devise goals and lifestyle modification program for weight loss including

integrated components of a dietary regimen, appropriate exercise, and behaviour modification and support

Total weight loss goal: e.g. 5-10% of body weight or 0.5-1 kg per week over

6-12 months

Satisfactory progress or goal achieved?

Regular monitoring- assist with weight maintenance and reinforce healthy eating and physical activity advice to prevent weight regain- address other risk factors

Pharmacotherapy as adjunct to lifestyle modification

- orlistatBMI >30 or 27 with other cardiovascular risk factors- phentermineBMI >30 or 27 with other cardiovascular risk factors; short-term use

Bariatric Surgerye.g. BMI ≥ 35 + risk factorsor BMI ≥ 40

MBS items:- 30511- 30512- 30518

Yes No No

Source: DoHA 2003a.

3.2 InterventionThe interventions considered are surgical procedures for obesity that are currently practised in Australia with the MBS item numbers under review (listed in Table 1.1). Figure 1.2 illustrates the possible patient pathways through the current MBS item numbers.

15



Figure 1.2: Flowchart of patient pathway through MBS items under review

Bariatric Surgery

30511 - gastric reduction or gastroplasty by any method

30512 - gastric bypass by any method

30518 - partial gastrectomy

Laparoscopic adjustable gastric banding (LAGB)

Vertical banded gastroplasty

(VBG)

30514 - surgical reversal, by any method, of procedure to which item 30511 or 30512 applies

14215 - adding or removing fluid via the implanted reservoir

to adjust the tightness of the

gastric band

31441 - repair,revision or

replacement of implanted reservoir

associated with adjustable gastric

band

Roux-en-Ygastric bypass (RYGB)

Laparoscopic gastric sleeve gastrectomy (SG)

Biliopancreaticdiversion with or

without duodenal switch (BPD with or

without DS)

Source: CWG. Dotted line indicates ‘if required’. Note: Roux-en-Y gastric bypass (RYGB) includes open and laparoscopic RYGB with or without duodenal switch.

Some of these MBS items include a range of different procedures, for example: item 30511 includes LAGB, VBG and SG; and item 30512 includes RYGB and BPD-DS.

On the other hand, for some procedures, various item numbers may be used. For example given the ambiguity of some of the MBS item descriptors SG is likely to have been claimed against item 30511 (as above) or item 30518. The complexity of the MBS data do not permit breakdown of item numbers into specific types of surgical procedure, nor enable a reclassification of item numbers. A key aim of the stakeholder consultations and literature review was to determine the extent to which each of these procedures is performed in Australia and which MBS item numbers they are likely to have been claimed against.

3.3 ComparatorThe safety and efficacy of bariatric procedures currently used in Australia was evaluated against one another using the following comparators:1. non-surgical intervention; 2. LAGB; and3. other relevant comparators as identified in the literature review and data analysis.

16



Non-surgical treatment comprises various combinations of behaviour modification (including exercise), pharmacotherapy and/or low-energy or very low-energy diets and in some instances, no treatment. The comparator enables interventions to be compared with each other (including LAGB) in terms of their absolute safety and efficacy.

Surgical procedures in current use would also be compared with LAGB (an intervention as well as a comparator) as this is by far the most widely used technique in Australian, making up the large majority of all procedures (DoHA, 2003a). This comparator enables other interventions to be compared with the most common current practice regarding the target population for whom bariatric surgery is indicated.

Other comparators in the literature and analysis were also reviewed for relevance for the economic evaluation (see section 5.4).

3.4 OutcomesSafety of the various bariatric procedures would be assessed in terms of procedure-related complications and adverse events. Procedural specific risks associated with bariatric surgery were evaluated based on: rate of converting laparoscopic to open procedure; rate of reoperation; post-operative length of hospital stay; short-and long-term procedural specific and non-specific morbidity associated with

surgery to be identified in the literature review; mortality (short term e.g. 90 days mortality, and long term e.g. 10 years mortality); and any other adverse events identified during the literature review and stakeholder

consultation process.

Procedural specific benefits associated with bariatric surgery were evaluated based on: weight loss, measured as percentage of starting weight, an absolute weight loss, or

‘excess weight’ (relative to a normal BMI); maintenance of weight loss (durations used to define clinically relevant outcomes are

discussed later in Chapter 7); quality of life; improvement in and resolution of obesity-related comorbidities (e.g. diabetes mellitus,

hypertension and hyperlipidaemia); and any other relevant efficacy outcomes indentified during the literature review and

stakeholder consultation process.

3.5 Research questionsThe research questions for this review were determined using the PICO criteria (section 3.1 to section 3.4). Listed below are the generalised research questions for this review. What is the safety of LAGB, VBG, SG and RYGB compared to non-surgical treatment of

obesity? What is the safety of VBG, SG and RYGB compared to LAGB in the treatment of obesity?

17



What is the effectiveness of LAGB, VBG, SG and RYGB compared to non-surgical treatment of obesity?

What is the effectiveness of VBG, SG and RYGB compared to LAGB in the treatment of obesity?

18



4 Key stakeholdersThis chapter identifies all key stakeholders involved in the review, including their roles, responsibilities and engagement.

4.1 Clinical Working Group

A Clinical Working Group (CWG) was established for the duration of the review comprising: Professor Joe Proietto, (nominated by Royal Australasian College of Physicians); Associate Professor John Dixon, NHMRC Senior Research Fellow, Obesity Research

Physician (nominated by Obesity Surgery Society of Australia and New Zealand); Dr Ken Loi, practising bariatric surgeon (nominated by Obesity Surgery Society of

Australia and New Zealand); Dr Anthony Moore, Clinical Advisor, DoHA; and Mr Craig Rayner, Director, Medical Benefits Reviews Task Group, DoHA (chair).

In addition, Professor Louise Baur is a clinical expert on the Deloitte Access Economics team. The CWG held its first meeting on 26 August 2010.

The CWG’s role is to help ensure that the review reflects an understanding of current Australian clinical practice and draws valid conclusions from the available evidence.

4.2 Clinical craft groups and othersClinical craft groups, representing those that provide the MBS services under the demonstration review, are key stakeholders. The draft protocol and draft review report were uploaded to the DoHA website and as such there are no exclusions from the consultation process. DoHA identified a number of organisations, including clinical craft groups, who were notified in writing of the review and asked to comment on the draft protocol and again on the draft review report. Stakeholder details are withheld due to confidentiality.

Deloitte Access Economics incorporated comments on the draft protocol in the final protocol, and similarly with the draft report.

4.3 Consumers and the general publicConsumers and the general public (which may include individual service providers, device manufacturers and sponsors of medical technologies) were given two opportunities to comment on elements of the review – the draft protocol and the draft review report. Twelve submissions were received when the draft protocol was released in 2010.

Key consumer and public stakeholders were identified by DoHA and notified in writing of the upload of the draft protocol and the draft review report, to facilitate an inclusive public comments process.

19



In addition, the CHF was contacted once the protocol was uploaded to the DoHA website, and a meeting to discuss the draft protocol was held during September 2010. The CHF was similarly notified of the upload of the draft review report with another meeting to discuss that milestone. DoHA staff were present at these meetings.

4.4 ConsultantsDeloitte Access Economics was responsible for drafting the review protocol and identifying, analysing and synthesising the evidence related to surgical obesity items through the agreed methodology. Deloitte Access Economics provided this review report at the completion of the project to help inform the Government whether the MBS items require amendment to support evidence-based clinical practice.

Deloitte Access Economics has the required technical ability to provide the services. The company is one of Australia’s leading and best known economic consultancies, highly regarded for its professional and high quality modelling and assessments and for commitment to sound policy analysis and advice. Deloitte Access Economics is a valued panel provider of health economics services to numerous Government departments, with many high profile projects (e.g. the national cancer screening program economic evaluations and aged care modelling for DoHA).

Deloitte Access Economics has demonstrated experience providing services similar to this review, in the area of obesity, analysis of surgical services, health service delivery reviews, MBS item and structure reviews, literature review and stakeholder consultation. Deloitte Access Economics has no conflict of interest in undertaking the review. Details of the operational personnel conducting the review were provided in the protocol.

4.5 The Department of Health and AgeingDoHA contracted Deloitte Access Economics to undertake this review and was responsible for the management of the contract. DoHA was also responsible for ensuring that the draft protocol and draft review report were made available online for public comment.

Following the finalisation of this review report, DoHA will be responsible for providing advice to the Minister for Health and Ageing on new or amended MBS items for the surgical treatment of obesity. This advice will be informed by the review report but will also draw on other information such as the Parliamentary Inquiry into Obesity in Australia and the review by National Health and Medical Research Council (NHMRC) on the Clinical Practice Guidelines on the Management of Overweight and Obesity, currently underway.

20



5 Review methodologyThis chapter outlines the agreed methodology for reviewing the MBS items against the clinical questions, further to the consultation process described in the previous chapter. It includes more detail on the methods for the literature review, MBS data analysis and economic evaluation, with the findings from these processes detailed in subsequent chapters. The review was undertaken over the timeframe July 2010 to June 2011.

5.1 Literature reviewThis section outlines databases that were searched, search terms used, supplementary search strategies and how evidence was classified. The aims of the literature review were: to identify contemporary clinical guidance on the use of bariatric surgery for obesity in

Australia and overseas; to obtain efficacy and safety data for bariatric surgeries for obesity commonly

performed in Australia; and to address the research questions in chapter 3 of this protocol using this information.

The databases searched for this review included EMBASE.com (which combined the Embase and Medline databases) and the Cochrane library. The search strategy, including search terms and limits, is presented in Appendix A.

Only studies meeting the NHMRC’s highest levels of evidence were included in the review, including meta-analyses and systematic reviews of randomised controlled trials (RCTs) (Grade I evidence) and any RCTs published since these reviews (Grade II evidence). Any recent RCTs were identified through the literature search and through discussions with the CWG members and professional bodies. This strategy ensured all RCTs were included either within the Grade I studies or more recent Grade II studies.

Relevance of the literature identified in the search was assessed according to the following criteria: study type: meta-analysis, systematic review, or evidence-based clinical guidelines; interventions: comparative analyses of bariatric surgeries that are commonly performed

in Australia; and outcomes: identifiable and comparative efficacy/safety data, and/or evidence-based

clinical practice recommendations.

The review team identified and excluded or caveated any studies that had been outdated or superseded by others identified in the search. For example, only the most recent versions of clinical guidelines and meta-analyses were included for review.

The clinical guidelines reviewed were restricted to Canada, New Zealand, the UK and the US (in so far as relevant guidelines are available). In addition, English language versions of guidelines from other European countries were reviewed. Guidelines were considered relevant where specifically providing recommendations for bariatric surgery for obesity. Greater priority was

21



given to guidelines based on evidence standards equivalent to NHMRC Grades I and II. The clinical guidelines review informed whether the MBS items under review require amendment or removal. Particular focus was on the recommended indications/restrictions for the use of surgery, views on surgery in adolescents, and minimum recommended thresholds for BMI, age and other patient characteristics when considering surgery.

The review of meta-analyses and systematic reviews informed the comparative safety and efficacy outcomes for different bariatric surgery procedures performed in Australia. These data enabled an assessment of which procedures provide the best outcomes for the Australian population with obesity.

The classification of evidence was relatively narrow, since the review only included the highest levels of clinical evidence. Grey literature was not included in the literature review except for specific papers suggested by the CWG for inclusion. Meta-analyses were given a higher classification than systematic reviews because they increase the precision of overall results. The classification of clinical guidelines depended on: (a) the levels of evidence used to support their recommendations; and (b) the relevance of the guidelines to the Australian population. Guidelines for Australia and countries with similar populations and health care systems were classified more highly.

5.2 MBS dataThe DoHA facilitated access to necessary data as appropriate. The relative usage and costs of MBS item numbers associated with bariatric surgery were evaluated using data on the following MBS item numbers: 14215 – long-term implanted reservoir associated with adjustable gastric band,

accessing of to add or remove fluid; 30511 – morbid obesity, gastric reduction or gastroplasty for, by any method (Anaes.)

(Assist.); 30512 – morbid obesity, gastric bypass for, by any method including anastomosis

(Anaes.) (Assist.); 30514 – morbid obesity, surgical reversal, by any method, of procedure to which item

30511 or 30512 applies (Anaes.) (Assist.); 30518 – partial gastrectomy; and 31441 – long term implanted reservoir associated with the adjustable gastric band,

repair, revision or replacement of (Anaes.).

Data were obtained for each item by subgroup, defined by gender and ten-year age group (e.g. males aged 25-34 years, females aged 25-34 years) and aggregations of these groups (e.g. males and females aged 25-34 years, males of all ages). Data were also aggregated for subgroups where MBS item numbers were too low to be reported due to confidentiality. This approach is more accurate than inputting MBS item numbers where data are reported as ‘fewer than n services’. Data were obtained for the last five years in order to evaluate trends in relative usage of different surgical procedures (i.e. five 12-month periods ranging from July 2005 – June 2006 to July 2009 – June 2010).

The CWG indicated a number of procedures that may be performed alongside these surgeries including reduction of hiatus hernia and anti-reflux operation by fundoplasty. Therefore, data

22



were also obtained for the ten most frequent MBS item numbers recorded under the same day claim as the item numbers above to evaluate additional usage/costs of MBS item numbers associated with surgery for obesity.

The types of analyses performed using these MBS data included subgroup analyses and trend analyses.

For the subgroup analyses, historical usage and expenditure for each MBS item numbers were established for each gender/age subgroup and aggregates of these subgroups. These data were used to inform the trend analyses described below. The subgroup analyses were also used to observe any increasing or decreasing trends in the use of MBS item numbers that are comparators for each other (e.g. item 30511 and 30512).

For the trend analyses, projections of future use/expenditure for these MBS item numbers were performed using the historical data trends, ABS projections of demographic changes, and current obesity prevalence rates by age and gender. The time horizon for these projections was five years. Projections incorporated reasonable assumptions about how any modifications to MBS item numbers might alter utilisation patterns (e.g. by restricting utilisation to particular age or risk groups).

Several issues emerged in identifying specific item numbers from the MBS data. Some MBS item numbers include procedures for different reasons. For example, partial

gastrectomy (item 30518) may be performed for several indications, such as stomach cancer rather than obesity. This issue was addressed using expert opinion or any available MBS-linked data that identified primary diagnoses/indications for surgery which could be accessed within the timeframe of this review.

Some MBS item numbers do not distinguish between types of surgery. For example, item 30512 does not distinguish between laparoscopic or open bypass, and item 30511 does not distinguish between AGB, gastroplasty or SG. To address this issue, supplementary data and expert opinion were used to disaggregate these usage figures based on current trends in surgery for obesity.

Some types of surgery may be recorded under different MBS item numbers; for example, the MBS item number for SG resection is unclear. Again expert opinion or supplementary data were used to determine the usage in Australia of newer procedures where the MBS item number allocation was unclear.

5.3 Economic evaluationIn addition to literature review of safety and efficacy, there was also a review of literature in relation to economic evaluation of services found to be safe and effective. While the recommendations of this review are independent of funding decisions, one key aim of this review is to determine whether the MBS items should be better defined in terms of patients and/or procedures. These decisions may be driven by consideration of cost effectiveness.

Within the timeframe and scope of this review, evidence of cost effectiveness was restricted to previous studies where procedures were found to be safe, effective, and commonly used in Australia. The literature search was performed similarly to that for safety and efficacy – using a variant of the search terms in Appendix A – to identify any published economic evaluations conducted in Australia or similar countries (i.e. countries for which clinical practice guidelines

23



have been examined – see section 5.1). Only economic studies published from the year 2000 onwards were included.

Economic evaluations included those comparing each surgical procedure with the most common procedure (LAGB) and/or other of no surgical intervention. The latter comparator ensured all procedures could be evaluated on their own merits and ranked by cost effectiveness.

The study inclusion criteria were full economic evaluations (rather than systematic reviews, meta-analyses, and clinical guidelines). A full economic evaluation was defined as a study that assessed both the incremental costs and incremental benefits of one or more types of bariatric surgery, and synthesised these to estimate an incremental cost effectiveness ratio (e.g. cost per life year gained).

Although there is a preference for evaluating final health outcomes using quality-adjusted life years (QALYs), the outcome metric(s) reflected the available literature, and thus could include other metrics also, such as disability-adjusted life years (DALYs) or natural health units such as BMI, deaths, or obesity cases.

More preference was also given to economic evaluations consistent with Medical Service Advisory Committee (MSAC) guidelines, such as: Australian studies (i.e. relevance of costs and outcomes to the national context); costs from the societal perspective, as defined by MSAC (this perspective is essentially

the health care payer i.e. excluding deadweight losses and in most cases productivity changes when interpreting the published evidence);

a lifetime horizon (to evaluate long term cost effectiveness); and accounting of cost offsets due to a reduction in other diseases and conditions associated

with obesity (e.g. diabetes, osteoarthritis, bowel cancer, breast cancer, coronary heart disease and stroke), where possible.

Results of economic evaluations are ideally expressed as the incremental cost effectiveness ratio (ICER), reported as the cost per QALY gained (or other common and suitable health outcome metric). ICERs for each surgical procedure relative to its comparator were assessed as either: ‘dominant’ (saving costs and improving health outcomes); highly cost effective (based on benchmarks such as the World Health Organization

(WHO) benchmark of costing less than gross domestic product (GDP) per capita per DALY averted);

cost effective (e.g. one to three times GDP per capita per DALY averted under the WHO benchmarks);

not cost effective (e.g. more than one to three times gross domestic product (GDP) per DALY averted under the WHO benchmarks); or

dominated (higher costs and worse health outcomes).

The economic analysis provides a strong evidential basis for any changes recommended from the review.

24

http://www.health.gov.au/internet/msac/publishing.nsf/Content/D81BE529B98B3DB6CA2575AD0082FD1B/$File/Economics%20Glines%20-%20FINAL%20at%20Aug%202008%20-%20endorsed%20MSAC%20ESC%20June%202009.pdf



6 Data analysisOver the past five years (2005-06 to 2009-10), there has been a rapid increase in the number of MBS items claimed in association with the surgical treatment of obesity, from around 55,000 services in 2005-06 to around 147,000 services in 2009-10, representing growth of 166%. Data on primary bariatric6 and maintenance surgeries are analysed separately throughout this chapter.

Growth in primary bariatric procedures was strong from 2006-07 to 2008-09, averaging around 33% before declining to a negative growth of 11% in 2009-10. This is primarily due to a decline in the number of gastric reductions or gastroplasties (item 30511) claimed during 2009-10 relative to 2008-09. Analysis of the proportion of primary bariatric services claimed relative to the Australian population found a high number of bariatric surgical procedures performed per 100,000 Australians from 2007-08 onwards. At this point in time, there are insufficient data to determine whether the increase in per capita bariatric surgeries performed under the MBS results from either an increase in obesity prevalence, an increase in awareness of bariatric surgery, both, or other reasons. To what extent bariatric surgery is being offered to patients who do not have clinically severe obesity or who are in a lower BMI bracket is not clearly known but, with adherence to strict clinical practice guidelines, ‘indication creep’ is less likely to occur. Chart 1.2 depicts past trends in the total number of services claimed for bariatric surgery and Table 1.8 summarises past annual growth in primary bariatric procedures, and the rate of utilisation relative to the general Australian population.

6 Primary bariatric procedures are defined here as the procedures that a patient undergoes for the treatment of clinically severe obesity (i.e. Item 30511, Item 30518 and Item 30512). In the case of AGB, although it may be said that it is the subsequent adjustments to the band that induce weight loss, not the placement of the band per se, for the purpose of tracking service utilisation, the initial procedure (i.e. the placement of the AGB) is classified as ‘treatment of obesity’ while band adjustments (Item 14215) and repair, revision or replacement of implanted reservoir associated with AGB (Item 31441) are classified as ‘maintenance therapy’.

25



Chart 1.2: Total number of MBS services claimed for the treatment of obesity by year

0

20

40

60

80

100

120

140

160

2005/06 2006/07 2007/08 2008/09 2009/10

Num

ber o

f ser

vice

s (th

ousa

nds)

Source: DoHA (data) Deloitte Access Economics calculations.

Table 1.8: Growth in the number of MBS items claimed under primary bariatric services* in the past five years

2005-06 2006-07 2007-08 2008-09 2009-10Number of primary bariatric services claimed 6,557 8,094 12,079 15,265 13,600 % growth 23% 49% 26% -11%Number of services per 100,000 people 32 39 57 70 62 % growth 22% 47% 24% -12%Source: DoHA (data) Deloitte Access Economics calculations. * See footnote Error: Reference source not found page 25 for the definition of ‘primary’.

6.1 Breakdown of MBS claims for the treatment of obesity

Over the past five years (2005-06 to 2009-10), a total of 506,264 services were claimed under MBS surgical obesity items. Of these, over 97% were claimed under two items – 14215 (gastric band adjustments) with 87.0% of the total and item 30511 (which includes LAGB, VBG and SG procedures) with 10.1%. A breakdown of the total number of services claimed by MBS item is shown in Chart 1.1.

26



Chart 1.1: Total number of services claimed in the last five years by MBS items for the treatment of obesity

0

50

100

150

200

250

300

350

400

450

14215 30511 30512 30514 30518 31441

Num

ber o

f ser

vice

s (th

ousa

nds)


A breakdown of MBS data by MBS items claimed over the past five years (2005-06 to 2009-10) shows a different trend for each of the three primary MBS items considered responsible for operatively induced weight loss (items 30511, 30518 and 30512) (Chart 1.2). The number of services claimed and annual growth rates for the respective item numbers are shown in Table1.9.

Chart 1.2: Total number of MBS surgical obesity procedures claimed, by year

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

2005/06 2006/07 2007/08 2008/09 2009/10

Item 30511

0

200

400

600

800

1,000

1,200

2005/06 2006/07 2007/08 2008/09 2009/10

Item 30518

0

50

100

150

200

250

300

2005/06 2006/07 2007/08 2008/09 2009/10

Item 30512


27



Table 1.9: Total number of MBS surgical treatment of obesity claims, by year

MBS item number 2005-06 2006-07 2007-08 2008-09 2009-1030511

Number of MBS items claimed 6,080 7,531 11,350 14,139 12,221% growth 23.9% 50.7% 24.6% -13.6%

30518Number of MBS items claimed 238 322 515 895 1,097% growth 35.3% 59.9% 73.8% 22.6%

30512Number of MBS items claimed 239 241 214 231 282% growth 0.8% -11.2% 7.9% 22.1%


A positive growth in the number of gastric reduction or gastroplasty procedures (item 30511) was observed from 2005-06 to 2008-09. Growth was most rapid in 2007-08, at around 51%, declined to around 25% in 2008-09, and slumped into a negative growth in 2009-10 (i.e. a reduced number of services claimed in 2009-10 compared to the previous year). MBS item 30511 includes a range of different procedures: LAGB, VBG and SG. As the item number does not distinguish between the types of procedures, analysis of trends by specific type of surgical procedure is not possible. This decline in the overall number of item 30511 procedures claimed may be due to a decrease in the number of VBGs performed and/or the number of SGs coded under item 30511 rather than a decrease in number of LAGBs performed, because trends in services relating to the maintenance of an AGB, specifically item 14215 (gastric band adjustments) and item 31441 (repair, revision and replacement of implanted reservoir associated with adjustable gastric band) have remained strong (see Chart1.1 and Table 1.10). However, it may also be the case, as suggested by stakeholders, that LAGB is falling out of favour, and the final year growth in maintenance procedures simply reflects the accumulated number of gastric bands in place from previous historical growth.

Every year during the past five year period, there was positive growth in the number of services claimed under item 30518. This suggests either an actual increase in the number of SGs performed and/or an increase in the coding of SG to this item by surgeons. Item 30512 has grown since a decline in number of services claimed in 2007-08. Similar to item 30511, this MBS item includes a range of different procedures (i.e. RYGB, BPD-DS and jejunoileal bypass); as a consequence, further analysis of trend by specific type of surgical procedure is not possible.

Growth in services claimed under item 14215 was generally higher than growth in services claimed under item 31441 (Table 1.10) suggesting, possibly, an underlying increase in the number of people with clinically severe obesity who have AGB, and/or an increase in the number of adjustments per AGB procedure performed over the past five years, and/or an enhancement in the procedure such that fewer repairs/revisions/replacements are required per capita over time.

28



Chart 1.1: Total number of services claimed relating to maintenance of AGB, by year

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2005/06 2006/07 2007/08 2008/09 2009/10

Item 14215

0

200

400

600

800

1,000

1,200

2005/06 2006/07 2007/08 2008/09 2009/10

Item 31441


Table 1.10: Total number of services claimed relating to maintenance of AGB, by year

Item number 2005-06 2006-07 2007-08 2008-09 2009-1014215

Number of MBS services claimed 47,128 61,222 84,799 117,322 129,837% growth 29.9% 38.5% 38.4% 10.7%

31441Number of MBS services claimed 579 643 771 968 1,089% growth 11.1% 19.9% 25.6% 12.5%


The number of MBS surgical reversals of bariatric surgery performed also increased steadily (with most of the surgical reversals related to a previous LAGB surgery performed7) – from around 700 procedures in 2005-06 to almost 2,000 procedures in 2009-10 (Chart 1.1). However, relative to the number of MBS items 30511 and item 30512 performed, the proportion of surgical reversal procedures has remained fairly constant, at around 11% from 2005-06 to 2008-09, with a spike (16%) observed in 2009-10 (Table 1.11) – noting there is of course a timing lag between the original procedure and any reversal (so comparing the ratio within the same year may not be particularly meaningful).

7 The majority of the surgical reversals are related to LAGB, although in some instances the item may be claimed for revising other historical types of obesity surgery performed long time ago (pers. comm., Dr Ken Loi, 4 March 2011).

29



Chart 1.1: Total number of MBS surgical reversal procedures claimed, by year

0

500

1,000

1,500

2,000

2005/06 2006/07 2007/08 2008/09 2009/10


Table 1.11: Total number of MBS surgical reversal procedures claimed, by year

2005-06 2006-07 2007-08 2008-09 2009-10Number of MBS services claimed under item 30514 721 868 1,189 1,557 1,976 Number of MBS services claimed under MBS items 30511 or 30512 6,319 7,772 11,564 14,370 12,503 % of surgical reversal 11.4% 11.2% 10.3% 10.8% 15.8%Source: DoHA (data) Deloitte Access Economics calculations.

6.2 Breakdown of items by broad age groupAn analysis of the proportion of MBS primary8 surgical procedure items claimed in the past five years by broad age group shows gastric reduction or gastroplasty (item 30511) to be the most common method of surgical treatment compared to partial gastrectomy (item 30518) and gastric bypass (item 30512) among people with clinically severe obesity across all age categories: children (aged 0-14), adolescents (aged 15-19), young adults (aged 20-44), older adults (aged 45-64) and the elderly (aged 65 and above), as shown in Chart 1.1. In people aged 65 years and above, although gastric reduction or gastroplasty (item 30511) remains the most common method of bariatric surgery employed (66%), the proportion of item 30511 claimed

8 See footnote Error: Reference source not found on page 25Error: Reference source not found for the definition of ‘primary’.

30



relative to other surgical methods is much lower than the high share of MBS claims made in the other age groups (which averaged around 95%). Instead, there was a higher proportion of services claimed under item 30518 (partial gastrectomy), which may be performed for other reasons, such as for the treatment of stomach cancer – see MBS-linked data discussed below (and noting that operative risk increases in older ages so starts to offset the risks of obesity itself). On the other end of the age spectrum, a total of 54 claims during the five year period for the surgical treatment of obesity were for children, of which 51 claims were attributed to either a gastric reduction or a gastroplasty procedure, while none were attributed to gastric bypass.

Chart 1.1: Proportion of bariatric surgery items claimed by item number and broad age group

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0-14 15-19 20-44 45-64 65+

30511 30518 30512


6.3 Breakdown of services claimed by ten-year age groups

Individual item utilisation data by age 0-24 and ten-year age groups for people aged 25 and above shows the use of MBS items for the surgical treatment of obesity to follow a bell shaped curve. From 2005-06 to 2009-10, the majority of MBS items for surgical treatment of obesity were claimed in patients aged 35-64, tapering off substantially in the younger and the older ages. Of the three surgical MBS items for the treatment of clinically severe obesity, item 30518 (partial gastrectomy) has the longest ‘tail’ into the older ages without a sudden drop in number of services at age 65. This pattern is likely to be confounded by the increasing prevalence of people with stomach cancer in the older population who have partial

31



gastrectomy (pers. comm., Dr Ken Loi, 10 November 2010). Chart 1.2 shows the number of MBS services claimed by item number, broken down by age.

Chart 1.2: Service utilisation by MBS Item number and ten-year age groups

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

0-24 25-34 35-44 45-54 55-64 65-74 75-84 85+

Item 14215

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

0-24 25-34 35-44 45-54 55-64 65-74 75-84 85+

Item 30511

0

50

100

150

200

250

300

350

400

450

0-24 25-34 35-44 45-54 55-64 65-74 75-84 85+

Item 30512

0

500

1,000

1,500

2,000

2,500

0-24 25-34 35-44 45-54 55-64 65-74 75-84 85+

Item 30514

0

100

200

300

400

500

600

700

800

0-24 25-34 35-44 45-54 55-64 65-74 75-84 85+

Item 30518

0

200

400

600

800

1,000

1,200

1,400

0-24 25-34 35-44 45-54 55-64 65-74 75-84 85+

Item 31441


6.4 Breakdown by genderThe majority of bariatric surgeries were performed on female patients. Of the total number of primary surgical procedures performed for the treatment of obesity during 2005-06 to 2009-10, 77.7% were female. Female patients also make up the majority (around 83%) of patients who had gastric band adjustments (item 14215) and repairs, revisions or replacements (item 31441) of their gastric band. If the gender proportions of those receiving gastric reduction or gastroplasty procedure is representative of those who have LAGBs (i.e. the number of VBGs

32



and SGs is negligible), then the higher proportion of AGB maintenance services performed in females (83%) compared to the proportion of primary bariatric procedures (i.e. LAGBs) (78%) may suggest that females are better at adhering to follow-ups (or live longer post-surgery). Note that there is also a higher proportion of surgical reversals in females – not unexpected given the higher proportion of females who have bariatric surgery (more specifically, item 30511 and 30512). A breakdown of MBS items by gender is detailed in Table 1.12.

Table 1.12: Total numbers of MBS items claimed by gender

30511 30518 30512 14215 31441 30514Male (number of claims) 11,109 1,019 272 74,024 670 956Female (number of claims) 40,212 2,048 935 366,284 3,380 5,355Male (%) 21.6 33.2 22.5 16.8 16.5 15.1Female (%) 78.4 66.8 77.5 83.2 83.5 84.9Source: DoHA (data) Deloitte Access Economics calculations.

Women of reproductive age

Women of reproductive age contribute to a large proportion of patients receiving surgical bariatric services. During 2005-06 to 2009-10, a total of 280,862 surgical services relating to the treatment of obesity have been provided to females aged between 15 and 49 9 . This represents 67.2% and 55.5% of all bariatric procedures performed in women and in all patients, respectively, during the five year period.

There was no significant difference in service utilisation trends between the three types of primary surgical procedures in females of reproductive age compared to females in general, and compared to all people receiving surgical bariatric services. Table 1.13 summarises the distribution of primary bariatric surgery by MBS items claimed in each population subgroup.

Table 1.13: Proportion of people receiving primary surgical procedures

Service utilisation 30511 30518 30512Female aged 15-49 94.2% 3.8% 2.0%Female of all ages 93.1% 4.7% 2.2%All persons 92.3% 5.5% 2.2%Source: DoHA (data) Deloitte Access Economics calculation.

6.5 MBS linked dataThis section provides an analysis of the ten most frequently claimed same day MBS items alongside each of the six MBS items under review between 2005-06 and 2009-10. Although these are self-reported data and a factor of mis-claiming may impact the validity of the data, nevertheless through linked data, some potential conclusions about clinical practice might be drawn. It is understood that the clinical practice of bariatric surgery is complex, the summary presented here is thus somewhat oversimplified, and the onus of accurate claiming is based on professional judgement and other clinical reasons.

9 Women of reproductive age defined as females aged between 15 and 49 (United Nations, 2003).

33



6.5.1 Primary bariatric procedures

Item 30511 – gastric reduction or gastroplasty by any method

Chart 1.1: Top 10 MBS item numbers claimed with item 30511

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

51303 20791 17610 105 22025 30393 30601 22012 31464 17603

Num

ber o

f ser

vice

s

MBS item numbers claimed in combination


Table 1.14: Descriptors of the top 10 MBS item numbers claimed with item 30511

MBS item number

Description

51303 Assistance at any operation identified by the word "Assist." for which the fee exceeds $537.15 or at a series of operations identified by the word "Assist." for which the aggregate fee exceeds $537.15

20791 Initiation of management of anaesthesia for gastric reduction or gastroplasty for the treatment of morbid obesity

17610 Anaesthetist, pre-anaesthesia consultation (Professional attendance by a medical practitioner in the practice of anaesthesia) a brief consultation involving a targeted history and limited examination (including the cardio-respiratory system) and of not more than 15 minutes s duration, not being a service associated with a service to which items 2801 - 3000 apply

105 Professional attendance by a specialist in the practice of his or her specialty where the patient is referred to him or her each attendance subsequent to the first in a single course of treatment where that attendance is at consulting rooms, hospital or residential aged care facility

22025 Intra-arterial cannulation when performed in association with the administration of anaesthesia

34



30393 Laparoscopic division of adhesions in association with another intra-abdominal procedure where the time taken to divide the adhesions exceeds 45 minutes (Anaes.) (Assist.)

30601 Diaphragmatic hernia, congenital, repair of, by thoracic or abdominal approach) (Anaes.) (Assist.)

22012 Blood pressure monitoring (central venous, pulmonary arterial, systemic arterial or cardiac intracavity), by indwelling catheter - once only for each type of pressure on any calendar day, up to a maximum of 4 pressures (not being a service to which item 13876 applies) when performed in association with the administration of anaesthesia

31464 Antireflux operation by fundoplasty, via abdominal or thoracic approach, with or without closure of the diaphragmatic hiatus, by laparoscopic technique - not being a service to which item 30601 applies (Anaes.) (Assist.)

17603 Examination of a Patient in preparation for the Administration of an anaesthetic relating to a clinically relevant Service. Subsequently split out to 17610.

Source: MBS.

The most commonly claimed MBS item in combination with item 30511 was assistance during the operation (item 51303), claimed for 12,909 occasions. Anaesthesia-related services provided on the same day were claimed in nearly 6,000 occasions (the sum of services provided under items 20791, 17610 and 17603). In terms of consultations by surgeons on the day of surgery, around 1,700 services were claimed.

MBS items commonly claimed include monitoring of patient’s blood pressure during the operation (anaesthetic item 22025 and item 22012), repair of hernia (item 30601 and item 31464) and laparoscopic division of adhesions to facilitate freeing of the relevant organs taking over 45 minutes (item 30393). Blood pressure monitoring was claimed alongside item 30511 on 3,027 occasions, i.e. in 5.9% of the gastric reduction or gastroplasty procedures performed. If the requirement for additional blood pressure monitoring during operation was considered a reflection of the level of patient complexity (e.g., presence of comorbidities, high cardiovascular risks and/or super obesity), then the proportion of surgery co-billing the two MBS items is a possible indication of the proportion of patients who are at a higher risk of peri-operative complications.

Procedures to repair or to prevent the occurrence of various types of hernia were separately billed by surgeons in 4.1% of patients (2,129 occasions) who had undergone a gastric reduction or gastroplasty surgery. Repair of hiatus hernia is generally considered as a part of the bariatric surgery by most surgeons; additional claiming for such a procedure may indicate a larger defect requiring more attention (as per comments from Prof. O’Brien and Sue O’Malley).

Division of adhesions exceeding 45 minutes was claimed alongside item 30511 in 1,591 instances (3.1%).

35



Item 30518 – partial gastrectomy


0

50

100

150

200

250

300

350

400

105 30390 30473 51303 30393 30530 30387 30515 31464 104

Num

ber o

f ser

vice

s




MBS item

Description


30390 Laparoscopy, diagnostic (Anaes.)30473 Oesophagoscopy (not being a service to which item 41816 or 41822 applies), gastroscopy,

duodenoscopy or panendoscopy (1 or more such procedures), with or without biopsy, not being a service associated with a service to which item 30476 or 30478 applies (Anaes.)



30530 Antireflux operation by cardiopexy, with or without fundoplasty (Anaes.) (Assist.)30387 Laparotomy involving operation on abdominal viscera (including pelvic viscera), not being a

service to which another item in this Group applies (Anaes.) (Assist.)30515 Gastroenterostomy (including gastroduodenostomy) or enterocolostomy or

enteroenterostomy (Anaes.) (Assist.)

36



31464 Antireflux operation by fundoplasty, via abdominal or thoracic approach, with or without closure of the diaphragmatic hiatus, by laparoscopic technique - not being a service to which item 30601 applies (Anaes.) (Assist.)

104 Specialist, referred consultation - surgery or hospital (professional attendance at consulting rooms or hospital by a specialist in the practice of his or her specialty where the patient is referred to him or her)- initial attendance in a single course of treatment, not being a service to which ophthalmology items 106, 109 or obstetric item 16401 apply.

Source: MBS.

Partial gastrectomy is not a procedure that is solely indicated for clinically severe obesity. It can be performed to treat a variety of diseases including cancers. This is reflected by the co-billing of MBS item 30387 or item 30515 alongside item 30518. Laparotomy involving operation on abdominal viscera (item 30387) and gastroenterostomy or enterocolostomy or enteroenterostomy (item 30515) suggest the involvement of other organs rather than an operation confined to the stomach area as would be the case for treatment of obesity (e.g., an indication of metastasised cancer cells). The claiming of endoscopy examination of the gastrointestinal tract (item 30473) and cardiopexy10 (item 30530) are also procedures not commonly related to the treatment of obesity. Billing for initial consultation (item 104) along with partial gastrectomy (item 30518) on the same day is considered unreasonable as bariatric surgery is often performed as an elective surgery. Co-billing of the two MBS item numbers suggests that the operation is not for bariatric reasons, but rather an emergency situation such as one involving bleeding from a tumour.

Of all other MBS item numbers claimed with partial gastrectomy, pre-surgery consultation (item 105) was most commonly claimed, with 358 services co-billed during the past five years. Laparoscopy as a diagnostic procedure was also frequently claimed; however, the need for surgeons to separately claim for diagnostic laparoscopy in relation to bariatric surgery is unclear.

Like other primary interventions for the treatment of obesity, assistance during the operation (item 51303) and laparoscopic division of adhesions to facilitate freeing of the relevant organs during the procedure taking over 45 minutes (item 30393) were among the top ten MBS items claimed, totalling 207 services and 156 services respectively. MBS item for the repair of hiatal hernia by fundoplasty (item 31464) was claimed in combination with partial gastrectomy on 14 occasions. In this instance, dividing the number of these procedures claimed by the total number of partial gastrectomy operations will not give an accurate representation of the percentage of patients undergoing partial gastrectomy surgery as it is difficult to exclude the number of partial gastrectomies not performed for the treatment of obesity.

10 Cardiopexy is a procedure more commonly indicated for use for the treatment of reflux associated with tumor in the proximal stomach

37



Item 30512 – gastric bypass


0

50

100

150

200

250

51303 30393 105 30515 30378 30391 30601 58503 30443 30412

Num

ber o

f ser

vice

s




MBS item

Description




30515 Gastroenterostomy (including gastroduodenostomy) or enterocolostomy or enteroenterostomy (Anaes.) (Assist.)

30378 Laparotomy involving division of adhesions in association with another intraabdominal procedure where the time taken to divide the adhesions is between 45 minutes and 2 hours (Anaes.) (Assist.)

30391 Laparoscopy, with biopsy (Anaes.) (Assist.)30601 Diaphragmatic hernia, congenital, repair of, by thoracic or abdominal approach) (Anaes.)

(Assist.)58503 CHEST (lung fields) by direct radiography (R)

38



30443 Cholecystectomy (Anaes.) (Assist.)30412 Liver biopsy by core needle, when performed in conjunction with another intra-abdominal

procedure (Anaes.)Source: MBS.

The most commonly claimed MBS item in combination with item 30512 was for assistance during the operation (item 51303), with 222 services claimed. Laparoscopic division of adhesions to facilitate freeing of the relevant organs during the procedure consuming over 45 minutes (item 30393) was claimed in 104 instances. Of all the gastric bypass surgery performed, 8.6% of the procedures required over 45 minutes of adhesion division. This is a higher percentage than the 3.1% claimed with item 30511, reflecting the complexity of the bypass surgery. It could also be related to conversion of another procedure such as LAGB or VBG to RYGB, requiring dividing significant amounts of adhesions. A pre-admission consultation with the bariatric surgeon (item 105) on the same day of the surgery was claimed on 45 occasions.

A more diverse combination of procedures performed with item 30512 (e.g. biopsy, imaging of the chest, surgical removal of gallbladder) was observed compared to gastric reduction or gastroplasty and partial gastrectomy procedures. Although the removal of gallbladder (item 30443) during a gastric bypass procedure is not commonly considered routine, data showed the removal of the gallbladder by a surgeon at the time of bariatric surgery was one of the top ten procedures claimed with a gastric bypass. Whether this is done by some surgeons because gastric bypass procedures are likely to result in the need for the removal of gallbladder at a later date or because the patient coincidently required such surgery requires closer investigation.

39



6.5.2 Maintenance procedures for AGB

Item 14215 – accessing of long-term implanted reservoir associated with the adjustable gastric band to add or remove fluid


0

50,000

100,000

150,000

200,000

250,000

105 23 104 58909 60503 58903 55054 116 55037 61109

Num

ber o

f ser

vice

s




MBS item

Description


23 Professional attendance at consulting rooms (not being a service to which any other item applies) by a general practitioner involving taking a selective history, examination of the patient with implementation of a management plan in relation to 1 or more problems, or a professional attendance of less than 20 minutes duration involving components of a service to which item 36 or 44 applies — each attendance

104 Specialist, referred consultation - surgery or hospital (professional attendance at consulting rooms or hospital by a specialist in the practice of his or her specialty where the patient is referred to him or her)- initial attendance in a single course of treatment, not being a service to which ophthalmology items 106, 109 or obstetric item 16401 apply.

58909 Barium or other opaque meal of 1 or more of pharynx, oesophagus, stomach or duodenum, with or without preliminary plain films of pharynx, chest or duodenum, not being a service associated with a service to which item 57939 or 57942 or 57945 applies - (R)

40



60503 Fluoroscopy, without general anaesthesia (not being a service associated with a radiographic examination)(R)

58903 Plain abdominal only, not being a service associated with a service to which item 58909, 58912, 58915 or 58924 applies (R)

55054 Ultrasonic cross-sectional echography, in conjunction with a surgical procedure using interventional techniques, not being a service associated with a service to which any other item in this group applies (R)

116 Professional attendance at consulting rooms or hospital, by a consultant physician in the practice of his or her specialty (other than psychiatry) following referral of the patient to him or her by a medical practitioner — each attendance (not being a service to which item 119applies) subsequent to the first in a single course of treatment

55037 Abdomen, ultrasound scan of, including scan of urinary tract when undertaken but not being a service associated with the service described in item 55600 or item 55603, where the patient is not referred by a medical practitioner, not being a service associated with a service to which an item in Subgroups 2 or 3 of this Group applies (NR)

61109 Fluoroscopy in an angiography suite with image intensification, in conjunction with a surgical procedure using interventional techniques, not being a service associated with a service to which another item in this table applies (R)

Source: MBS.

MBS items relating to consultations make up a large proportion of the MBS numbers claimed with item 14215. The three most frequent MBS items claimed with item 14215 over the period from 2005-06 to 2009-10 were consultations with either specialists (i.e., bariatric surgeons) or a general practitioner. A total of around 243,000 follow-up consultations were claimed by a specialist, of which around 237,000 were for subsequent visits and around 6,000 were claimed for initial consultations. A legitimate claiming of the latter would be in instances where, instead of returning to the surgeon who placed the gastric band, the patient may have visited another surgeon for the adjustment of their gastric band post-surgery. Of all consultations claimed in combination to item 14215, the proportion of consultations claimed by general practitioners was around 32.4% (116,866 services). MBS item 14215 with a physician consultation after referral from a medical practitioner (item 116) was also observed; this possibly reflects multidisciplinary patient management in practice.

Of the total number of MBS item 14215 claimed during the past five years, around 2.6% of the occasions (11,570 services) were accompanied by an MBS imaging item number (i.e., Item 58909, item 60503 or item 61109) claimed on the same day. Imaging can be used to assess patency of passage and assist in the diagnosis of band slippage or a malfunctioning band. MBS items relating to a port examination – plain abdominal examination (item 58903) or an ultrasound scan of the abdomen (item 55054 and item 55037) – were claimed in around 1.2% (over 5,000 services) of the occasions at the time of band adjustment were made. These proportions may act as a rough indication of the rate of band and port complications experienced by patients who have AGB.

41



Item 31441 – repair, revision or replacement of long-term implanted reservoir associated with the adjustable gastric band


0

50

100

150

200

250

300

350

400

105 30390 20703 17610 30473 30023 51300 30393 35637 30177

Num

ber o

f ser

vice

s




MBS item

Description


30390 Laparoscopy, diagnostic (Anaes.)20703 Initiation of management of anaesthesia for all procedures on the nerves, muscles, tendons

and fascia of the upper abdominal wall, not being a service to which another item in this Subgroup applies


30473 Oesophagoscopy (not being a service to which item 41816 or 41822 applies), gastroscopy, duodenoscopy or panendoscopy (1 or more such procedures), with or without biopsy, not being a service associated with a service to which item 30476 or 30478 applies (Anaes.)

42



30023 Wound of soft tissue, traumatic, deep or extensively contaminated, debridement of, under general anaesthesia or regional or field nerve block, including suturing of that wound when performed (Anaes.) (Assist.)

51300 Assistance at any operation identified by the word "Assist." for which the fee does not exceed $537.15 or at a series or combination of operations identified by the word "Assist." where the fee for the series or combination of operations identified by the word "Assist." does not exceed $537.15


35637 Laparoscopy, involving puncture of cysts, diathermy of endometriosis, ventrosuspension, division of adhesions or similar procedure - 1 or more procedures with or without biopsy - not being a service associated with any other laparoscopic procedure or hysterectomy (Anaes.) (Assist.)

30177 LIPECTOMY radical abdominoplasty (Pitanguy type or similar), with excision of skin and subcutaneous tissue, repair of musculoaponeurotic layer and transposition of umbilicus, not being a service performed within 12 months after the end of a pregnancy and not being a service associated with a service to which item 45564, 45565 or 45530 applies

Source: MBS.

Of all MBS items claimed with ‘surgical intervention for the repair, revision or replacement of implanted reservoir associated with an adjustable gastric band’, pre-surgery consultation by a bariatric surgeon (item 105) was most commonly claimed, with 355 services paid during the past five years. Assistance received during surgery was claimed on 86 occasions.

In terms of procedures, 431 services relating to endoscopic methods such as laparoscopy to examine band integrity (item 30390) and gastroscopy to assess band position and stomach integrity (item 30473) were claimed at the time of repair, revision or replacement of the implanted reservoir. Of all MBS item 31441 claimed, 10.6% required endoscopic examination. Anaesthesia-related claims (item 20703 and item 17610) totalled 206 occasions (i.e., accompanied 5.1% of all adjustable band repair, revision or replacement procedures). Debridement of the wound area was claimed on 91 occasions; suggesting the occurrence of infection to the banding area requiring debridement and repositioning of the reservoir.

The need for more than 45 minutes of laparoscopic division of adhesion (item 30393) was claimed on 36 occasions. The clinical reasoning for performing extensive adhesion division at time of repair, revision or replacement of implanted reservoir is unclear. It is possible that the surgeons undertook the procedure in order to investigate other potential complications inside the abdomen (e.g. bowel obstruction or pain) which may or may not be related to obesity.

Other surgical procedures included in the top ten most commonly claimed MBS items alongside item 31441 were laparoscopy around the lower abdomen area (item 35637) and lipectomy (item 30177). As these are not MBS item numbers commonly claimed by bariatric surgeons, it is possible that item 35637 and item 30177 were billed by a gynaecologist and plastic-surgeon, respectively, in situations where ports need to be revised or repositioned after undertaking other surgical procedures not related to obesity.

43



6.5.3 Surgical reversal

Item 30514 – surgical reversal of bariatric surgery by any method


0

500

1000

1500

2000

51303 30393 105 30473 20791 30601 17610 22014 22025 30378

Num

ber o

f ser

vice

s




MBS item

Description




30473 Oesophagoscopy (not being a service to which item 41816 or 41822 applies), gastroscopy, duodenoscopy or panendoscopy (1 or more such procedures), with or without biopsy, not being a service associated with a service to which item 30476 or 30478 applies (Anaes.)

20791 Initiation of management of anaesthesia for gastric reduction or gastroplasty for the treatment of morbid obesity

30601 Diaphragmatic hernia, congenital, repair of, by thoracic or abdominal approach) (Anaes.) (Assist.)

44




22014 Blood pressure monitoring (central venous, pulmonary arterial, systemic arterial or cardiac intracavity), by indwelling catheter - once only for each type of pressure on any calendar day, up to a maximum of 4 pressures (not being a service to which item 13876 applies) when performed in association with the administration of anaesthesia relating to another discrete operation on the same day

22025 Intra arterial cannulation when performed in association with the administration of anaesthesia

30378 Laparotomy involving division of adhesions in association with another intraabdominal procedure where the time taken to divide the adhesions is between 45 minutes and 2 hours (Anaes.) (Assist.)

Source: MBS.

Similar to the primary surgical procedures for the treatment of obesity, the most frequently claimed MBS item in combination with item 30514 was for assistance during the operation (item 51303), which was claimed on 1,974 occasions. This represents 31.3% of all procedures relating to surgical reversal of bariatric surgery. Laparoscopic division of adhesions to facilitate freeing of the relevant organs during the procedure exceeding 45 minutes (item 30393), pre-surgery consultation by a bariatric surgeon (item 105) and anaesthesia related management, administration of and pre-anaesthesia consultation (item 20791 and item 17610), and blood pressure monitoring (item 22014 and 22025) were among the top ten items claimed on the same day alongside item 30514.

This MBS item may be claimed with either item 30511 or 30512, whichever is relevant. However, data from Medicare on same day claims from 2005-06 to 2009-10 did not reveal any instances where MBS item 30514 was claimed alongside these two primary bariatric procedures.

45



7 Systematic review of meta-analyses, systematic reviews and evidence-based clinical guidelines

A key aim of the review of existing MBS items for the surgical treatment of obesity is to determine whether these MBS items should be better defined in terms of patients and/or procedures. To inform these decisions, it is necessary to assess the safety and efficacy of each surgical procedure funded under the existing MBS items in different patient groups.

This clinical review is restricted to those procedures practised in Australia. Further, within the timeframe and scope of this review, the assessment included published meta-analyses, systematic reviews and evidence-based clinical practice guidelines, including RCTs and other studies. A systematic literature search was performed to identify the evidence.

The draft protocol for this review was originally to include studies comparing surgical procedures for obesity against non surgery (i.e. conventional treatment11) or AGB, which is the most common surgical procedure for the treatment of obesity in Australia. However, given the relative youthfulness of the bariatric surgery field as well as for both practical and ethical reasons, high level evidence was relatively scarce for surgery compared with no surgery. Therefore, and in light of comments on the draft protocol, meta-analyses and systematic reviews reporting safety and efficacy were also included where comparing surgeries against each other and/or conventional therapy. For similar reasons, meta-analyses and systematic reviews were included even if they included studies that were not higher tier RCTs.

7.1 Methods for the systematic literature searchThe aim of the literature search was to identify all published systematic reviews and meta-analyses of bariatric surgery (recall Section 5.1). Detailed description of the search strategy including search strings are in Appendix A. The study was not limited by publication date. However, it should be noted that there is a learning curve associated with bariatric surgery and outcomes from older reviews may be less generalisable to the current Australian context.

In total, 469 publications were identified using EMBASE.com and 43 publications were identified in the Cochrane library. Accounting for 7 duplicate publications in both databases, 505 publications were identified altogether.

A systematic study inclusion/exclusion approach was used for the 505 studies using pre-specified criteria. The inclusion criteria are directly related to the PICO (Population, Intervention, Comparator, Outcomes – Chapter 3) criteria used to develop the clinical questions for the review protocol – to summarise: population: people with obesity requiring bariatric surgery for weight loss; intervention: an evaluation of one or more bariatric surgery procedures;

11 Such as various combinations of diet, behaviour modification (e.g. exercise) and/or pharmacotherapy.

46



comparator: other bariatric procedures or no surgery (including non-surgical treatment such as diet, exercise and pharmacotherapy); and

outcomes: reporting at least one of a procedure-specific risk (complication or adverse event) or a benefit measure.

It was pre-specified in the review protocol (Section 3.4) that the key procedural risks associated with bariatric surgery would include: rate of converting laparoscopic to open procedure; rate of reoperation; post-operative length of hospital stay; short- and long-term procedural specific and non-specific morbidity associated with

surgery to be identified in the literature review; and mortality (short term e.g. 90 days mortality, and long term e.g. 10 years mortality).

Similarly, the key procedural benefits associated with bariatric surgery were pre-specified as: weight loss, measured as percentage of starting weight, an absolute weight loss, or

‘excess weight’ relative to a normal BMI – termed excess weight loss (EWL); maintenance of weight loss (durations used to define clinically relevant outcomes are

established as part of the review below); quality of life; and improvement in and resolution of obesity-related comorbidities (e.g. diabetes mellitus,

hypertension and hyperlipidaemia).

To capture as many relevant studies as possible, no exclusion criteria were specified by country, age group or the number of studies included in any meta-analysis or systematic review. The only exclusion criterion was the language of publication, which was restricted to English. Where study inclusion/exclusion could not be made on the basis of the title and abstract alone, the full paper was retrieved and reviewed in more detail.

Additionally, a hand search of reference lists in the included systematic reviews and meta-analysis was undertaken to ensure all relevant studies had been identified. From the hand search, two systematic reviews and six clinical practice guidelines were identified. In total, 62 studies were identified for the review. This includes ten meta-analyses, 33 systematic reviews and 19 guidelines. A full breakdown of the numbers of studies identified in the literature search and included in the review is provided in Table 1.20.

Table 1.20: Numbers of clinical studies identified and included in the literature review

Source Number of studiesEmbase + Medline 469Cochrane library +43References identified in the literature search including duplicates 512Duplicates -7References identified in the literature search excluding duplicates 505Total studies excluded for the reasons in Table 1.21 -451Studies included from initial references 54

47



Source Number of studiesAdditional studies identified and included using a hand search +8Total studies included 62Source: Deloitte Access Economics.

A breakdown of the reasons for excluding studies is provided in Table 1.21. Many studies were excluded on the basis of more than one of the inclusion/exclusion criteria listed above. For reporting purposes in Table 1.21, the hierarchy of exclusion criteria was: the intervention(s) studied is not bariatric surgery; the study is not a systematic review, meta-analysis or clinical guideline; the population is not people with obesity; the outcome reported is not related to safety or efficacy of the bariatric procedure(s); there is either no comparator, or the comparator is not a bariatric surgery or no surgery; the study is not original i.e. refers to another systematic review, meta-analysis, or

clinical guideline; and other reasons such as a comparison between various brands of surgical products,

variation of similar techniques or post-surgical management of patients.

Of the 505 references identified in the literature search, 112 studies were excluded as the interventions studied were not bariatric surgery. Most studies (253/451) were excluded on the basis that they were not meta-analyses, systematic reviews or evidence-based clinical guideline. One study utilised an animal model, hence was excluded under the population criterion. Ten studies examined other aspects of bariatric surgery not relating to safety or efficacy and 16 studies did not use an appropriate comparator. Fifty-nine further studies were excluded for other reasons such as not published in English language, not an original study, full study of abstract presented at conferences not available and updated version of study available.

Table 1.21: Reasons for exclusion

Reason for exclusion Number of studies excludedWrong intervention 112Wrong study type (not a systematic review, meta-analysis or clinical guideline) 253Wrong population 1Wrong outcome 10Wrong comparator 16Not an original study 15Not English language 2Conference abstract/ protocol (insufficient information) 3Other 39Total studies excluded 451

Source: Deloitte Access Economics.

48



To ensure the review was focused on the highest quality evidence, a brief quality assessment was made for each meta-analysis and systematic review as well as for guidelines with a systemic review component. The quality assessment was based on a simplified version of two UK methodology checklists (the Critical Skills Appraisal Skills Programme and the Scottish Intercollegiate Guidelines Network), which are referred to by the NHMRC (2009) in their guidance on systematic reviews. The questions asked of each study were: Did the reviewers try to identify all relevant studies? Did the reviewers show that they assessed the quality of the studies? If the study results were statistically combined (i.e. meta-analysis), was it reasonable to

do so? Can the results be applied/generalised to the Australian population?

The reviewers deemed the studies to be of a higher quality if the authors of the study rated the methodological quality of all the trials included in their review according to some set criteria (for example, the Jadad Quality Score, the Quality Criteria Checklist as part of the Evidence Analysis Approach designed by the American Dietetic Association, models used by US Preventive Services Task Force and the criteria from the Cochrane Handbook for Systematic Reviews of Interventions).

In total, 22 high quality studies were included in the clinical literature review as detailed in Table 1.22. These included five meta-analyses, 11 systematic reviews (i.e. studies where a systematic approach was taken to identify studies, but the data were not combined statistically), and six guidelines (i.e. guidelines which included a systematic literature search and reported on safety and efficacy outcomes).

A total of 19 clinical guidelines were identified, of these, six guidelines papers included a systematic search of the literature, graded the evidence, and reported statistics on safety and efficacy outcomes. Data extracted from these systematic reviews were reviewed along with other systematic reviews and meta-analyses in Section 7.2, while key messages in Section 7.3 were drawn from all 19 guidelines. The six guideline papers reviewed along with other systematic reviews are listed in Table 1.22 while the remaining 13 guidelines are listed in Table1.23.

Table 1.22: Systematic reviews and meta-analyses included in the clinical literature review

Author(s) and year Study title Citation

Meta-analysesBuchwald et al, 2007 Trends in mortality in bariatric surgery: a

systematic review and meta-analysisSurgery, 142:621-35

Buchwald et al, 2009 Weight and type 2 diabetes after bariatric surgery: systematic review and meta-analysis

The American Journal of Medicine, 122: 248-256

Garb et al, 2009 Bariatric surgery for the treatment of morbid obesity: a meta-analysis of weight loss outcomes for laparoscopic adjustable gastric banding and laparoscopic gastric bypass

Obesity Surgery, 19: 1447-1455

49




Shekelle et al, 2004 Pharmacological and Surgical Treatment of Obesity

Evidence Report/Technology Assessment No. 103, Prepared by the Southern California-RAND Evidence-Based Practice Center, Santa Monica, CA

Treadwell et al, 2008 Systematic Review and Meta-Analysis of Bariatric Surgery for Pediatric Obesity

Annals of Surgery, 248:763-776

Systematic ReviewsBrethauer et al, 2009 Systematic review of sleeve gastrectomy as

staging and primary bariatric procedureSurgery for Obesity and Related Diseases, 5: 469-475.

Chapman et al, 2004 Laparoscopic adjustable gastric banding in the treatment of obesity: a systematic literature review

Surgery, 135: 326-51

Colquitt et al 2009 Surgery for obesity Cochrane Database of Systematic Reviews, Issue 2. Art. No.: CD003641

De Groot et al, 2009 Systematic review: the effects of conservative and surgical treatment for obesity on gastro-oesophageal reflux disease

Alimentary Pharmacology and Therapeutics, 30: 1091-1102

Douketis et al, 2005 Systematic review of long-term weight loss studies in obese adults: clinical significance and applicability to clinical practice

International Journal of Obesity, 29:1153-1167

Farrell et al, 2009 Clinical application of laparoscopic bariatric surgery: an evidence-based review

Surgical Endoscopy, 23:930-949

Gentileschi et al, 2002 Evidence-based medicine: open and laparoscopic bariatric surgery

Surgical Endoscopy, 16: 736-744

Glenny and O’Meara, 1997

Systematic review of interventions in the treatment of prevention of obesity

NHS Centre for Reviews and Dissemination, University of York, CRD Report 10

Manterola et al, 2005 Surgery for Morbid Obesity: Selection of Operation Based on Evidence from Literature Review

Obesity Surgery, 15: 106-113

Schneider, 2000 Laparoscopic adjustable gastric banding for clinically severe (morbid) obesity

Alberta Heritage Foundation for Medical Research, Canada, HTA 7: Series B

Tice et al, 2008 Gastric Banding or Bypass? A Systematic Review Comparing the Two Most Popular Bariatric Procedures

The American Journal of Medicine, 121:885-893

Guidelines (Country to which guideline refers)

50




Department of Health and Ageing (DoHA), 2003 (Australia)

Clinical practice guidelines for the management of overweight and obesity in adults

National Health and Medical Research Council (NHMRC)

Department of Health and Ageing (DoHA), 2003 (Australia)

Clinical practice guidelines for the management of overweight and obesity in children and adolescents

National Health and Medical Research Council (NHMRC)

Kelly et al, 2005 (USA) Best Practice Recommendations for Surgical Care in Weight Loss Surgery

Obesity Research, 13:227-233

Kelly et al, 2009 (USA) Best Practice Updates for Surgical Care in Weight Loss Surgery

Obesity, 17:863-870

McTigue et al, 2003 (USA)

Screening and Interventions for Obesity in Adults: Summary of the Evidence for the US Preventive Services Task Force

Annals of Internal Medicine, 139:933-949

Pratt et al, 2009 (USA) Best Practice Updates for Pediatric/Adolescent Weight Loss Surgery

Obesity, 17:901-910

Source: Deloitte Access Economics

Table 1.23: Guidelines reviewed

Author(s), year and country to which guideline refers

Study title Citation

GuidelinesApovian et al, 2005 (USA)

Best Practice Guidelines in Pediatric/AdolescentWeight Loss Surgery

Obesity Research, 13: 274-282

August et al, 2008 (USA)

Prevention and treatment of pediatric obesity: an Endocrine Society clinical practice guideline based on expert opinion

Journal of Clinical Endocrinology & Metabolism, 93:4576-4599

Baur et al, 2010 (Australia and NZ)

Recommendations for bariatric surgery in adolescents in Australia and New Zealand

Journal of Paediatrics and Child Health, 46(12):704-707

Buchwald, 2005 (USA) Bariatric surgery for morbid obesity: healthimplications for patients, health professionals,and third-party payers

Journal of the American College of Surgeons, 200(4): 593-604

Fried et al, 2007 (European countries)

Interdisciplinary European Guidelines for Surgery for Severe (Morbid) Obesity

Obesity Surgery, 17:260-270

Laville et al, 2005 (France)

Recommendations regarding obesity surgery Obesity Surgery, 15:1476-1480

Mechanick et al, 2008 (USA)

American Association of Clinical Endocrinologist, The Obesity Society, and American Society for Metabolic and Bariatric Surgery medical guidelines for clinical practice for the perioperative nutritional, metabolic, and nonsurgical support of the bariatric surgery patient

Endocrine Practice 14(S1): 1-83.

51



Author(s), year and country to which guideline refers

Study title Citation

National Institutes of Health (NIH), 1998 (USA)

Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence report

National Heart, Lung, and Blood Institute in cooperation with the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health Publication, No 98-4083

National Institutes of Health (NIH) 1996 (USA)

NIH Consensus Statement: Gastrointestinal surgery for severe obesity - NIH Consensus Development Conference March 25-27 1991

Nutrition, 12(6): 397-402

National Institute for Health and Clinical Excellence (NICE) 2006 (UK)

Obesity – guidance on the prevention, identification, assessment and management of overweight and obesity in adults and children

NICE clinical guideline 43

Sauerland et al, 2005 (European countries)

Obesity surgery: Evidence-based guidelines of the European Association for Endoscopic Surgery (E.A.E.S.)

Surgical Endoscopy, 19: 200-221

Snow et al, 2005 (USA) Pharmacologic and Surgical Management of Obesity in Primary Care: A Clinical Practice Guideline from the American College of Physicians

Annals of Internal Medicine, 142:525-531

Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) 2008 (USA)

SAGES guideline for clinical application oflaparoscopic bariatric surgery

SAGES Practice/Clinical Guidelines, published on: 06/2008 by,

Note: All guidelines are peer-reviewed. Source: Deloitte Access Economics.

7.2 Review of meta-analyses and systematic reviews

This section provides a critical review of the high quality systematic reviews and meta-analyses identified in the literature search detailed in Section 7.1. The literature review below is structured by procedure type to more closely align with the individual MBS item numbers. Appendix B provides details of each study included in the review within a data extraction template, including key data.

Unless specified otherwise, studies focused on bariatric surgery outcomes for the obese adult population with a BMI ≥ 40 kg/m2, or ≥ 35 kg/m2 with comorbidities.

7.2.1 Gastric banding

LAGB has been noted as the most common weight loss operation performed worldwide. In the US, LAGB has not been approved by the US Food and Drug Administration for use in children; however, more recent trends show increasing utilisation in adolescents undergoing

52



bariatric surgery (Jen et al 2010). The procedure involves no bowel transection or anastomosis, is reversible, and has lower operative mortality and morbidity compared with combination weight loss procedures Thirteen studies evaluated the efficacy of gastric banding for obese patients (Buchwald et al 2007, Buchwald et al 2009, Chapman et al 2004, Colquitt et al 2009, Farrell et al 2009, Gentileschi et al 2002, Kelly et al 2005, Kelly et al 2009, McTigue et al 2003, Pratt et al 2009, Schneider 2000, Tice et al 2008, Treadwell et al 2008).

Evidence based guidelines for best practices for surgical care in weight loss surgery (Kelly et al 2009) report short-term data showing promising outcomes with LAGB; however, long-term studies raise questions on durability and re-operative rates. Kelly et al (2009) note a five year failure rate (EWL <25%, or major reoperation) of approximately 40%, and that 57% of patients will not achieve >50% EWL. Although there is some concern about long-term weight loss efficacy and re-operative rates associated with LAGB, it may be considered safer than combination procedures, with a lower mortality rate and faster recovery period (Kelly et al 2005, Kelly et al 2009). Data also show marked improvements in obesity-related comorbidities and quality of life measures post surgery (Kelly et al 2005). These advantages may make LAGB a suitable option for patients preferring a relatively safer operation rather than one with greater weight loss.

An Australian study by Chapman et al (2004) provided findings from a systematic literature review of LAGB in the treatment of obesity. Percentage of EWL at four years ranged from 44%-68% for LAGB, in comparison with VBG, which ranged from 40%-77%, and RYGB which ranged from 50%-67%. The results of this systematic review in terms of relative efficacy may therefore be considered inconclusive given the rates reported from the individual studies overlap (this is one disadvantage of systematic reviews in comparison with meta-analyses).

Chapman et al (2004) also report LAGB to be associated with a lower risk of reoperation (most reported a risk less than 8%). The greatest risk of reoperation pertains to VBG with reoperation rates of 20-53% reported. Other efficacy measures reported by Chapman et al (2004) include: revision rates, which varied from 0.6%-71% for LAGB, compared with 0.2%-10% for

RYGB; and post operative duration of hospital stay, which ranged from 1.2-11.8 days for LAGB

compared with 2.9 -11.4 days for open VBG and 1.6-8.4 days for RYGB.

The relative risk of death within 18 months after surgery from LAGB versus VBG (0.16, 95% confidence interval (CI): 0.04-0.61) and RYGB (0.10, 95% CI: 0.03-0.33) were significantly in favour of LAGB (P=.0001 and P=.007, respectively).

Buchwald et al (2009) reported that the mean change in percentage EWL up to two years after gastric banding surgery (95% CI) was 43.85% (40.25%, 47.46%), and for over two years was 48.98% (44.00%, 53.96%). A statistical test for the heterogeneity of gastric banding outcomes between the studies included in the meta-analysis was not significant. These outcomes were lower than for all types of surgery combined, including gastric banding, gastroplasty, gastric bypass, BPD, and DS. Absolute weight loss was also lower for gastric banding than for all surgeries combined. The mean change (95% CI) in absolute weight loss for gastric banding was -31.97kg (-35.14kg, -28.80kg), and for all surgeries combined was -38.47kg (-40.36kg, -36.38kg). Therefore, mean weight loss with gastric banding is reported as significantly lower than for all surgeries. However, given the comparator group includes gastric banding, results

53



for the relative efficacy of banding versus all other surgeries should be considered inconclusive from this study.

A Cochrane Review (Colquitt et al 2009) compared LAGB against laparoscopic isolated SG. The proportion of EWL was significantly greater with SG than LAGB at one year (57.7% vs 41.4%, p=.0004) and at three years (66% vs 48%, p=.0025). At three years, weight loss (29.5kg vs 17kg, p<0.0001) and reduction in BMI (27.5 vs 18, p<0.0001) were also both significantly greater with SG. All data were reported by Colquitt et al (2009) as medians and ranges, and therefore care should be taken when interpreting the results as average outcomes.

Colquitt et al (2009) also compared LAGB with no surgery (i.e. non-surgical treatment) in people with a BMI 30-35 with co-morbidity. Some 98% of those people in the LAGB group had achieved satisfactory weight loss (EWL >25%) at two years, compared to 35% of people in the nonsurgical group. BMI decreased in the LAGB group from 33.7 at baseline to 26.4 at two years (87.2% of excess weight) compared with a reduction from 33.5 to 31.5 (21.8% of excess weight) in the non surgical group over the same period. There was a significant (p<.001) difference in the weight of participants in both groups at 12, 18 and 24 months.

In Schneider’s (2000) systematic review, all studies reported decreased BMI, weight loss and/or EWL after LAGB surgery. However, there was a statistically significant difference in EWL in favour of the RYGB over both VBG and LAGB. Tice et al (2008) also evaluated the patient clinical outcomes for LAGB and RYGB. LAGB was associated with shorter hospital length of stay of about two days. Fewer deaths were reported with LAGB although mortality rates were low for both surgeries (0.06% vs 0.17%). Rates of conversion to open procedures, perforation, bleeding and anastomotic leaks were low for both types of surgery. However, overall, the reported difference in major early complications favoured LAGB. Although safety outcomes were better with LAGB, the reported observational evidence demonstrates greater weight loss and improvements in obesity-related conditions with RYGB.

The clinical application of laparoscopic bariatric surgery was reviewed by Farrell et al (2009). This study was intended to guide surgeons applying laparoscopic techniques to the practice of bariatric surgery. Farrell et al (2009) reported that weight loss after LAGB occurs in a gradual manner, with approximately 35% EWL by 6 months, 40% EWL by 12 months, and 50% EWL by 24 months. This weight loss appears to remain stable after 3–8 years based on the few studies reporting this length of follow up. In comparison, patients who undergo laparoscopic RYGB typically experience 60–70% EWL. In general, these outcomes are better than for banding procedures, which have 45–50% EWL and a less predictable improvement in comorbidities, but poorer than for BPD-DS, which has 70–80% EWL with excellent control of comorbidities. Farrell et al (2009) report open and laparoscopic RYGB as being associated with similar efficacy. Farrell et al (2009) found type 2 diabetes to be improved for about 90% of patients when assessed according to increased insulin sensitivity and increased pancreatic beta-cell function. Diabetic medications were reported to be eliminated for 64% of patients.

Farrell et al (2009) also reported life threatening complications to be less frequent with LAGB compared with laparoscopic RYGB. Case series and systematic reviews reported early mortality after LAGB to be 0.05–0.4% compared with 0.5–1.1% for laparoscopic RYGB, 0.5–1.0% for open RYGB, 1.1% for open BPD, and 2.5–7.6% for laparoscopic BPD. This suggests that mortality is lowest with AGB and highest with BPS/DS. Farrell et al (2009) report only limited comparative morbidity data from a single centre, where major and total complication rates were lower with LAGB than with either laparoscopic RYGB or laparoscopic BPD.

54



In one systematic review, Gentileschi et al (2002) found laparoscopic adjustable silicone gastric banding (ASGB) provided a shorter hospital stay and lower readmission rate than open ASGB. However, RCT data showed laparoscopic ASGB to be associated with lower weight loss than open VBG. Long-term efficacy of laparoscopic ASGB could not be determined at the time of the study because of incomplete and poor clinical evidence.

Colquitt et al (2009) also evaluated outcomes for laparoscopic ASGB, but against open ASGB. No significant difference in weight loss was reported between the procedures at 12 months. However, both laparoscopic and open ASGB were associated with a statistically significant (p<0.05) reduction in weight compared with baseline (35kg and 34.4kg, respectively). Participants undergoing the laparoscopic procedure did, however, have a significantly shorter hospital stay, and were significantly less likely to be readmitted. Early post-operative complications were similar between the groups but incisional hernia complications were experienced by 12% of patients who received the open procedure and none of the laparoscopic group.

McTigue et al (2003) described results from the large, controlled Swedish Obese Subjects (SOS) study where subjects in the surgery group underwent nonadjustable or adjustable banding, VBG, and gastric bypass. The authors noted that at two years, mean weight reduction (± standard deviation (SD)) after gastric banding was 21% (±12%), compared with 23% (±10%) for VBG and 33% (±10%) for gastric bypass. The SOS study reported that an increase in weight was observed in all surgical groups following the maximal weight loss noted after one to two years, but weight gain plateaued at eight to ten years post surgery (Sjostrom et al 2007). After 15 years, the corresponding weight losses were 13% (±12%), 18% (±11%), and 27% (±12%), respectively (Sjostrom et al 2007). McTigue et al (2003) also noted fewer surgical complications with laparoscopic compared with open procedures, although mortality was low with either technique. In cohorts receiving VBG, peri-operative mortality ranged from 0%-1.5% (6 deaths in 1165 patients using pooled data). Similar mortality rates were observed among patients undergoing either gastric bypass (0%-1.5% per data series) or AGB (0%-1.5%).

Buchwald et al (2007) conducted a systematic review and meta-analysis specifically to assess mortality after bariatric surgery. The study showed mortality after gastric banding to be relatively lower than after gastroplasty, gastric bypass and BDP-DS. Mortality rates up to 30 days after gastric banding were 0.3% for open surgery compared with 0.1% for laparoscopic surgery (meta-analysed mean difference of 0.18%, 95% CI: 0.00%-0.49%). The mortality rate between day 30 and two years was one death per 5,145 patients for LAGB (meta-analysed mean of 0.0%, 95% CI: 0.00%-0.06%). No deaths were reported with open gastric banding.

Patient subgroup data are limited among the studies of gastric banding included in the literature review. In one systematic review and meta-analysis of bariatric surgery for paediatric obesity (patients aged under 21 years), Treadwell et al (2008) reported no inpatient or postoperative deaths with LAGB or RYGB. Reoperations were performed on 8% of patients to correct various complications, including band slippage, gastric dilation, intragastric band migration, psychologic intolerance of band, hiatal hernia, cholecystitis and tubing crack. No reoperation rates were reported for RYGB, but one banded bypass study noted revision surgery was needed for gastro-gastric fistula in two cases and one each for cholecystectomy and recurrent ulcers. This information was not statistically reported. No studies reported the impact of surgery on patient growth or development.

55



Pratt et al (2009) recommended AGB for adolescent patients because of its relative safety and lower risk of postoperative vitamin deficiencies compared with RYGB or BPD. Patients undergoing AGB were aged 9-19 years and experienced 37–63% EWL during follow-up periods ranging between six months and seven years. In comparison EWL for RYGB ranged from 60.8%–64% and 70.1%–78% for BDP. Complication rates were 6–10% with no reported deaths for AGB. Complication rates for RYGB ranged from 5.5% –27%. All but one study reported no deaths for RYGB and BDP. In the study that did reported mortality rates, these were 0.5% and 1.1% for RYGB and BDP respectively. Pratt et al (2009) noted that weight loss devices should only be used in paediatric populations in a controlled clinical trial setting after investigational device exemption and institutional review board approval.

The systematic reviews did not capture one of the more recent randomised controlled trial conducted in Australia which involved assigning 50 adolescents between 14 and 18 years with a BMI greater than 35 and identifiable comorbidity to either a supervised lifestyle intervention or to undergo gastric banding (O’Brien et al 2010). O’Brien et al (2010) reported improvement in weight measures and health status in both study groups; however, the extent of the weight loss was substantially higher in the gastric banding group. The mean percentage EWL at two years in the lifestyle group was 13.2% (95% CI, 2.6%-21.0%) compared to 78.8% (95% CI, 66.6%-91.0%) in the gastric banding group. Of the ten participants with metabolic syndrome who were assigned into the lifestyle group, this resolved in six of the participants by two years, but in all nine adolescents with metabolic syndrome upon entry into the gastric banding. In the gastric banding group, 28% of the patients required a revisional procedure during the study duration. There were two unplanned pregnancies in each intervention arm.

7.2.2 Gastroplasty

VBG is now used infrequently. The technique partitions the stomach using surgical staples, creating a small segment at the top of the stomach which is partially separated from the remainder with only a small gap (stoma) remaining (Colquitt et al 2009). It is less easily reversed than gastric banding (DoHA 2003a). Several studies have shown gastroplasty to be effective in inducing weight loss (DoHA 2003a) although is not as effective as gastric bypass and weight regain is common (DoHA 2003a). Although VBG achieves significant weight loss in super-obese individuals, they remain obese, with a BMI over 35 (DoHA 2003a).

Six high quality studies evaluated the efficacy of gastroplasty for obese patients (Buchwald et al 2007, Buchwald et al 2009, Colquitt et al 2009, Kelly et al 2009, McTigue et al 2003, Schekelle et al 2004).

As reported above, McTigue et al (2003) noted that in the SOS study at two years mean weight reduction (±SD) after VBG was 23% (±10%), compared with 21% (±12%) for gastric banding. Longer term mean weight loss from the SOS study was also mentioned above in 7.2.1. The main complications with VBG were reoperation (20% to 25% over 3 to 5 years) and wound infection (8% to 32% of patients). Less frequent events (<6%) included gastric leaks, stomal stenosis, and pouch dilatations. A lower overall rate of wound infection (8%-20%) was reported for patients who underwent gastric bypass.

A second article referencing the SOS study (Schekelle et al 2004) reported that at eight years follow-up, 251 surgical patients (most of whom were treated with VBG), experienced an average weight loss of 20 kg, whereas 232 medically treated patients experienced no weight

56



loss on average. Patients undergoing VBG (or banding procedures) did, however, lose less weight than patients undergoing RYGB.

Kelly et al (2009) also found weight reduction to be greater after laparoscopic RYGB than after VBG, although both procedures had comparable operative safety and postoperative recovery when performed laparoscopically. Kelly et al (2009) considered VBG and LAGB to achieve similar results, although VBG is associated with increased peri- and postoperative complications. Kelly et al (2009) noted that VBG is generally considered when alternative weight loss surgeries are not safe or possible.

Gentileschi et al (2002) compared open and laparoscopic VBG, and concluded that laparoscopic VBG led to significantly fewer wound infections and incisional hernias. Shorter hospital stay was also reported for laparoscopic VBG. A more recent study also compares open VBG against laparoscopic VBG and reports similar EWL at 12 months (open 55% vs. laparoscopic 47%) resulting in a median BMI of 33 for both groups (Colquitt et al 2009). Longer median surgical time was reported with laparoscopic VBG (2.1 hours vs. 1.45 hours, p<.002), although there was no significant difference in the median length of hospital stay (four days in both groups).

In an update of their 2005 evidence-based guidelines for weight loss surgery, Kelly et al (2009) compared VBG and AGB and found VBG to be associated with improved EWL (70% vs. 55%, p<.001) but increased peri-operative complications (18% vs. 6%) at two years.

The Cochrane Review by Colquitt et al (2009) found significantly lower BMI with laparoscopic VBG compared with LAGB at one year (30.1 vs. 35.5, p<0.05). However, at two and three years follow-up the difference was not statistically significant (two years: 29.7 vs. 34.8; three years: 30.7 vs. 35.7). A similar result was observed in terms of percentage EWL, which was significantly greater with laparoscopic VBG compared with AGB at one year (62.3% vs. 39.2%, p<0.05) but was not significantly different at later follow-up (two years: 63.5% vs. 41.4%, three years: 58.9% vs. 39.0%, p values not reported) (Colquitt et al 2009).

Efficacy was also reported in terms of quality of life at five years follow-up. Patients were asked if they were satisfied with, or regretted having undergone, the operation. Only 56% of VBG patients were satisfied with the result of their procedure, while 81% of AGB patients were satisfied (p-value for the difference not reported) (Colquitt et al 2009).

Regarding safety, early morbidity was similar between the two laparoscopic procedures (AGB 6.1% vs. VBG 9.8%, p=0.754). However, late complications were more common following laparoscopic AGB (32.7% vs. 14%, p<0.05). No laparoscopic VBG patients required late reoperation, whereas 24.5% of LAGB patients required late reoperation (p<0.001), most commonly due to band slippage (Colquitt et al 2009).

Buchwald et al (2009) reported weight and type 2 diabetes outcomes after bariatric surgery. For gastroplasty, mean percentage EWL (95% CI) was 54.58% (46.70%, 62.46%) up to two years post-surgery, and 56.48% (52.47%, 60.49%) beyond two years. These outcomes were similar to the results for all types of surgery combined (including gastric banding, gastroplasty, gastric bypass and BPD-DS). Mean absolute weight loss in kilograms (95% CI) with gastroplasty was slightly less than for all surgeries combined being -36.07 (-39.75, -32.38) compared with -38.47 (-40.36, -36.38). Diabetes resolution (cases resolved) was greatest for patients undergoing BPD-DS (95.1%), followed by gastric bypass (80.3%), gastroplasty (79.7%), and LAGB (56.7%).

57



In their meta-analysis of mortality data after bariatric surgery, Buchwald et al (2007) reported that mortality from gastroplasty was similar to gastric banding, but lower than for gastric bypass and BPD-DS. Mortality rates up to 30 days after gastroplasty were 0.3% for open surgery compared with 0.2% for laparoscopic surgery (meta-analysed mean difference of 0.33%, 95% CI: 0.15%, 0.51%). In comparison, mortality rates with BDP/DS were 0.9% and 0.7% for open and laparoscopic surgery respectively, and for gastric banding were 0.3% and 0.1% for open and laparoscopic surgery respectively. The mortality rate between day 30 and two years was 0.3% with open gastroplasty and zero with laparoscopic gastroplasty (meta-analysed mean difference of 0.23%, 95% CI: 0.00%, 0.86%).

Systematic review conducted by Colquitt et al (2009) found seven of the ten trials comparing VBG to other surgical options reported no deaths or immediate deaths related to surgery. Colquitt et al (2009) noted that one trial comparing open VBG to open gastric bypass found a 10% mortality rate in the gastric bypass group. Another trial comparing open VBG to LAGB found a 4% mortality rate in participants undergoing open VBG while no deaths were reported in the LAGB group. The other trial that reported deaths did not specify the procedure after which the death occurred.

7.2.3 Gastric bypass

Gastric bypass was assessed in all but two of the 16 studies (Brethauer et al 2009 and Schneider 2000 did not assess gastric bypass). In the past, RYGB accounted for more than 80% of bariatric operations in the US. This is because gastric bypass was considered to offer a balance between effectiveness and risk i.e., the procedure was considered to have acceptably low morbidity and mortality rates and superior long term weight loss (Kelly et al 2005, Kelly et al 2009). For adolescents in the US, RYGB is a much more commonly available form of surgery for adolescents as LAGB is off-label (Baur et al 2010). However, with the advent of the LAGB procedures, that proportion is changing (Mechanick et al 2008, Jen et al 2010).

Douketis et al (2005) conducted a systematic review of long-term weight loss studies in obese adults, comparing surgeries with non-surgical intervention. Surgeries included in the analysis were predominantly gastric bypass surgeries. The authors found dietary and lifestyle therapy to provide less than 5kg weight loss after two to seven years, pharmacologic therapy to provide 5-10 kg weight loss after one to two years, and surgical therapy to provide 25-75 kg weight loss after two to four years with one study reporting 20 kg weight loss after eight years. Douketis et al (2005) noted that the large weight loss with surgery may depend on ongoing dietary/lifestyle interventions as surgical therapy is usually combined with dietary therapy and/or a behaviour counselling/lifestyle modification program, although not usually clearly specified. Benefits of weight loss from surgery should be balanced against a 0.5–1% postoperative mortality risk and increased risk of wound dehiscence, venous thromboembolism, and cardio respiratory insufficiency.

According to Colquitt et al (2009), hospitalisation time for LAGB is significantly shorter than for laparoscopic RYGB. However, weight loss at 10 and 15 years follow-up were greatest following gastric bypass than following VBG or (adjustable or non-adjustable) gastric banding, although statistical tests of these differences were not performed. Laparoscopic gastric bypass was also found to be superior in terms of percentage EWL compared with LAGB (66.6% vs. 47.5% at five years).

58



Kelly et al (2009) found gastric bypass for the treatment of severely obese patients to have a profoundly positive impact on obesity-related comorbidities and quality of life. The authors also found open and laparoscopic RYGB to produce similar short term weight loss and improvements in co-morbid medical conditions. The laparoscopic approach improves short-term recovery from surgery and has a lower incidence of incisional hernias than open RYGB.

As of 2002 it had not been proved (levels of evidence three and four) whether laparoscopic RYGB is as effective as open RYGB (Gentileschi et al 2002). An earlier study by Schekelle et al (2004) compared open versus laparoscopic RYGB. The study found weight loss with both approaches to be substantial and not significantly different between procedures. Because the final anatomic reconfiguration is the same for laparoscopic and open RYGB, weight loss and comorbidity outcomes should be identical. However, these procedures involve very different technical approaches that result in different types/rates of complications. In 2004 Schekelle et al concluded laparoscopic RYGB is as safe as open RYGB and is associated with lower impairment of pulmonary function and postoperative pain than open RYGB. More recent comparison in terms of safety outcomes between laparoscopic and open procedures is captured in registry data discussed in Section 7.4 below

Garb et al (2009) conducted a meta-analysis of weight loss outcomes for LAGB compared with laparoscopic gastric bypass and concluded bypass to be superior to LAGB in terms of percentage EWL at one year, two years, and more than three years. The results identified a composite EWL of 49.4% for LAGB and 62.6% for laparoscopic gastric bypass. The difference in effect sizes between the two types of surgery was statistically significant (p<0.001).

Results from the SOS study showed that at two and eight years surgical patients (all procedures) had lost 28 ± 15 kg (95% CI: 26.9 kg, 29.1 kg) and 20 ± 16 kg (95% CI: 18.0 kg, 22.0 kg) while controls had lost 0.5±8.9 kg (95% CI: -0.2 kg, 1.2 kg) and gained 0.7 ± 12 kg (95% CI: -0.8 kg, 2.2 kg), respectively (McTigue et al 2003, Torgerson and Sjostrom 2001). Mean weight reduction (±SD) at two years for gastric bypass was 33% ± 10%. Overall, surgery promoted substantial, prolonged weight loss (10 kg to 159 kg over one to five years) in patients with extreme obesity (McTigue et al 2003). These results clearly show gastric bypass to be highly effective when compared with no surgery.

Glenny and O’Meara (1997) compared RYGB with gastroplasty in a systematic review of obesity treatments. At three years follow up, 66% of gastric bypass patients achieved greater than 50% EWL versus 44% of gastroplasty patients. Glenny and O’Meara (1997) noted that compared with VBG, bypass appears to be the most effective surgical intervention in terms of maintained weight loss and low early mortality (regarding this latter finding it should be considered that this study was conducted prior to widespread use of gastric banding). The study also reports that gastric bypass may result in both vitamin and mineral deficiency, but these can be overcome by supplementation.

Bariatric surgery generally has a low mortality rate (Buchwald et al 2007). Regarding gastric bypass specifically, open surgery has much higher 30 day mortality than laparoscopic surgery (0.41% vs 0.16%). Further, although mortality with gastric bypass is lower than with malabsorptive BPD-DS, it is notably higher than for restrictive LAGB.

De Groot et al (2009) conducted a systematic review comparing the effect of various weight reducing surgeries on symptoms of GORD. All but one study found gastric bypass to have a positive effect on GORD symptoms. Comparative studies showed RYGB to provide better

59



results than gastric banding and VBG with regard to GORD symptom reduction. While all procedures resulted in weight loss, for VBG this positive effect may be counteracted by the negative effect of acid accumulation on the VBG pouch. Recently, Lee et al (2011) undertook a randomised controlled trial comparing rates of type 2 diabetes remission achieved between gastric bypass and SG. The study found higher diabetes remission amongst participants randomised to gastric bypass.

There is a paucity of information regarding outcomes for RYGB in adolescent patients. Pratt et al (2009) reported the results of one controlled multicenter study comparing laparoscopic RYGB with a one-year, family- based paediatric behavioural treatment program. In the RYGB group, BMI decreased from 56.5 to 35.8, with a significant resolution of comorbidities, whereas there was no significant BMI change in the comparator group. Pratt et al (2009) recommended RYGB as a safe operation in adolescents, with outcomes being similar to those observed in adults. However, every effort should be made to avoid postoperative vitamin deficiency and maximise postoperative compliance because adolescence is a time of increased growth and development and decreased compliance. Note these conclusions regarding adolescents are in an American context and may be less appropriate in Australia where other options are available.

The systematic review and meta-analysis of bariatric surgery for paediatric obesity by Treadwell et al (2008) reported no in-hospital deaths after gastric bypass, and no deaths likely to be directly related to bariatric surgery up to six years. However, postoperative complications following bypass included shock, pulmonary embolism, severe malnutrition, immediate postoperative bleeding and gastrointestinal obstruction. Significant weight loss was reported after gastric bypass along with hypertension resolution rates of almost 80%. BMI reduction for paediatric patients appears to be larger after RYGB than LAGB (Treadwell et al 2008). However, this may be because RYGB patients had larger pre-surgical BMIs compared to LAGB patients.

Colquitt et al (2009) compared different gastric bypass methods, and found percentage EWL at three years to be significantly greater in patients who received banded gastric bypass (a combination of RYGB and VBG) compared with those who received non-banded gastric bypass. However, differences in the resolution of comorbidities were not shown to be statistically significant.

7.2.4 Biliopancreatic diversion with duodenal switch

BPD is a partially reversible procedure which involves removing a portion of the stomach and connecting the remaining stomach to the distal part of the small intestine, the ileum. By bypassing the first two segments of the small intestine, the duodenum and jejunum, the small intestine is shortened to induce malabsorption. The gastric pouch created in a BPD is larger than that of gastric bypass or the restrictive procedure, therefore allowing larger meals and a less restricted diet following the procedure as compared to undergoing bypass (Colquitt et al 2009). Recently, BPD has been more commonly performed in conjunction with DS. The DS procedures retain a proportion of the duodenum such that the pyloric valve stays intact. The preserved pyloric valve regulates the release of stomach contents into the small intestine, facilitating greater nutritional uptake and reduces the occurrence of the dumping syndrome. BPD is effective in inducing weight loss and may be considered for the most obese patients, but can cause significant complications (DoHA 2003a, Kelly et al 2005, Kelly et al 2009). Of the 11 studies that included BPD, seven evaluated BPD with DS.

60



Kelly et al (2009) reported short-term data indicating laparoscopic BPD-DS to be the most effective surgery in terms of weight loss, having the potential to achieve greater weight loss than RYGB or LAGB. Manterola et al (2005) observed that among laparoscopic operations, BPD showed the largest decrease in BMI at one year. However, the data reported by Kelly et al (2005, 2009) also indicated that long-term nutritional and vitamin deficiencies occur at a significant rate following BPD. The increased incidence of stomal ulceration, severe protein-energy malnutrition, and dumping has resulted in the limited widespread acceptance of BPD.

The BPD-DS technique initiates dramatic weight loss during the first 12 postoperative months, which then continues at a slower rate over the next six months (Farrell et al 2009). Weight loss is durable up to at least five years. Farrell et al (2009) reported that 95% of patients with a BMI less than 50kg/m2, and 70% of patients with a BMI exceeding 50kg/m2 (‘super obese’) achieve greater than 50% EWL. The available data suggest the weight loss effect of BPD to be greater and more durable than for LAGB. Likewise, BPD may be superior to RYGB for patients with a BMI exceeding 50 kg/m2.

Buchwald (2004, 2009) confirmed these results, finding BPD-DS to result in the largest percentage EWL, which is maintained for at least two years. Further, BPD-DS has the most pronounced resolution of comorbidities, in particular diabetes.

Farrell et al (2009) also found BPD to have a dramatic impact on comorbidities. At least 90% of patients with type 2 diabetes cease diabetic medications by 12–36 months. Between 50% and 80% of hypertensive patients are cured, with another 10% experiencing improvement in hypertension. The study reports BPD to result in greater improvement of diabetes, hyperlipidemia, hypercholesterolemia, hypertriglyceridemia, and obstructive sleep apnoea than any other bariatric procedure for obesity. Up to 98% of patients with obstructive sleep apnoea experience improvement in sleep apnoea symptoms.

Whilst BPD may result in greater weight loss and resolution of comorbidities compared with other bariatric surgeries, it has the highest mortality rate (Farrell et al 2009). For open procedures, mortality risk at 30 days is 0.76% for BPD-DS, 0.44% for gastric bypass and 0.33% for gastric banding and 0.18% for gastroplasty. For laparoscopic procedures, mortality risk at 30 days ranges from 1.11% for BPD-DS to 0.06% for gastric banding. In terms of open and laparoscopic surgery, mortality risk at 30 days for open BPD-DS was 0.9% compared to 0.7% for laparoscopic (Buchwald et al 2007). Overall mortality for bariatric surgeries is 0.28% at 30 days (Buchwald et al 2007). Longer-term mortality risk (30 days to two years) is 0.85% for BPD-DS.

Outcomes of BPD surgery in adolescents suggest the risks outweigh the potential benefits of greater weight loss with BPD-DS and other procedures that cause significant malabsorption compared with RYGB or AGB (Pratt et al 2009). The other studies reviewed in this section suggest this finding may hold for the wider (adult) obese population.

7.2.5 Sleeve gastrectomy

Sleeve gastrectomy (SG) is a relatively new, non-reversible procedure increasingly being undertaken laparoscopically. In the current literature, five systematic reviews and meta-analyses studies report comparative outcomes for SG and were all published in 2009 (Brethauer et al 2009, Colquitt et al 2009, Farell et al 2009, Kelly et al 2009, and Pratt et al 2009). Brethauer et al (2009) is the only study to pool data across studies to compare SG complication and mortality rates.

61



SG divides the stomach vertically to reduce its size to about 25%. Normal stomach function and digestion are retained leading to fewer nutritional risks compared to RYGB (Pratt et al 2009). However, a recent study by Aarts et al (2011) suggests that there is a lack of evidence as research in this area is scarce. The authors found patients who had undergone laparoscopic SG to be at serious risk for developing micronutrient deficiencies and required regular monitoring. The operation is relatively short in duration, reducing the risk of operative complications. The procedure was developed as the first part of a two-part surgical procedure for patients who are at high risk from bariatric surgery, being followed at a later date (after a six to 12 months period) by a gastric bypass or a duodenal switch. When used as a first-stage procedure, laparoscopic SG has been shown to reduce weight, comorbidities, and operative risk compared to immediate DS/bypass (Farrell et al 2009).

However, for some, effective weight loss is achieved with SG alone and the procedure is now increasingly being seen as an effective stand-alone restrictive procedure for weight loss (Kelly et al 2009). Morbidity following laparoscopic SG is relatively low compared with BPD-DS or RYGB and a reduction in comorbidities has been noted (Kelly et al 2009). However, greater severities of complications have been reported with SG compared with LAGB.

Brethauer et al (2009) reported mean percentage change in EWL after SG as 55.4% (ranging from 33% to 85%) with a follow-up period ranging from three months to five years. Post-operative comorbidity data showed improvement or remission of type 2 diabetes in more than 70% of patients as well as significant improvements in other components of the metabolic syndrome (i.e. hypertension, hyperlipidemia), sleep apnoea and joint pain. Major postoperative complication rates ranged from 0% to 15.3% in studies with more than 100 patients. The overall mortality rate within 30 days was 0.19%.

Colquitt et al (2009) reported comparative data for SG relative to other surgical interventions. Laparoscopic SG was found to be superior to laparoscopic gastric bypass, in terms of EWL at 12 months (69.7% vs 60.5%, p=0.05). Laparoscopic SG was also superior to LAGB in terms of EWL at one year (57.7% vs 41.4%, p=.0004) and at three years (66% vs 48%, p=.0025). There were no statistically significant differences in BMI between procedures at 12 months follow-up. However, at three years BMI was significantly reduced in laparoscopic SG patients compared with LAGB patients (-27.5 vs -18, p<.0001). Comparisons between laparoscopic SG and gastric bypass in terms of BMI reductions at three years were not reported. No conversions to open surgery and no intra-operative and postoperative complications were reported for laparoscopic SG.

Pratt et al (2009) is the only study identified that evaluates the effect of SG on adolescents. The authors concluded that because current evidence in laparoscopic SG is significant initial weight loss with low operative risk in adults, until techniques are standardised and proof of longer-term efficacy becomes available, the procedure should only be offered to adolescents within the context of a controlled prospective study.

7.3 Review of clinical guidelinesIn total, 19 clinical guidelines were identified for inclusion in the clinical literature review. A summary of the key messages in these clinical guidelines is provided below.

62



Comparative outcomes for different bariatric surgeries Bariatric surgery is the most effective treatment for clinically severe obesity and can

result in an improvement or resolution of comorbidities, and lead to longer survival (Buchwald 2005, Mechanick et al 2008, SAGES 2008).

BPD with or without DS is effective in producing weight loss. The procedure may be associated with a variety of post-operative nutritional and metabolic complications (Mechanick et al 2008). Morbidity and mortality rates are increased in patients with a pre-operative BMI >65 kg/m2 undergoing BPD-DS (DoHA 2003a, Mechanick et al 2008).

RYGB produces greater long term weight loss than most other surgical procedures such as those attained with pure gastric restriction procedures: gastric partitioning alone, LAGB or VBG (Kelly et al 2005, Mechanick et al 2008, DoHA 2003a).

Open & laparoscopic RYGB produce similar short term weight loss and improvements in co morbid medical conditions (Kelly et al 2005). However, laparoscopic RYGB improves short term recovery from surgery and has lower incidence of internal hernia than open RYGB.

Long limb (>150cm) RYGB may produce superior short term weight loss in patients who are >200lb (approximately 90 kg) overweight or BMI≥50kg/m2 (however benefit of long limb decreases over time) (Kelly et al 2005) but this is not yet conclusive (Kelly et al 2009). However, long limb RYGB and -very very long limb-RYGB may increase risk of protein and micronutrient deficiencies (Kelly et al 2009).

LAGB produces variable short term weight loss and improvements in obesity-related comorbidities. The role of VBG is limited and the procedure has been largely supplanted by LAGB (Kelly et al 2005).

Malabsorptive procedures such as gastric bypass produce greater weight loss than restrictive procedures (e.g. banding or gastroplasty) alone (Laville et al 2005). Banded RYGB may be subject to long term complications related to reintervention, reoperation and quality of life (Kelly et al 2009).

There is evidence that a laparoscopic approach is advantageous for AGB, VBG, and gastric bypass, and probably also for BPD (Sauerland et al 2005).

Expected average weight loss and weight maintenance increases (lowest to highest) as follows: AGB, VBG, gastric bypass, BPD-DS, BPD (Fried et al 2007). However, surgical complexity and potential surgical and long-term metabolic risks of procedures also increase in that order.

Mortality risks are greatest with BPD (1%), followed by gastric bypass (0.5%), LAGB and VBG (0.1% for both) (Buchwald 2005). Bariatric surgical procedures performed by more experienced surgeons are associated with lower operative mortality (DoHA 2003a, Mechanick et al 2008).

Morbidity risks are similar, with operational morbidity risk for gastric bypass, LAGB, VBG and BPD being around 5% (Buchwald 2005).

Weight loss is greatest with BPD-DS (70% EWL and 35% BMI reduction), followed by RYGB (65%-70% EWL and 35% BMI reduction), VBG (50%-60% EWL and 25%-30% BMI reduction), LAGB (50% EWL and 25% BMI reduction) (Buchwald 2005).

63



Surgery selection AGB, VBG, RYGB and BPD are all effective in treating clinically severe obesity, but differ

in the degree of weight loss and range of complications. The choice of procedure should be tailored to the individual’s situation (Sauerland et al 2005).

Gastric bypass should be reserved for heavier patients because of the potential for metabolic complications related to malabsorption (DoHA 2003a).

Staged bariatric surgical procedures may be considered for patients at high risk for complications (Mechanick et al 2008).

At the present time (2008), treatment decisions are driven by patient and surgeon preferences, as well as considerations of the degree and timing of necessary outcomes versus tolerance of risk and lifestyle change (SAGES 2008).

A doctor–patient discussion of surgical options should include the long-term side effects, such as the possible need for reoperation, gall bladder disease, and malabsorption (Snow et al 2005).

Combination procedures lead to greater EWL and resolution of comorbidities than restrictive procedures (Kelly et al 2009).

Laparoscopic SG may be considered if other weight loss surgery options are ruled out for reasons of safety or preference (Kelly et al 2009).

Patient selection Bariatric surgery is not uniformly a ‘low-risk’ procedure, and judicious patient selection

and diligent peri-operative care are critical (Mechanick et al 2008). Each patient should be assessed based on various factors including BMI, comorbidities,

age, gender, fat distribution and GORD to determine the most appropriate treatment (Fried et al 2007; NIH 1996, NIH 1998).

Pre-operative care is extremely important to assess medical, surgical, psychiatric and nutritional needs, as well as any existing comorbidities, patient expectations and desired outcomes. The full range of treatment options and their likely outcomes should also be discussed (NIH 1996).

For adults, bariatric care should be indicated for people (Fried et al 2007; NIH 1996, NICE 2006, SAGES 2008; McTigue et al 2003): with BMI >40 kg/m2; with BMI 35-40 kg/m2 with comorbidities; who have attempted and failed to lose weight or maintain weight loss via other

methods such as lifestyle and behavioural changes and pharmacological treatment, as well as shown compliance with medical appointments (Buchwald 2005);

who are motivated and well informed, and are free of significant psychological disease; and

where the expected benefits of operation will outweigh the risk of surgery. Bariatric surgery is recommended as a first-line option (instead of lifestyle interventions

or pharmacotherapy) for adults with a BMI over 50 kg/m2 in whom surgical intervention is considered appropriate (NICE 2006).

Bariatric surgery is not recommended for (Baur et al 2010): children under the age of 14 years;

64



pregnant or breastfeeding adolescents and adults; patients with significant cognitive disabilities; patients with untreated or untreatable psychiatric or psychological disorder; or patients with Prader-Willi syndrome or other similar hyperphagic conditions.

Adults over 60 years of age should be assessed on a case by case basis and the objective of surgery should be to improve their quality of life: General nutritional safety of weight reduction at older ages is of concern because

restrictions on overall food intake could result in inadequate intake of protein or essential vitamins and minerals (NIH 1998).

Laparoscopic surgery has been performed in patients older than 55–60 years of age, but with comparatively lower weight loss, longer length of stay, higher morbidity and mortality, and lower complete resolution of comorbidities compared with younger patients. Reductions in comorbidities support the use of laparoscopic RYGB or LAGB in well-selected older patients (SAGES 2008).

Age may remain an independent risk factor for complications following weight loss surgery but procedures can be safe and effective in patients over 60 years of age (Kelly et al 2009).

Adolescent patient population There is evidence that gastric restrictive or gastric bypass surgery induces weight loss in

adolescents, with a reduction in obesity-related comorbidity that is comparable to that found in adult studies. However the evidence is based on few patient numbers and long-term follow-up data are limited (DoHA 2003b)

Bariatric surgery in severely obese adolescents should only be considered within the context of an ongoing and coordinated multidisciplinary approach (Baur et al 2010)

In Australia and New Zealand, bariatric surgery for adolescents should only be considered if all of the following criteria are met: the patient is over the minimum age of 15 years, although surgery may be

considered in exceptional circumstances at age 14 years; the patient has attained Tanner 4 or 5 pubertal development and final or near-

final adult height (August et al 2008); the patient has a BMI >40 kg/m2, or above 35 kg/m2 plus severe obesity-

associated complications; severe obesity and comorbidities persist despite a formal multidisciplinary and

supervised program of lifestyle modification and pharmacotherapy over a minimum of 6 months; and

informed consent for the surgery from the adolescent. In the US, BMI cut off points are 35 kg/m2 or above with severe comorbidities (i.e. type 2

diabetes, moderate or severe obstructive sleep apnoea, severe and progressive non-alcoholic fatty liver disease and non-alcoholic steatohepatitis, pseudotumor cerebri) and 40 kg/m2 or above with other comorbidities (e.g., hypertension, insulin resistance, glucose intolerance, substantially impaired quality of life or activities or daily living, dyslipidaemia and sleep apnoea) (Pratt et al 2009).

It is important that patients and their families understand bariatric surgery requires commitment pre- and post-surgery in terms of altering patient’s lifestyle and behaviour to maintain their weight loss (Fried et al 2007).

65



Depression, anxiety or eating disorders are not exclusion criteria for weight loss surgery. However, adolescents with mental disabilities should demonstrate ability to comply with treatment regimens and medical monitoring before weight loss surgery (Pratt et al 2009).

Surgeons performing bariatric surgery on adolescents should be experienced, credentialed for bariatric surgery and affiliated with a team experienced in the long-term follow-up and management of the metabolic and psychosocial needs of the adolescent bariatric patient and family. The institution should be one that is either participating in a study of the outcomes of bariatric surgery, or sharing data (Baur et al 2010, August et al 2008).

LAGB is the primary bariatric surgical procedure of choice for adolescents in Australia and New Zealand (Baur et al 2010).

RYGB is considered reasonably safe and highly efficacious for extremely obese adolescents in the US provided long-term follow-up occurs (Pratt et al 2009).

BPD and DS procedures should not be recommended in adolescents (Pratt et al 2009). SG should be considered investigational – existing data are not sufficient to recommend

it (Pratt et al 2009)Evidence suggests an approximate change in BMI of -20 kg/m2 after approximately two years can occur in obese adolescents undergoing bariatric surgery (NICE 2006).

Adolescent patient should be followed-up on a 4-6 weekly basis post-surgery. Long-term follow-up should extend beyond 10 years, and ideally for the whole of life (Baur et al 2010).

Post-operative management To ensure maintained weight loss, complementary long-term follow-up pathways should

be provided (both surgically and medically) through joint interdisciplinary teams (Fried et al 2007, Buchwald 2005, Laville et al 2005, Mechanick et al 2008, NIH 1996). Nutrients and vitamins should be regularly prescribed to compensate for reduced absorption and intake (DoHA 2003a).

Patients must adhere strictly to all post-operative and follow-up rules to maintain weight loss and ensure behavioural and lifestyle changes take place (Fried et al 2007).

Patients must be continuously educated regarding diet and exercise, and it should be clear that after a surgical procedure patients cannot resume their previous eating habits (Snow et al 2005).

Surgeons should be aware that post-operative complications may have an atypical presentation in the obese and early detection and timely management are necessary to prevent deleterious outcomes (Sauerland et al 2005).

Frequency of follow-up depends on the bariatric performed and the severity of comorbidities (Mechanick et al 2008)

BPD-DS require diligent lifelong patient follow-up (Kelly et al 2009). Bariatric surgery in adolescents require on-going follow-up extending beyond 10 years,

and ideally for the whole of life (Baur et al 2010). Revisional surgery can address unsatisfactory weight loss or complications after primary

weight loss surgery. It may also enhance weight loss and further improve comorbidities (Kelly et al 2009).

66



Complications, length of stay, and mortality are higher for revisional weight loss surgery, but can be safe and effective when performed by experienced weight loss surgeons.

Multidisciplinary team For patients under 18 years of age, a multidisciplinary team of specialists with

experience in adolescent care is needed for optimal pre-operative decision making and post-operative management (Apovian et al 2005, Baur et al 2010). Ideal weight loss surgery team should include a minimum of 4-5 professionals and

have at least one preoperative face-to-face meeting to prepare a treatment plan for each patient (Pratt et al 2009).

Recommendations should be developed on anaesthesia and intensive care for obese subjects (Laville et al 2005; Kelly et al 2005).

Medical imaging, lifting and transport equipment as well as beds should be adapted for patients whose corpulence is incompatible with standard models (Laville et al 2005).

The aetiology of clinically severe obesity seems to involve genetic, environmental, metabolic, and psychosocial factors. Therefore, treatment of the bariatric patient lends itself to a team approach for systematic evaluation and management (SAGES 2008).

Ideally there should be surgeons who perform bariatric surgeries frequently (50-100 cases per year) operating in properly equipped, high volume weight loss centres (100 cases per year) with integrated and multidisciplinary treatment (Kelly et al 2005), as there is a steep learning curve associated with bariatric surgery.

Rigorous training is required that strongly emphasises patients’ safety and includes close monitoring and early supervision of surgeons during the steep learning curve (Kelly et al 2005).

For gastric banding, obesity medicine specialists, nurse practitioners, physician assistants, residents and bariatric nurse specialists should be able to safely adjust bands under the supervision of a weight loss surgeon (Kelly et al 2009).

Future steps Research needs for the future include prospective data collection and interpretation of

long-term outcomes of adolescents undergoing weight loss surgery (Apovian et al 2005; NIH 1996).

Registries should be set up and procedures assessed to monitor safety and ensure good practice (Laville et al 2005).

Better reporting of clinical outcomes over the long term is required to assess each surgical treatment and to develop recommendations on which surgery type is appropriate for which patient (Apovian et al 2005; Laville et al 2005; Buchwald 2005; NIH 1996).

Patient selection criteria should be updated regularly to reflect new technologies and ongoing refinement in surgical techniques (Kelly et al 2005).

7.4 Review of registry and other relevant dataThe evidence reviewed and consultation process revealed there is currently no bariatric surgery registry to capture prospective, longitudinal patient data in Australia. This section

67



provides a summary of the findings from an analysis of the US registry data to supplement and strengthen the clinical evidence review in relation to long-term safety of bariatric surgery.

The Longitudinal Assessment of Bariatric Surgery projects (LABS-1, LABS-2 and LABS-3) are prospective, multicenter, observational studies funded by the National Institutes of Health to evaluate the risks and benefits of bariatric surgery (Belle et al, 2007). The goal of LABS-1 was to evaluate the short-term safety of bariatric surgery. The project involved 4,776 patients over 18 years of age who had undergone first-time bariatric surgery by a LABS-certified surgeon (Flum et al, 2009). The rate of major adverse outcomes (defined as death; venous thromboembolism; reintervention; and failure to be discharged from the hospital within 30 days of surgery) was 4.1%. The overall 30-day mortality rate among patients who had a RYGB or LAGB was 0.3% overall – being zero, 0.2% and 2.1% for LAGB, laparoscopic RYGB and open RYGB respectively (Flum et al, 2009). The second phase, LABS-2, set out to evaluate the longer term safety and efficacy of bariatric surgery. A subset of approximately 2,400 LABS-1 patients was recruited for this phase of the project (US Department of Health and Human Services, 2010). The project is ongoing and results are yet to be published. The third phase, LABS-3 aims to include detailed mechanistic studies in the LABS subject. Recruiting for the LABS study began in April 2007 and is also an ongoing project.

Bariatric Outcomes Longitudinal Database (BOLD) is a repository of bariatric surgery patient information put together by all participants in the Bariatric Surgery Center of Excellence program. Surgical procedure data on a total of 57,918 patients collected between June 2007 and May 2009 shows the most common bariatric surgical procedure performed in these centres of excellence was gastric bypass (54.68%), followed by gastric banding (39.62%), SG (2.29%), and BPD (0.89%) (DeMaria, 2010). The vast majority of index procedures were completed using laparoscopic surgery techniques, except for BPD, which was primarily done with an open approach. The overall mortality rate after bariatric surgery in the BOLD patient population was 0.14%, and the all-cause 30- and 90-day mortality rates were 0.09% and 0.11%, respectively (DeMaria, 2010). Mortality prior to discharge from hospital was 0.05%. BOLD suggests even lower bariatric surgery mortality is possible in quality accredited centres with a high volume of patients and specialist surgeons experienced with the appropriate procedures.

7.5 Conclusions of the clinical literature reviewThe key conclusions drawn from the systematic review of clinical studies are detailed below. Surgery for obesity is generally only recommended for clinically severe obese patients

for whom non-surgical treatments have failed to achieve and sustain adequate weight loss. Surgery for adolescents is only recommended in circumstances involving appropriate pre-operative education and post-operative follow-up, long-term multidisciplinary care, and adequate engagement of the young person and the family. Bariatric surgery for clinically severe obesity is universally reported to be more effective in inducing weight loss than non-surgical techniques. However, the common and unique risks with each type of surgery should be balanced against their benefits.

As expected, there are few relatively long-term study data and quality of life post surgery studies available for newer surgical techniques such as SG, compared with LAGB, gastric bypass and VBG. Similarly, there are limited outcome data for subgroups of the morbidly obese, including adolescents and patients with particular comorbidities.

68



LAGB is more effective than non-surgical intervention, and leads to weight reduction, where non-surgical intervention may not. LAGB is associated with lower initial operative mortality and morbidity than other surgical procedures for obesity. However, many studies find LAGB less effective than other procedures, and it may therefore be preferred for people who want a safer operation with potentially lower weight loss.

Although VBG is more effective than medical treatment (see SOS study) since it leads to weight loss when medical treatment may not, it is now infrequently used. This is because compared to LAGB, VBG is associated with increased peri- and post-operative complications and only achieves similar weight loss. Weight regain is also more common with VBG. When compared to RYGB, both procedures had comparable operative safety and postoperative recovery, however weight reduction is greater after RYGB.

Gastric bypass is considered highly effective in achieving large and rapid weight loss. However, complication rates including mortality are generally reported to be greater with bypass than with gastric banding or gastroplasty techniques.

The few clinical studies reviewed that report gastric bypass outcomes for adolescents report broadly similar efficacy compared with other procedures in the adult obese population. However, postoperative vitamin deficiency resulting from bypass is a concern given that adolescents are still undergoing bodily growth and development. Moreover, current guidelines in Australia only recommend LABG.

Laparoscopic approaches to gastric bypass, adjustable gastric banding and VBG have been demonstrated to achieve similar efficacy with improved safety compared with open techniques of the same procedure. This is because laparoscopic procedures result in the same gastrointestinal transformation as open procedures. Laparoscopic procedures may not, however, be possible for all patients (e.g. the super obese or patients with previous open abdominal operations).

BPD with or without DS is reported to lead to greater long-term weight loss and comorbidity resolution compared to gastric bypass. The highest rates of diabetes resolution have also been reported for BPD. The procedure is, however, associated with the highest mortality rates of all bariatric surgery techniques for obesity. This, coupled with higher incidence rates of malnutrition has resulted in limited widespread acceptance of BPD in Australia and most parts of the world.

The long-term clinical evidence for SG is limited, but generally shows SG leads to greater early EWL than LAGB. Before more research emerges on this rather new bariatric procedure, it should be used cautiously by bariatric surgeons as a stand alone bariatric procedure.

The lower risk of complications and favourable relative efficacy means that SG may be utilised as a first stage procedure for patients at a high risk of complications from more invasive bariatric surgery such as RYGB or BPD-DS. However, evidence of SG’s comparative long-term effectiveness is not yet established.

69



8 Systematic review of economic evaluations

A key aim of the review of MBS items for the surgical treatment of obesity is to determine whether these items should be better defined in terms of patients and/or procedures. In addition to safety, efficacy and equity considerations, these decisions may be driven by the cost effectiveness of different procedures in different patient groups. Within the timeframe and scope of the review, evidence of cost effectiveness is restricted to previous studies of procedures commonly used in Australia and demonstrated in this report to be safe and effective. Although the recommendations of this review are independent of funding decisions, the findings presented in this section may help inform such decisions in the future.

A literature search was performed to identify primary (original) full economic evaluations and reviews of primary economic evaluations conducted in Australia and overseas. Full economic evaluations included in the review compare surgical procedures for obesity against each other and/or patient management without surgery. A full economic evaluation assesses the incremental costs and incremental benefits of each surgical procedure against its comparator, and synthesises these to estimate the incremental cost effectiveness ratio (ICER) for that procedure (e.g. the cost per life year gained compared with its comparator). A common type of ICER is the incremental cost utility ratio (ICUR) i.e. the cost per QALY gained or DALY averted. The studies for inclusion were not limited by comparator, enabling each procedure to be evaluated on its own merits against no surgery and against other procedures.

8.1 Methods for the systematic literature search8.1.1 Search strategy

The aim of the literature search was to identify all published full economic evaluations of bariatric surgery for obesity. The literature search was undertaken using EMBASE.com (combined Embase and Medline databases) and the Cochrane library. A detailed description of the search strategy including search strings in presented in Appendix A. The study was not limited by publication date, although it should be noted that older economic evaluations may be less generalisable to the current context.

In total, 233 publications were identified using EMBASE.com and 89 publications were identified in the Cochrane library. Accounting for 53 duplicate publications in both databases, 269 publications were identified altogether.

It was hypothesised that the number of primary full economic evaluations would be relatively low. Therefore, broad inclusion/exclusion criteria were pre-specified to select relevant studies from the 269 references. The inclusion criteria were: population: people with obesity requiring bariatric surgery for weight loss; interventions and comparators: an evaluation of one or more bariatric surgery

procedures against other bariatric procedures or no surgery;

70



outcomes: at a minimum, a full economic evaluation with clear reporting of incremental costs and incremental benefits for each intervention versus the other(s), and synthesis of incremental costs and benefits in one or more ICERs.

To maximise the number of relevant studies for review, no exclusion criteria were specified by country, age group, or type of outcome (as long as at least one ICER was reported).

To ensure all relevant primary studies were identified, previous reviews of cost effectiveness studies were also marked for retrieval. These included systematic and non-systematic reviews of full economic evaluations of bariatric surgery. The reference lists in these review papers were searched to identify any studies not captured in the literature search.

Where study inclusion/exclusion could not be made on the basis of the title and abstract alone, the full paper was retrieved and reviewed in more detail. In total, 18 primary studies were included after two rounds of inclusions/exclusions. From the reference lists of these papers and 15 review papers, an additional five primary studies were identified. A full breakdown of the numbers of studies identified in the literature search and included in the review is provided in Table 1.20.

Table 1.24: Numbers of economic studies identified and included in the literature review

Number of studiesEmbase + MEDLINE 233Cochrane library 89References identified in the literature search including duplicates 322Duplicates 53References identified in the literature search excluding duplicates 269Primary studies included from initial references 18Additional primary studies identified from reviews and other primary studies and included 5Total primary studies included 23Source: Deloitte Access Economics.

A breakdown of the reasons for excluding studies is provided in Table 1.25. Many studies were excluded on the basis of more than one of the inclusion/exclusion criteria listed above. For reporting purposes in Table 1.25, the hierarchy of exclusion criteria are: the paper does not detail an economic study; the intervention(s) studied is not bariatric surgery; the population is not people with obesity; there is no comparator, or the comparator is not a bariatric surgery or no surgery; the study is not a full economic evaluation i.e. does not report incremental costs and

benefits and synthesise these into one or more ICERs; other reason (see below).

Of the 269 references identified in the literature search, 10 were retained as review papers to identify further primary studies. Of the remaining 259 references, 241 were excluded and 18 were included as primary studies in the literature review.

71



Most excluded studies (167/241) were excluded simply by them not reporting an economic study. After excluding nine additional studies for not including bariatric surgery, a further 65 studies were excluded due to them not undertaking a full economic evaluation. Two studies did not use an appropriate comparator. Six further studies were excluded for other reasons (not available through any source, not in English language, erratum for another study, refers to previous study only, wrong abstract published).

Table 1.25: Reasons for exclusion

Reason for exclusion Number of studies excludedNot an economic study 167Wrong intervention 8Not a full economic evaluation 59Wrong comparator 1Other 6Total studies excluded 241


In total, as outlined in Table 1.20, a total of 23 primary studies were included in the literature review (Table 1.26).

Table 1.26: Total numbers of studies included in the literature review

Reason for exclusion Number of studies excludedTotal references identified in the initial search excluding duplicates 269Relevant reviews identified 10Studies excluded 241Primary studies included 18Additional primary studies included (identified from other primary studies and reviews) 5Total primary studies included 23

Source: Deloitte Access Economics.*Two studies (Clegg et al 2002 and Picot et al 2009) include a primary study and a systematic review.

The full citation list for the 23 primary studies included in the economic literature review is included in Table 1.27. Four studies were only available as conference abstracts, but were included to increase the number of primary studies in the review (Campbell et al 2008, Maklin et al 2009, Minshall et al 2009a, Minshall et al 2009b).

No additional primary studies published up to the date of the systematic literature search commencement (2 September 2010) were identified through the literature search and a review of published systematic reviews, including the recent review by Picot et al (2009). Therefore, it is claimed with a high degree of confidence that no additional published primary studies exist beyond the 23 identified here.

72



Table 1.27: Primary economic studies included in the literature review


Ackroyd et al 2006 Cost-effectiveness and budget impact of obesity surgery in patients with type-2 diabetes in three European countries

Obesity Surgery 2006; 16(11):1488-1503

Ananthapavan et al 2010

Assessing cost-effectiveness in obesity: laparoscopic adjustable gastric banding for severely obese adolescents

Surgery for Obesity and Related Disorders 2010; 6(4):377-385

Anselmino et al 2009 Cost-effectiveness and budget impact of obesity surgery in patients with type 2 diabetes in three European countries (II)

Obesity Surgery 2009;19 (11):1542-1549

Campbell et al 2008 (poster abstract)

Cost-effectiveness of laparoscopic adjustable gastric banding & laparoscopic Roux-en-Y gastric bypass in the treatment of morbid obesity.

Value in Health 2008;11(3):A159

Campbell et al 2010 Cost-effectiveness of laparoscopic gastric banding and bypass for morbid obesity

American Journal of Managed Care 2010;16(7):e174-e187

Clegg et al 2002 The clinical effectiveness and cost-effectiveness of surgery for people with morbid obesity: A systematic review and economic evaluation

Health Technology Assessment 2002;6(12)

Craig & Tseng 2002 Cost-effectiveness of gastric bypass for severe obesity

American Journal of Medicine 2002;113(6):491-498

Hoerger et al 2010 Cost-effectiveness of bariatric surgery for severely obese adults with diabetes

Diabetes Care;33(9):1933-1939

Ikramuddin et al 2009 Cost-effectiveness of Roux-en-Y gastric bypass in type 2 diabetes patients

American Journal of Managed Care 2009;15(9):607-615

Jensen & Flum 2005 The costs of nonsurgical and surgical weight loss interventions: Is an ounce of prevention really worth a pound of cure?

Surgery for Obesity and Related Diseases 2005;1(3):353-357

Keating et al 2009a Cost-efficacy of surgically induced weight loss for the management of type 2 diabetes

Diabetes Care 2009;32(4):580-584

Keating et al 2009b Cost-effectiveness of surgically induced weight loss for the management of type 2 diabetes: Modeled lifetime analysis

Diabetes Care 2009;32(4):567-574

Maklin et al 2009 (poster abstract)

Cost-utility of bariatric surgery in the treatment for morbid obesity in Finland

Value in Health 2009;12(7):A381

McEwen et al 2010 The cost, quality of life Impact, and cost–utility of bariatric surgery in a managed care population

Obesity Surgery 2010;20(7):919–928

Minshall et al 2009a (poster abstract)

Cost-effectiveness of the Roux-en-Y gastric bypass surgery compared with medical management for treatment of Type 2 Diabetes Mellitus (T2DM) in Spain, Italy, and Sweden

Obesity Surgery 2009;19 (8):1003-1004

Minshall et al 2009b (poster abstract)

Cost-effectiveness of the Roux-en-Y gastric bypass surgery compared with medical management for treatment of Type 2 Diabetes Mellitus (T2DM) in The UK, France, and Germany

Obesity Surgery 2009;19(8):1003

Medical Services Advisory Committee (MSAC) 2003

Laparoscopic adjustable gastric banding for morbid obesity

Canberra: Medical Services Advisory Committee 2003

73



Paxton & Matthews 2005

The cost effectiveness of laparoscopic versus open gastric bypass surgery

Obesity Surgery 2005;15(1):24-34

Picot et al 2009 The clinical effectiveness and cost-effectiveness of bariatric (weight loss) surgery for obesity: A systematic review and economic evaluation

Health Technology Assessment 2009;13(41)

Salem et al 2008 Cost-effectiveness analysis of laparoscopic gastric bypass, adjustable gastric banding, and nonoperative weight loss interventions

Surgery for Obesity and Related Diseases 2008;4(1):26-32

Siddiqui et al 2006 A comparison of open and laparoscopic Roux-en-Y gastric bypass surgery for morbid and super obesity: A decision-analysis model

American Journal of Surgery 2006;192(5):e1-e7

van Gemert et al 1999 A prospective cost-effectiveness analysis of vertical banded gastroplasty for the treatment of morbid obesity


van Mastrigt et al 2006

One-year cost-effectiveness of surgical treatment of morbid obesity: Vertical banded gastroplasty versus Lap-Band



8.1.2 Data extraction

The 23 primary studies were critically reviewed and summarised using the data extraction template in Appendix C. Because of the low number of primary studies, the data extracted for the review of cost effectiveness was kept quite broad. However, a number of preferences were pre-specified with regard to the studies, relating to the methods of economic evaluation required by other government agencies such as the Pharmaceutical Benefits Advisory Committee (PBAC) and MSAC: final health outcomes expressed using quality-adjusted life years (QALYs). Where QALYs

are not reported, the review focuses on the alternative outcome metric(s) used to estimate the ICER.

Australian studies, where costs and outcomes are more relevant to the national context; ‘societal’ perspective for costs, as defined by the PBAC and MSAC. This perspective is

essentially the health care payer, where health care costs are included regardless of whether government, the patient or a third party payer bears that cost. Deadweight losses and productivity changes are ignored when interpreting the published data;

a lifetime horizon, to evaluate long term cost effectiveness; and inclusion of cost offsets due to reduced rates of other conditions associated with obesity

(e.g. diabetes, osteoarthritis, bowel cancer, breast cancer, coronary heart disease and stroke).

8.2 Results of the systematic literature searchThis section provides a brief overview of the economic studies and all full economic evaluations that have been published for bariatric surgery.

8.2.1 Overview of excluded economic studies

As noted in earlier reviews, few full economic evaluations of bariatric surgery procedures for obesity have been published. Hauri et al (1999) stated that prior to 1999 there were no

74



published cost effectiveness analysis (CEAs) of bariatric surgery. Flum (2005) reported only three cost utility analysis (CUAs) having been published by 2005, all of which are included here (Clegg et al 2003, Craig & Tseng 2002, van Gemert et al 1999). Salem et al (2008) noted that by 2008 there was only one published formal CEA evaluating RYGB (Craig & Tseng 2002) and none for laparoscopic bypass or gastric banding.

It is therefore unsurprising that recent systematic reviews found similar numbers of studies to this review. For example, Stephenson & Hogan (2007) and Picot et al (2009) both identified only five eligible original economic studies. Picot et al’s (2009) inclusion criteria are similar to the current study and the five studies they identified are included here (Ackroyd et al 2006, Craig & Tseng 2002, Jensen & Flum 2005, Salem et al 2008, van Mastrigt et al 2006).

Of the 23 studies included in the current review, 14 studies were published since 2008 and were therefore not captured in earlier reviews (some of these studies are only available as conference abstracts). This review includes nine studies published prior to 2008, which is greater than the number included in the most recent reviews (e.g. Picot et al 2009) and represents the full systematic approach and inclusion criteria adopted here.

Although the literature search identified a large number of studies that reported having undertaken an economic analysis, most did not report a full economic evaluation and were excluded. Some of the excluded studies only included procedure costs among other outcomes such as length of stay and complication rates. These studies, at most, claimed cost effectiveness on the basis of lower procedure costs and better outcomes (dominance). For example, Nguyen et al (2007) reported lower procedural costs and better outcomes with laparoscopic versus open gastric bypass. These types of study are not full economic evaluations, since they do not synthesise incremental costs and benefits for different treatment options, nor consider costs beyond the surgical procedure.

Other studies limit the economic data following the procedure to changes in pharmaceutical costs and/or productivity after surgery (e.g. Monk et al 2004 and Narbro et al 2002). These studies typically use short time horizons and narrow perspectives (e.g. costs accruing to the hospital, or pharmaceutical costs only).

8.2.2 Overview of included economic studies

All identified full economic evaluations are presented in Table 1.22, totalling 23 studies. Of these, 19 studies compare different surgical procedures with non-surgical management. Non-surgical options differ between studies, and include watchful waiting, diet and exercise programmes, medical management, or patient management prior to surgery. The 19 studies include: eight studies evaluating gastric bypass and gastric banding (Ackroyd et al 2006,

Anselmino et al 2009, Campbell et al 2008, Campbell et al 2010, Clegg et al 2002, Hoerger et al 2010, Picot et al 209, Salem et al 2008), of which one study (Clegg et al 2002) also evaluates vertical banded gastroplasty (VBG);

five studies evaluating gastric bypass only (Craig & Tseng 2002, Ikramuddin et al 2009, Jensen & Flum 2005, Minshall et al 2009a, Minshall et al 2009b);

three studies evaluating gastric band only (Ananthapavan et al 2010, Keating et al 2009a, Keating et al 2009b);

75



two studies evaluating bariatric surgery in general i.e. a pooled analysis of gastric bypass, gastric band and other surgeries (Maklin et al 2009, McEwen et al 2010); and

one study evaluating vertical banded gastroplasty only (van Gemert et al 1999).

Of the four studies without a non-surgical comparator: two studies compare gastric banding with VBG (MSAC 2003, van Mastrigt et al 2006), of

which one study (MSAC 2003) also includes gastric bypass; two studies compare open and laparoscopic gastric bypass (Paxton & Matthews 2005,

Siddiqui et al 2006).

Cost utility analysis (CUA) – analysis of the cost per quality-adjusted life year (QALY) gained with surgery – was performed in 19 studies (Ackroyd et al 2006, Anselmino et al 2009, Campbell et al 2008, Campbell et al 2010, Clegg et al 2002, Craig & Tseng 2002, Hoerger et al 2010, Ikramuddin et al 2009, Jensen & Flum 2005, Keating et al 2009b, Maklin et al 2009, McEwen et al 2010, Minshall et al 2009a, Minshall et al 2009b, MSAC 2003, Picot et al 2009, Salem et al 2008, van Gemert et al 1999, van Mastrigt et al 2006).

The remaining four studies undertook CEA. Two studies evaluated gastric banding against no surgery in Australia, and estimated the cost per disability-adjusted life year (DALY) averted (Ananthapava et al 2010) and cost per diabetes case remitted (Keating 2009a). The trial-based analysis published by Keating et al (2009a) was adapted to a lifetime CUA in another publication (Keating et al 2009b). The other two CEAs evaluated laparoscopic versus open gastric bypass (Paxton & Matthews 2005, Siddiqui et al 2006). In both studies the analysis was in effect restricted to cost minimisation analysis, since dominance of the laparoscopic procedure was claimed on the basis of a lower procedure cost and similar efficacy to the open procedure.

Of the 23 economic studies, analyses were performed for the following countries/regions: ten studies in the US; nine studies in Western Europe, including the UK (four studies), France, Germany, Italy,

Netherlands, Spain (two studies each), Austria, Finland, and Sweden (one study each) ; and

four studies in Australia.

No full economic evaluations were performed for other countries with similar health care systems to Australia, for example Canada and New Zealand, or for other world regions including Asia. However, it is expected that procedure costs in the countries listed above would be quite similar to Australia with the exception of the US.

Eight of the 23 studies were undertaken specifically for the obese population with type 2 diabetes mellitus (T2DM) (Ackroyd et al 2006, Anselmino et al 2009, Hoerger et al 2010, Ikramuddin et al 2009, Keating et al 2009a, Keating et al 2009b, Minshall et al 2009a, Minshall et al 2009b). Other studies defined obese populations using BMI with or without age restrictions, or using a higher BMI without comorbidities and a lower BMI with comorbidities (which could include diabetes).

76



8.3 Systematic review of the primary economic studies

This section provides a critical review of the data extracted from the economic evaluations. An overview of the 23 studies is presented in Appendix C within the data extraction template. The tabulated descriptions of the populations and interventions are as provided in the publications. For example, studies that only refer to the ‘obese population’ without specifying BMI, or ‘gastric bypass surgery’ without specifying the surgical technique (e.g. Roux-en-Y), have no further details tabulated.

The literature review is structured in order of relevance.

First, the four Australian studies are discussed, since their results are considered most transferable to this report. These studies include four analyses LAGB compared with either no surgery (three studies) or VBG (one study).

Second, the 17 other CUAs are discussed, with results disaggregated by surgical technique for reference to the MBS item numbers. As detailed in Section 1.1.2, CUAs are the preferred economic analysis, since (a) they are consistent with the measurement of cost effectiveness used by the PBAC and MSAC to make health care funding recommendations, and (b) QALYs incorporate both mortality and morbidity and can therefore be compared to all other health care interventions. CUAs report the ICUR: the cost per QALY gained with the study surgery versus an appropriate comparator.

Third, the two other CEAs are reviewed, which both evaluate laparoscopic versus open RYGB.

Importantly, when comparing the cost effectiveness of different surgeries, for example bypass versus banding, it is not appropriate to compare ICERs for each procedure versus no surgery, either across or within studies. Instead, the more effective surgery should be compared with the less effective surgery within the same study. This minimises bias and ensures an appropriate surgical comparator. In the discussion below, the cost effectiveness of different procedures is only compared in the context of studies that directly calculate an ICER for one procedure versus another.

8.3.1 Australian cost effectiveness studies

All four published Australian cost effectiveness studies have evaluated LAGB only. Ananthapavan et al (2010) have performed the only economic evaluation of LAGB in the severely obese adolescent population. The study extrapolated limited case series data for 28 severely obese adolescents reporting the change in BMI at three years post-surgery to a lifetime horizon. The cost per DALY saved with LAGB was estimated to be $4,400 (95% confidence interval [CI]: $2,900, $6,120). Similar to other published economic studies with extended time horizons, the assumption that BMI change will continue into the future may be overly simplistic, and modelled BMI changes using long term follow up data from other studies would be more appropriate.

Keating et al (2009a, 2009b) undertook an analysis of LAGB versus medical management in Australians with BMI 30-40 (kg/m2) with T2DM diagnosed in the previous two years. A head-to-head RCT analysis compared resource costs (surgery costs, outpatient consultations and

77



tests, weight loss assistance therapies, and medications) with the number of diabetes cases remitted at two years. This within-trial analysis demonstrated a cost of $16,600 per diabetes case remitted with LAGB. This outcome is also difficult to interpret in the context of the overall cost effectiveness of LAGB, particularly since the long term cost and health consequences of remitted diabetes are not captured.

In a further study, Keating et al (2009b) extended this analysis to a lifetime horizon by applying an annual diabetes relapse rate based on a mean remission duration of 13 years, and annual mortality. Accounting for intervention costs and the annual cost of diabetes, LAGB was projected to save $2,444 and confer an additional 1.2 discounted QALYs over a patient lifetime. Hence, LAGB was claimed to dominate medical management and to be cost effective.

The other study – an MSAC assessment report of LAGB (MSAC 2003) - compared LAGB with VBG in the Australian obese population. This economic evaluation attempted to update the UK study by Clegg et al (2002) with Australian procedure, re-operation, and band adjustment costs and thus estimate the cost effectiveness of LAGB in Australia. Using Clegg et al’s (2002) incremental gain of 0.14 QALYs over 20 years with LAGB vs VBG, and an estimated procedure cost difference of $3,665, MSAC (2003) estimated the maximum incremental cost utility ratio (ICUR) to be $26,178 with LAGB. Unfortunately, the MSAC analysis has several major limitations including the omission of health care costs beyond the surgery and QALYs being assessed for the UK population. Hence, costs and benefits are not matched with regard to the time horizon and population.

In summary, only one CUA has been undertaken in Australia using appropriate methods (Keating et al 2009b). However, this study is restricted to the Australian obese population with recently diagnosed T2DM, and the costs and benefits are related to diabetes remission rather than a wider range of health sequelae associated with obesity.

8.3.2 Non-Australian cost utility studies

Other CUAs undertaken outside Australia are discussed below, by surgical procedure.

8.3.2.1 Gastric bypass

Thirteen of the CUAs performed outside Australia evaluated the cost effectiveness of gastric bypass (Ackroyd et al 2006, Anselmino et al 2009, Campbell et al 2008, Campbell et al 2010, Clegg et al 2002, Craig & Tseng 2002, Hoerger et al 2010, Ikramuddin et al 2009, Jensen & Flum 2005, Minshall et al 2009a, Minshall et al 2009b, Picot et al 2009, Salem et al 2008). The UK results for Clegg et al (2002) are not discussed in this section since Picot et al (2009) updated the analysis using the same model.

It should also be noted that MSAC (2003) did not evaluate the cost effectiveness of RYGB due to concerns over whether bypass or LAGB has superior benefits. RYGB was reported as conferring greater weight loss but also greater post-surgical mortality.

Of the 12 studies that evaluated the cost effectiveness of gastric bypass, half were performed for obese populations with T2DM.

In patients with a BMI over 35 and T2DM, at five years, Ackroyd et al (2006) and Anselmino et al (2009), found gastric bypass to dominate conventional management (one year diet and four

78



years watchful waiting) in Austria, France, Germany and Italy, and have a favourable ICUR in Spain and the UK. Even under worst case scenarios, the ICURs for bypass increased to a maximum of £2,599 in the UK and €4,347 in Spain.

Using a lifetime Markov model, Hoerger et al (2010) estimated ICURs of US$7,000 and US$12,000 for US patients newly diagnosed with diabetes and with established diabetes, respectively. Most convincingly, in a probabilistic sensitivity analysis (PSA) 95% of ICURs were below US$23,000.

Ikramuddin et al (2009), Minshall et al (2009a), and Minshall et al (2009b) all used the Centre for Outcomes Research (CORE) diabetes Markov model to evaluate the cost effectiveness of RYGB compared with medical management of diabetes. At 35 years, the ICURs were US$21,973 (US), £2,922 (UK), €298 (France), €2,034 (Spain), €364 (Italy), and SEK24,437 (Sweden). Bypass dominated medical management in Germany. These results were evaluated at 35 years for the US, and a lifetime horizon for all other countries. Ikramuddin et al (2009) report, however, that RYGB is not cost effective over shorter time horizons, with an ICUR of US$122,001 at ten years and dominance by medical management at five years. Although Minshall et al (2009a, 209b) do not report their sensitivity analyses it is assumed this finding holds for the other countries.

It should be considered that the CORE model, which examines the impacts of diabetes, may capture benefits of surgery for obesity differently that a model specifically designed to capture the impacts of weight loss. This is one shortcoming of diabetes-specific analyses such as those reported above.

Of the other six studies undertaken for the general morbidly obese population, five studies were undertaken for the US and five were undertaken for the UK.

The earliest US study (Craig & Tseng 2002) estimated cost effectiveness in a subset with BMI 40-50 aged 35-55 years, based on data from a single study with 14 years follow up. Craig & Tseng (2002) found the ICUR to vary markedly by gender and BMI, between US$5,700 in women aged 35 years with BMI 50 to US$35,600 in men aged 55 years with BMI 40. Although Craig & Tseng (2002) performed a limited range of one-way and two-way sensitivity analyses, they noted that bypass was not cost effective under all potential parameter values. In a later decision tree analysis based on Craig & Tseng’s (2002) approach, Salem et al (2008) reported similar or lower base case ICURs for the same subgroups, and ICURs below US$25,000 in all one-way sensitivity analyses.

In two publications, Campbell et al (2008, 2010) reported a CUA based on a lifetime Markov model and a single five-year head-to-head RCT of laparoscopic RYBG (LRYGB) versus LAGB. No treatment was assumed to confer no weight change, which may bias the results for or against surgery. Unsurprisingly, both publications report similar results, although the results are presumably updated in Campbell et al (2010). For all patient subgroups defined by gender and BMI (35-40, 40-50, 50+), the ICUR for LRYGB versus no surgery in people aged 40 years remained below US$27,000. LRYGB was also shown to be highly cost effective using probabilistic sensitivity analysis (PSA).

Jensen & Flum (2005) also used a lifetime Markov model to estimate the cost effectiveness of bypass versus diet and exercise in morbidly obese white women in the US. However, this study has some critical flaws including a lack of control for differences in the underlying studies

79



(an indirect comparison of two studies was used), and an unconventional approach of modelling interventions that occur at different times during a person’s lifetime. The rationale for the selection and applicability of the two efficacy studies is not presented. The study showed the base case ICUR to be US$7,126 and the worst case ICUR to be US$35,000, although these results should be interpreted with caution due to the approach.

Picot et al (2009) designed a Markov model to evaluate the cost effectiveness of bariatric surgery, including key health states of T2DM, stroke, coronary heart disease (CHD), remission of comorbidity, and death. Weight loss with surgery was derived from a systematic review of studies, and a targeted search identified studies linking BMI with adverse events and changes in utility. The Picot et al (2009) study appears to report the best designed economic model of those reviewed for this report, at least for a UK cohort analysis.

Even under ‘pessimistic’ assumptions for weight reduction following surgery, Picot et al (2009) found the ICUR for bypass to be £4,127 in the base case analysis for people with BMI > 40, and less than £10,000 in all one-way sensitivity analyses for that population. Unfortunately, Picot et al (2009) only conducted subgroup analyses and PSA for LAGB (see Section 1.3.2.2) and not bypass.

8.3.2.2 Gastric banding

Nine CUAs performed outside Australia evaluate the cost effectiveness of gastric banding (Ackroyd et al 2006, Anselmino et al 2009, Campbell et al 2008, Campbell et al 2010, Clegg et al 2002, Hoerger et al 2010, Picot et al 2009, Salem et al 2008, van Mastrigt et al 2006).

The methods and limitations of eight of these studies were described in Section 1.3.2.1, since these studies also evaluated bypass. The other study by van Mastrigt et al (2006) was a one-year within trial analysis comparing LAGB and VBG for Netherlands patients with BMI > 40, or 35-40 and significant comorbidity. The study evaluated excess weight loss (EWL) at 12 months, utility (quality of life) at three, six, and 12 months, and actual cost data from billing records during the previous year. Bootstrapping analysis revealed no significant difference between the costs or QALYs with either procedure, but significantly higher EWL with VBG (17.82%; 95% CI: 9.60%, 26.05%). The study results were therefore non-conclusive. The authors recommended the surgery choice be based on efficacy, safety, clinical aspects, and long term cost effectiveness. In particular, the analysis was only conducted for one year and patients lose weight more slowly with LAGB than with VBG.

All eight studies previously described in Section 1.3.2.1 showed LAGB to be cost effective compared with no surgery across all sensitivity analyses.

In three US studies of the morbidly obese, the ICURs for LAGB were all below US$25,000 when considering different genders, ages, BMIs, and one-way sensitivity analyses (Campbell et al 2008, Campbell et al 2010, Salem et al 2008). Furthermore, in PSA, Campbell et al (2010) found all simulations to report higher costs and QALYs with LAGB, and an ICUR below US$50,000.

Picot et al (2009) used an updated version of the 20-year Markov model reported by Clegg et al (2002) to evaluate LAGB against non-surgical treatment in the UK. Even under ‘pessimistic’ assumptions for weight reduction following surgery, the ICUR ranged between £1,367 in

80



people with BMI 30-40 and T2DM and £12,673 in people with a BMI 30-35. In sensitivity analyses, the ICUR remained below £34,000 for all subgroups.

In patients with BMI > 35 and T2DM, at five years, Ackroyd et al (2006) and Anselmino et al (2009) found adjustable gastric banding to dominate conventional management in Austria, France, Germany and Italy, and have a favourable ICUR in Spain and the UK. Even under worst case scenarios, the ICURs for banding increased to a maximum of £3,251 in the UK and €3,142 in Spain.

Using the model described in Section 1.3.2.1, Hoerger et al (2010) estimated the cost effectiveness of banding against usual care for diabetes. The study reported ICURs of US$11,000 and US$13,000 for US patients newly diagnosed with diabetes and with established diabetes, respectively and, convincingly, 95% of ICURs were below US$30,000 in a PSA.

Three studies compared LAGB with bypass, in addition to comparing each surgical procedure with no surgery (Campbell et al 2008, Campbell et al 2010, Clegg et al 2002). All found LAGB to not be cost effective compared with bypass (or found bypass to be cost effective compared with LAGB).

Campbell et al (2008) estimated the ICUR for LRYGB versus LAGB to be US$16,540 or US$5,780 for women aged 40 years with a BMI 35-40 or 40-50, respectively, which is within the bounds of cost effectiveness (recall Section 5.3 regarding WHO benchmarks for cost effectiveness). This reflects higher lifetime costs with bypass (considering procedure and complication costs only) but greater lifetime QALY gains.

Campbell et al (2010) further showed this result to hold across extensive sensitivity analyses. The ICUR for LRYGB versus LAGB was below US$15,000 across all subgroups of the US population aged 18-74 years with BMI > 40, or BMI > 35 with significant comorbidity. More convincingly, PSA showed all ICURs for LRYGB versus LABG to be below US$50,000 (conventionally considered an upper limit for cost effectiveness in the US).

Clegg et al (2002) found LAGB to be more effective than bypass over a 20 year horizon, but with an ICUR of £256,856. In the same study, however, bypass was reported to be cost effective compared with VBG or no surgery (see Section 1.3.2.1).

As a qualification to these results, all three studies state that not all future benefits of LAGB may be captured in their analyses, since long term studies have been undertaken for bypass but not for banding. Further, few head-to-head controlled study data are reported for bypass and banding (at least in economic evaluations), and results therefore reflect substantial modelling and assumptions.

8.3.2.3 Vertical banded gastroplasty

Four CUAs performed outside Australia considered VBG (Clegg et al 2002, MSAC 2003, van Gemert et al 1999, van Mastrigt et al 2006).

The study by van Gemert et al (1999) is the earliest full economic evaluation identified in the literature search. This is to be expected, since VBG is now used less often in favour of bypass or banding. The cost effectiveness of newer procedures has received more attention in the economic literature in recent years.

81



Van Gemert et al (1999) performed their analysis for the Netherlands population with BMI > 40, using intention-to-treat (ITT) patient data (n = 21) before and two years after VBG. The comparator is therefore inferred to be continued management in the absence of surgery. Life expectancy gains were estimated by applying an unspecified adjustment (using the Declining Exponential Approximation of Life Expectancy or ‘DEALE’ method) to a study of mortality in clinically severe obesity published in 1979.

The authors estimated a lifetime gain of 12 QALYs based on a utility increase of 0.25 following VBG, a standard life expectancy of 47.8 years, an increased life expectancy of 3.6 years with VBG, and a 5% discount rate. Lifetime discounted cost of illness savings were reported to exceed the procedure cost by US$2,164 (US$8,029 - US$5,865). Thus, the authors concluded VBG to be dominant and cost effective, even without considering productivity gains.

The van Gemert et al (1999) study results should be interpreted with caution due to some major methodological shortcomings. First, a before-and-after study was used rather than measuring outcomes within a controlled study. Second, the utility change at two years was applied over a lifetime, and does not consider differences in long term health status for either patient group. Third, the costs of illness for clinically severe obesity appear to only be applied in the no surgery group.

The UK study by Clegg et al (2002), which used a model later updated by Picot et al (2009) as described in Section 1.3.2.1, evaluated the cost effectiveness of VBG compared with no surgery. The study also used VBG as a comparator for bypass or banding. As an aside, VBG was omitted from the more recent Picot et al (2009) study, most likely since “expert opinion indicates that VBG is almost never undertaken (in the UK) now” (Picot et al 2009). The ICUR for VBG was £10,237 at 20 years, which would be considered cost effective by conventional UK and Australian standards.

The studies by MSAC (2003) and van Mastrigt (2006) both included VBG as a comparator for LAGB, but did not specifically evaluate the cost effectiveness of VBG. These studies were detailed in Section 1.3.1 and Section 1.3.2.2, respectively.

8.3.2.4 Pooled intervention studies

Two other CUA studies performed outside Australia compared bariatric surgery (generally) with non-surgical management in Finland (Maklin et al 2009) and the US (McEwen et al 2010). Both studies did not specify a patient population, but implied this to be patients undergoing bariatric surgery for weight loss in general.

Details of the Maklin et al (2009) study are scarce, since it is only reported as a conference abstract. However, bariatric surgery included gastric bypass, gastric banding, and SG. As such, this study is the only identified economic evaluation that considers SG, albeit not as a distinct procedure. Bariatric surgery was reported to dominate non-surgical management, conferring an additional 0.54 QALYs and cost saving of €13,000 at ten years.

The limited information in the abstract suggests only the impacts of re-operation, abdominoplasty, and death were considered (in addition to survey reported quality of life) since these were the health states in the Markov model. Surprisingly, the abstract does not refer to the modelling of weight change. This suggests the study may not have considered all events of interest following surgery for obesity and as such the results may be of limited value.

82



McEwen et al (2010) collected cost and utility data at 12 months post surgery (33% LRYGB, 64% open RYGB, 3% unspecified other procedure) and extrapolated these data to two years or a patient lifetime using assumed changes in treatment costs. Outcomes with no surgery were estimated by extrapolating cost data for the same patient group prior to surgery and using utilities collected from a different patient cohort requesting surgery. McEwen et al (2010) concluded bariatric surgery to have an ICUR of US$48,662 at two years and US$1,425 over a patient lifetime. This suggests surgery is cost effective soon after the procedure, with cost effectiveness improving over time.

However, McEwen et al’s (2010) approach limits the study’s credibility compared with other studies using head-to-head data derived from similar patient groups over the same time period. Furthermore, the ICURs were estimated by extrapolating data at 12 months rather than modelling future events based on long term weight changes and associated costs and utilities.

8.3.3 Other non-Australian cost effectiveness studies

Two US studies compared laparoscopic RYGB (LRYGB) and open RYGB using CEA.

Paxton & Matthews (2005) undertook a meta-analysis of clinical studies published between 1984 and 2000 and reported similar EWL at three years but lower complication rates for LRYGB versus open bypass. Using US cost data the authors estimated net cost savings of US$2,783 (2004 prices) with LRYGB; although procedure costs were higher there were cost offsets due to lower hospital costs for recovery and lower complications rates. The cost of complications included hospital bed days and lost patient income due to convalescence and peri-operative death.

Paxton & Matthews (2005) concluded that LRYGB is more cost effective than open RYGB on the basis of a lower cost, similar efficacy, and fewer complications. However, published data suggest that had income losses been excluded (e.g. restricting the study to a hospital perspective), LRYGB would have a greater cost. This could potentially change the findings of the study. The authors note that long term efficacy may be similar between procedures, which rely on the same surgical alterations to produce their effects, and differences are restricted to complication rates and associated costs.

Siddiqui et al (2006) performed a similar restrictive analysis of laparoscopic RYGB versus open RYGB. Weighted mean mortality and complication rates within one year of surgery were derived from 35 studies identified in a literature review of clinical studies published since 1990. In three BMI groups (35-49, 50-59, and 60+), LRYGB was associated with a higher rate of success (defined as no major complications within one year of surgery), lower mortality, and lower costs (including the procedure and complications). Similar to the Paxton & Matthews (2005) study, Siddiqui et al (2006) thus concluded the laparoscopic RYGB procedure to be more cost effective.

Although both studies are limited to the first three years post surgery and the costs of procedures and their complications, this is appropriate since different methods of performing the same surgical adjustment are being compared. At most, the results of these studies suggest that if these procedures are performed in similar ways in Australia, then LRYGB may be a more cost effective method of gastric bypass. This appears to be in line with current trends in Australia to perform bypass laparoscopically when possible.

83



8.4 Conclusions from the economic literature review

The key conclusions drawn from the systematic review of economic studies are detailed below. Most economic evaluations of bariatric surgery for obesity have been published since

2005 and evaluated LAGB and gastric bypass. Importantly, there is a lack of well-performed Australian studies and CUAs.

There is limited economic evidence for VBG; however, published data suggest more recently developed surgical techniques are cost effective when compared with VBG.

Bariatric surgery for obesity is universally reported to be cost effective compared with no surgery even across extensive deterministic and probabilistic sensitivity analyses. The likelihood of publication bias should be considered; that is, researchers finding surgery less cost effective may be less likely to seek publication. However, Picot et al (2009) performed arguably the most comprehensive economic evaluation for a National Institute for Health and Clinical Excellence (NICE) review, which is less likely to have vested interests, and found surgeries to be cost effective.

Many studies show surgery to be cost saving after several years, although the scope of costs in these studies should be carefully considered.

On balance, surgery appears to be more cost effective in women and younger people (due to greater life expectancy over which benefits accrue), and people with higher BMIs and comorbidities such as diabetes (in whom surgery makes the greatest clinical difference). Surgery is also more cost effective in people with newly diagnosed diabetes compared with established diabetes (at least two years since diagnosis).

Generally, lower ICERs are reported for LAGB than for bypass, when compared with no surgery. However, the cost effectiveness of one procedure versus another should only be compared using the incremental costs and benefits for one procedure versus another procedure, and within the same study to control for other factors.

When compared directly, bypass appears cost effective relative to banding (or banding is not cost effective relative to bypass) with a favourable cost for the additional clinical benefits. However, this outcome may in part be driven by the underlying data since there are (a) a lack of head-to-head study data, and (b) a lack of evidence on long term outcomes with banding.

Laparoscopic bypass appears to be cost effective compared with open bypass, assuming similar outcomes, since savings in complication costs outweigh any additional procedure costs. From an economic viewpoint, LRYGB should potentially be used over open bypass unless laparoscopic procedures are contra-indicated in the patient or conversion is required during surgery.

84



9 ConclusionsThis section of the report outlines the key conclusions drawn from the review of the clinical and economic literature, and the analysis of MBS data showing bariatric surgery usage for private patients in Australia.

9.1 Conclusions regarding the MBS items

Consideration be given to splitting MBS item 30511 (for gastric reduction and gastroplasty) into several items, each describing a specific procedure.

MBS item 30511 potentially covers a range of surgical procedures for obesity including AGB, VBG, and other gastric reduction procedures that cannot be claimed under other MBS item numbers. These different procedures are associated with different efficacy, risks and supportive data, and their aggregation into a single MBS item and benefit level may not therefore be appropriate. Therefore, consideration should be given to respecifying the term ‘gastric reduction’ as ‘adjustable gastric banding’ avoids any ambiguity with other reductive procedures that may have less clinical support, or may be miscoded (e.g. SG, of which the majority is currently claimed under MBS item 30518).

Further consultation with clinical experts will determine whether gastroplasty (i.e. VBG) continues to be funded through the MBS. The use of gastroplasty has declined over the past decade in favour of other forms of surgery, and is rarely used in Australia and countries with similar health care systems such as the UK and the US. Further, several economic studies have demonstrated RYGB and LAGB to be cost effective relative to VBG. If gastroplasty is retained on the MBS, the procedure should be respecified as VBG, for which there is clinical evidence of efficacy in the surgical treatment of obesity.

It is preferable that AGB and VBG (if remaining on the MBS) should be separated into two MBS item numbers, since they are associated with different surgical risks, and potentially different long-term efficacy.

Consideration be given to splitting MBS item 30512 into separate items for gastric bypass and BPD with or without DS.

Three main types of gastric bypass can be used to treat obesity: RYGB: this is a surgical procedure for obesity in Australia and is associated with

relatively higher efficacy compared to gastric banding. There is reasonable amount of literature to support continued existence of this procedure; however, gastric bypass should be treated differently from other truly mal-absorptive procedures.

BPD with or without DS: although this type of surgery is associated with higher efficacy than RYGB, it incurs a higher risk of serious surgical and nutritional complications. Consideration should be given in conjunction with clinical experts, whether BPD should potentially be funded on the MBS, as is currently possible under

85



the present descriptor for MBS item 30512. Alternatively, if could continue to be funded under a separate MBS item number, such that utilisation in Australia can be better monitored.

Loop gastric bypass: this version of gastric bypass has been abandoned by bariatric surgeons for the last 20 years since its side effects (reflux, ulcerations, leakage etc) are not considered to outweigh its benefits in weight loss management.

Based on the above, it is concluded that there be separate MBS item numbers specifying gastric bypass and BPD with or without DS (if funded through the MBS), to avoid any ambiguity about the types of surgery that should be performed and funded in Australia.

Consideration be given to specifying in MBS item 30518 the type of gastrectomy operation and the surgical indication for this.

Potentially, MBS item number 30518 can be used to claim for different methods of ‘partial gastrectomy’ and different indications including not only treatment of obesity, but also treatment of stomach cancer and peptic ulcers. SG is an emerging surgery for obesity, for which long-term comparative safety and efficacy data are not available.

Consideration should be given to whether there is sufficient clinical evidence (as presented in this report) to justify the funding of SG for the treatment of obesity in Australia. Further, there is an absence of published economic data reporting the cost effectiveness of SG relative to current surgical practice (e.g. bypass or banding). Although level of evidence may not be strong, decision to fund SG under the MBS should be balanced with the importance for the Department to support new technology in the light of emerging evidence.

If SG is to be funded through the MBS for the treatment of obesity, a unique MBS item number should be established or MBS item 30518 respecified to refer specifically to for obesity purposes only, rather than ‘partial gastrectomy’ with no restriction on surgical indication. If not, MBS item 30518 should clearly indicate the surgical indication(s) for partial gastrectomy to exclude the potential for funding SG when there is currently insufficient clinical and economic support.

Consideration be given to splitting all relevant MBS items for obesity surgery (currently items 30511 and 30512) into separate items for laparoscopic and open procedures.

The findings of this review suggest that, where possible, laparoscopic surgery should ideally be performed rather than open surgery. Laparoscopic approaches to gastric banding and bypass are demonstrated in a range of studies to be as effective, safer, and more cost effective than open versions of the same procedure. Laparoscopic operations achieve the same result with a lower risk of complications and shorter inpatient stay (lower costs). For various reasons, it is not possible to perform laparoscopic operations on all patients, and MBS funding for open surgery should naturally be continued.

Currently most bariatric procedures are performed laparoscopically. Hence, consideration be given to separating out the current MBS items into distinct item numbers for

86



laparoscopic and open versions of each procedure. The purpose of separating out items into laparoscopic versus open is not to attach a higher benefit to the laparoscopic item numbers but, rather, to ensure consistency with other MBS items. A consideration is that a proportion of banding and bypass surgeries for obesity are converted from laparoscopic to open approaches during the procedure. Consideration should therefore be given as to whether the descriptions of the proposed MBS items should specify whether the approach (laparoscopic or open) corresponding to that item number is the one performed or the one attempted.

Consideration be given to specifying in the MBS items for surgical treatment of obesity special considerations that must be met prior to offering such surgery in the adolescent population.

A range of MBS items on the current schedule (November 2010) specify treatment of adults only, including investigations for sleep apnoea, botolinum toxin injection, and pyloroplasty. Whilst the clinical support for bariatric surgery for obesity is relatively strong for the adult population, comparatively few high-quality studies have been undertaken in adolescents.

Any special restrictions for adolescent populations in the MBS items should be consistent with the key recommendations of clinical practice guidelines developed in consultation with local clinical experts.

The indication of ‘morbid obesity’ should be redefined for MBS items relating to bariatric surgery for obesity, and the terminology changed.

The current MBS (November 2010) defines the indications for items 30511, 30512 and 30514 as ‘morbid obesity’. However, there appears to be some ambiguity as to terminology (“clinically severe obesity” is preferred) and to what constitutes morbid obesity in terms of BMI levels and comorbidities. Further, patients classified as super obese (BMI > 50) also benefit from gastric bypass surgery, which can lead to large and rapid weight loss.

Consideration should be given to redefining ‘morbid obesity’ in the MBS item descriptions using a combination of BMI range and/or presence/absence of specific comorbidities.

The greatest volume of clinical support for the resolution of comorbidities with surgery, and cost-effectiveness of surgery, is for patients with T2DM and hypertension. There are fewer clinical data for resolution of other comorbidities such as osteoarthritis or obesity-related cancers.

There should be periodic reviews of the long term efficacy of emerging surgical techniques such as SG, and the long-term efficacy and cost effectiveness of other forms of bariatric surgery (including an analysis of reoperation and band adjustment rates for gastric banding).

There is limited long-term clinical data on gastric banding and as such, evidence for gastric banding should be reviewed periodically. In particular, it should be considered whether in the future the relevant MBS items require:

87



a change in benefit levels to reflect the long term costs of band adjustments, reoperation etc;

modification to the MBS item descriptions, to reflect the most appropriate methods and populations for these surgeries;

addition of a separate MBS item for post-bariatric surgery follow-up and management; and

addition/removal from the MBS to reflect the findings of future clinical studies.

MBS item 14215 (gastric band adjustments) and item 30514 (surgical reversal of procedures for morbid obesity) should be retained on the MBS as long as the surgeries to which they apply continue to be funded through the MBS. This is important for ensuring the continued efficacy of these procedures. However, benefit levels for these items should potentially be revisited. In particular, if gastroplasty ceased to be funded via the MBS, the benefit level and descriptor for MBS item 30514 should not incorporate reversal of gastroplasty.

9.2 Conclusion regarding the adolescent patient population

Based on the evidence presented in this review: Bariatric surgery in severely obese adolescents should only be considered within the

context of an ongoing and coordinated multidisciplinary approach. Surgeons performing bariatric surgery on adolescents should be experienced,

credentialed for bariatric surgery and affiliated with a team experienced in the long-term follow-up and management of the metabolic and psychosocial needs of the adolescent bariatric patient and family. The institution should be one that is either participating in a study of the outcomes of bariatric surgery, or sharing data.

LAGB is the bariatric surgery of choice for adolescent patients because of its relative safety and its potential reversibility.

SG should only be considered investigational and BPD and DS procedures should not be recommended in adolescents

Adolescent patients should be followed-up on a 4-6 weekly basis post-surgery with long-term follow-up extended beyond 10 years, and ideally for the whole of life.

9.3 General/other recommendations Bariatric surgery is not recommended for:

children under the age of 14 years; pregnant or breastfeeding adolescents and adults; patients with significant cognitive disabilities; patients with untreated or untreatable psychiatric or psychological disorder; or patients with Prader-Willi syndrome or other similar hyperphagic conditions.

Adults over 60 years of age should be assessed on a case by case basis and the objective of surgery should be to improve their quality of life.

88



Bariatric surgery may be considered as a first-line option (instead of lifestyle interventions or pharmacotherapy) for adults with a BMI over 50 kg/m2 in whom surgical intervention is considered appropriate.

Morbidity and mortality rates are increased in patients with a pre-operative BMI >65 kg/m2 undergoing BPD-DS, in these patients, staged bariatric surgical procedures may be an option.

Bariatric operations should ideally be performed by a surgeon who has substantial experience, perform bariatric surgeries frequently (50-100 cases per year), operating in properly equipped, high volume weight loss centres (100 cases per year) with integrated and multidisciplinary treatment, as there is a steep learning curve associated with bariatric surgery and this reduces operative mortality.

AGB, RYGB and BPD are all effective in treating clinically severe obesity, but differ in the degree of weight loss and range of complications. The choice of procedure should be tailored to the individual’s situation weighing necessary outcomes versus tolerance or risk and lifestyle change.

Gastric bypass and BPD (with or without DS) should be reserved for heavier patients because of the potential for metabolic complications related to malabsorption.

Bariatric surgery is not uniformly a ‘low-risk’ procedure, and judicious patient selection and diligent peri-operative care are critical (Mechanick et al 2008).

Registries should be considered to monitor safety and ensure good practice. Patient selection criteria should be updated regularly to reflect new technologies and

ongoing refinement in surgical techniques. Recommendations should be developed on anaesthesia and intensive care for obese

subjects. Medical imaging, lifting and transport equipment as well as beds should be adapted

for patients whose corpulence is incompatible with standard models. Further consideration should be given to the merits of allowing suitable trained and

qualified staff, such as practice nurses, nurse practitioners, physician assistants and residents, to adjust gastric bands under the supervision of a medical practitioner. The medical practitioner under whose supervision the adjustment is provided would retain responsibility for the health, safety and clinical outcomes of the patient.

89



ReferencesAarts EO, Janssen IMC, Berends FJ 2011, ‘The gastric sleeve: losing weight as fast as

micronutrients?’, Obesity Surgery, 21(2): 207-211.

Access Economics (2008) The growing cost of obesity in 2008: three years on, Report for Diabetes Australia, August. www.accesseconomics.com.au/publicationsreports/showreport.php?id=172&searchfor=2008&searchby=year, accessed 20 August 2010.

Ackroyd R, Mouiel J, Chevallier JM, Daoud F 2006, ‘Cost-effectiveness and budget impact of obesity surgery in patients with type-2 diabetes in three European countries’, Obesity Surgery, 16(11): 1488-1503.

Ananthapavan J, Moodie M, Haby M, Carter R 2010, ‘Assessing cost-effectiveness in obesity: laparoscopic adjustable gastric banding for severely obese adolescents’, Surgery for Obesity and Related Disorders, 6(4): 377-385.

Anselmino M, Bammer T, Fernandez Cebrian JM, Daoud F, Romagnoli G, Torres A 2009, ‘Cost-effectiveness and budget impact of obesity surgery in patients with type 2 diabetes in three European countries (II)’, Obesity Surgery 19 (11): 1542-1549.

Apovian CM, Baker C, Ludwig D, Hoppin AG, Hsu G, Lenders C et al 2005, ‘Best practice guidelines in pediatric/adolescent weight loss surgery’, Obesity Research, 13: 274-282.

Aucott L, Poobalan A, Smith WCS, Avenell A, Jung R, Broom J 2005, ‘Effects of weight loss in overweight/obese individuals and long-term hypertension outcomes: a systematic review’, Hypertension, 45: 1035-1041.

August GP, Caprio S, Fennoy I, Freemark M, Kaufman FR, Lustig RH et al 2008, ‘Prevention and treatment of pediatric obesity: an endocrine society clinical practice guideline based on expert opinion’, Journal of Clinical Endocrinology & Metabolism, 93: 4576-4599.

Australian Institute of Health and Welfare (2006) Australia’s Health 2006, AIHW.

Bales CW, Buhr GT 2009, ‘Body mass trajectory, energy balance, and weight loss as determinants of health and mortality in older adults’, Obesity Facts, 2(3), 171-178.

Batsis JA, Sarr MG, Collazo-Clavell ML, Thomas RJ, Romero-Corral A, Somers VK et al 2008, ‘Cardiovascular risk after bariatric surgery for obesity’, American Journal of Cardiology, 102: 930-937.

Baur LA, Fitzgerald DA 2010, ‘Recommendations for bariatric surgery in adolescents in Australia and New Zealand’, Journal of Paediatrics and Child Health; 46(12):704-707.

Begg S, Vos T, Barker B, Stevenson C, Stanley L, Lopez AD (2007) The burden of disease and injury in Australia in 2003, AIHW

90

http://www.accesseconomics.com.au/publicationsreports/showreport.php?id=172&searchfor=2008&searchby=year

http://www.accesseconomics.com.au/publicationsreports/showreport.php?id=172&searchfor=2008&searchby=year



http://www.aihw.gov.au/publications/hwe/bodaiia03/bodaiia03-c01.pdf, accessed 20 August 2010.

Belle SH, Berk PD, Courcoulas AP, Flum Dr, Miles CW, Mitchell JE et al 2007, ‘Safety and efficacy of bariatric surgery: Longitudinal Assessment of Bariatric Surgery’, Surgery for Obesity and Related Diseases, 3(2): 116-26.

Booth M, Okely AD, Denney-Wilson E, Yang B, Hardy L, Dobbins T (2006) NSW schools physical activity and nutrition survey (SPANS) 2004, NSW Department of Health, Sydney.

Booth M, Chey T, Wake M, Norton K, Hesketh K, Dollman J, Robertson I (2003) ‘Change in the prevalence of overweight and obesity among young Australians, 1969-1997.’ American Journal of Clinical Nutrition, 77: 29-36.

Brethauer SA, Hammel JP, Schauer PR 2009, ‘Systematic review of sleeve gastrectomy as staging and primary bariatric procedure’, Surgery for Obesity and Related Diseases, 5: 469-475.

Buchwald, H 2005, ‘Bariatric surgery for morbid obesity: health implications for patients, health professionals, and third-party payers’, Journal of the American College of Surgeons, 200(4): 593-604.

Buchwald H, Avidor Y, Braunwald E, Jensen M, Pories W, Fahrbach K, Schoelles K 2004, ‘Bariatric surgery: a systematic review and meta-analysis’, Journal of the American Medical Association, 292(14): 1724-1737.

Buchwald H, Estok R, Fahrbach K, Banel D, Sledge I 2007, ‘Trends in mortality in bariatric surgery: a systematic review and meta-analysis’, Surgery, 142:621-35.

Buchwald H, Estok R, Fahrback K, Banel D, Jensen M, Pories W 2009, ‘Weight and type 2 diabetes after bariatric surgery: systematic review and meta-analysis’, The American Journal of Medicine, 122: 248-256.

Campbell J, McGarry LJ, Thompson D, Gilmore K, Hill G, Lee J, et al 2008 (poster abstract) ‘Cost-effectiveness of laparoscopic adjustable gastric banding and laparoscopic Roux-en-Y gastric bypass in the treatment of morbid obesity’, Value in Health, 11(3): A159.

Campbell J, McGarry LJ, Shikora SA, Hale BC, Lee JT, Weinstein MC 2010 ‘Cost-effectiveness of laparoscopic gastric banding and bypass for morbid obesity’, American Journal of Managed Care, 16(7): e174-e187.

Centers for Disease Control (2000) ‘CDC Growth Charts: United States’ Advance Data, number 314, http://www.cdc.gov/nchs/data/ad/ad314.pdf, accessed 6 September 2010.

Chapman AE, Kiroff G, Game P, Foster B, O’Brien P, Ham J et al 2004, ‘Laparoscopic adjustable gastric banding in the treatment of obesity: a systematic literature review’, Surgery, 135: 326-51.

91

http://www.cdc.gov/nchs/data/ad/ad314.pdf

http://www.aihw.gov.au/publications/hwe/bodaiia03/bodaiia03-c01.pdf



Chaston TB, Dixon JB, O’Brien PE 2007, ‘Changes in fat-free mass during significant weight loss: a systematic review’, International Journal of Obesity, 31: 743-750.

Chavez-Tapia NC, Tellez-Avila FI, Barrientos-Gutierrez T, Mendez-Sanchez N, Lizardi-Cervera J, Uribe M 2010, ‘Bariatric surgery for non-alcoholic steatohepatitis in obese patients’, Cochrane Database of Systematic Reviews, Issue 1. Art. No.: CD007340.

Clegg AJ, Colquitt J, Sidhu MK, Royle P, Loveman E, Walker A 2002, ‘The clinical effectiveness and cost-effectiveness of surgery for people with morbid obesity: A systematic review and economic evaluation’, Health Technology Assessment, 6(12).

Clegg A, Colquitt J, Sidhu M, Royle P, Walker A 2003, ‘Clinical and cost effectiveness of surgery for morbid obesity: a systematic review and economic evaluation’, International Journal of Obesity, 27: 1167-1177.

Cole TJ, Bellizzi MC, Flegal KM, Dietz WH 2000. ‘Establishing a standard definition for childhood overweight and obesity worldwide: international survey’, British Medical Journal, 320: 1240–1243.

Colquitt JL, Picot J, Loveman E, Clegg AJ 2009, ‘Surgery for obesity’, Cochrane Database of Systematic Reviews, Issue 2. Art. No.: CD003641.

Compher CW, Badellino KO, Boullata JI 2008, ‘Vitamin D and the bariatric surgical patient: a review’, Obesity Surgery, 18: 220-224.

Craig BM and Tseng DS 2002, ‘Cost-effectiveness of gastric bypass for severe obesity’ American Journal of Medicine, 113(6): 491-498.

De Groot NM, Burgerhart JS, Van De Meeberg PC, De Vries DR, Smout AJPM, Siersema PD 2009, ‘Systematic review: the effects of conservative and surgical treatment for obesity on gastro-oesophageal reflux disease’, Alimentary Pharmacology and Therapeutics, 30: 1091-1102.

DeMaria EJ, Pate V, Warthen M, Winegar DA 2010, ‘Baseline data from American Society for Metabolic and Bariatric Surgery – designated Bariatric Surgery Centers of Excellence using the Bariatric Outcomes Longitudinal Database’, Surgery for Obesity and Related Diseases, 6: 347-355.

Denney-Wilson, E, Booth ML, Baur, LA (2003), ‘Development of the Australian standard definition of child/adolescent overweight and obesity’ Nutrition and Dietetics, 60:74-77.

Department of Health and Ageing (2003a), Clinical practice guidelines for the management of overweight and obesity in adults, National Health and Medical Research Council (NHMRC) http://www.health.gov.au/internet/main/publishing.nsf/Content/obesityguidelines-guidelines-adults.htm, accessed 2 July 2010.

- (2003b), Clinical practice guidelines for the management of overweight and obesity in children and adolescents, National Health and Medical Research Council (NHMRC)

92

http://www.health.gov.au/internet/main/publishing.nsf/Content/obesityguidelines-guidelines-adults.htm

http://www.health.gov.au/internet/main/publishing.nsf/Content/obesityguidelines-guidelines-adults.htm



http://www.health.gov.au/internet/main/publishing.nsf/Content/893169B10DD846FCCA256F190003BADA/$File/children.pdf, accessed 6 September 2010.

Douketis JD, Macie C, Thabane L, Williamson DF 2005, ‘Systematic review of long-term weight loss studies in obese adults: clinical significance and applicability to clinical practice’, International Journal of Obesity, 29:1153-1167.

Farrell TM, Haggerty SP, Overby DW, Kohn GP, Richardson WS, Fanelli RD 2009, ‘Clinical application of laparoscopic bariatric surgery: an evidence-based review’, Surgical Endoscopy, 23:930-949.

Ferchak CV, Meneghini LF 2004, ‘Obesity, bariatric surgery and type 2 diabetes – a systematic review’, Diabetes/Metabolism Research and Reviews, 20: 438-445.

Flum DR 2005, ‘Cost and survival outcomes assessment of gastric bypass surgery’, Surgery for Obesity and Related Diseases, 1(1): 41-44.

Flum DR, Belle SH, King WC, Wahed AS, Berk P, Chapman W et al 2009, ‘Perioperative safety in the Longitudinal Assessment of Bariatric Surgery’, The New England Journal of Medicine, 361(5): 445-54.

Fried M, Kasalicky M, Melechovsky D, Kormanova K 2002, ‘Current status of non-adjustable gastric banding’, Obesity Surgery, 12: 395-398.

Fried M, Hainer V, Basdevant A, Buchwald H, Deitel M, Finer N et al 2007, ‘Interdisciplinary European guidelines for surgery for severe (morbid) obesity’, Obesity Surgery, 17: 260-270.

Garb J, Welch G, Zagarins S, Kuhn J, Romanelli J 2009, ‘Bariatric surgery for the treatment of morbid obesity: a meta-analysis of weight loss outcomes for laparoscopic adjustable gastric banding and laparoscopic gastric bypass’, Obesity Surgery, 19: 1447-1455.

Gentileschi P, Kini S, Catarci M, Gagner M 2002, ‘Evidence-based medicine: open and laparoscopic bariatric surgery’, Surgical Endoscopy, 16: 736-744.

Glenny, A, O’Meara S 1997, ‘Systematic review of interventions in the treatment of prevention of obesity’, NHS Centre for Reviews and Dissemination, University of York, CRD Report 10.

Guelinckx I, Devlieger R, Vansant G 2009, ‘Reproductive outcome after bariatric surgery: a critical review’, Human Reproduction Update, 15(2): 189-201.

Gumbs AA, Pomp A, Gagner M 2007, ‘Revisional bariatric surgery for inadequate weight loss’, Obesity Surgery, 17: 1137-1145.

Guo B, Harstall C 2005, ‘Laparoscopic adjustable gastric banding for the treatment of clinically severe (morbid) obesity in adults: an update’, Alberta Heritage Foundation for Medical Research, Health Technology Assessment Unit, Information Paper No. 26.

Hauri P, Horber FF, Sendi P 1999, ‘Is bariatric surgery with its costs?’, Obesity Surgery, 9: 480-483.

93

http://www.health.gov.au/internet/main/publishing.nsf/Content/893169B10DD846FCCA256F190003BADA/$File/children.pdf

http://www.health.gov.au/internet/main/publishing.nsf/Content/893169B10DD846FCCA256F190003BADA/$File/children.pdf



Herpertz S, Kielmann R, Wolf AM, Langkafel M, Senf W, Hebebrand J 2003, ‘Does obesity surgery improve psychosocial functioning? A systematic review’, International Journal of Obesity, 27: 1300-1314.

Hoerger TJ, Zhang P, Segel JE, Kahn HS, Barker LE, Couper S 2010, ‘Cost-effectiveness of bariatric surgery for severely obese adults with diabetes’, Diabetes Care, 33(9): 1933-1939.

Jen HC, Rickard DG, Shew SP, Maggard MA, Slusser WM, Dutson EP et al 2010, ‘Trends and outcomes of adolescent bariatric surgery in California 2005-07’, Pediatrics, 126(4):746-753.

Iannelli A, Facchiano E, Gugenheim J 2006, ‘Internal hernia after laparoscopic Roux-en-Y gastric bypass for morbid obesity’, Obesity Surgery, 16: 1265-1271.

Ikramuddin S, Klingman CD, Swan T, Minshall ME 2009, ‘Cost-effectiveness of Roux-en-Y gastric bypass in type 2 diabetes patients’, American Journal of Managed Care, 15(9): 607-615.

International Diabetes Institute (2001) ‘Diabesity and Associated Disorders in Australia – 2000: The Accelerating Epidemic’ The Australian Diabetes, Obesity and Lifestyle Study (AusDiab), Melbourne.

Jensen C and Flum DR 2005, ‘The costs of nonsurgical and surgical weight loss interventions: Is an ounce of prevention really worth a pound of cure?‘, Surgery for Obesity and Related Diseases, 1(3): 353-357.

Keating CL, Dixon JB, Moodie ML, Peeters A, Playfair J, O’Brien PE 2009a, ‘Cost-efficacy of surgically induced weight loss for the management of type 2 diabetes’, Diabetes Care, 32(4): 580-584.

Keating CL, Dixon JB, Moodie ML, Peeters A, Bulfone L, Maglianno DJ et al 2009b ‘Cost-effectiveness of surgically induced weight loss for the management of type 2 diabetes: Modeled lifetime analysis’, Diabetes Care, 32(4): 567-574.

Kelly J, Tarnoff M, Shikora S, Thayer B, Jones DB, Forse RA et al 2005, ‘Best practice recommendations for surgical care in weight loss surgery’, Obesity Research, 13: 227-233.

Kelly J, Shikora S, Jones DB, Hutter MH, Robinson MK, Romanelli J, et al 2009, ‘Best practice updates for surgical care in weight loss surgery’, Obesity, 17: 863-870.

Larsen SH, Wagner G, Heitmann BL 2007, ‘Sexual function and obesity’, International Journal of Obesity, 31: 1189-1198.

Laville M, Romon M, Chavrier G, Guy-Grand B, Krempf M, Chevallier JM, et al 2005, ‘Recommendations regarding obesity surgery’, Obesity Surgery, 15: 1476-1480.

Lee WJ, Chong K, Ser KH, Lee YC, Chen SC, Chen JC et al 2011, ‘Gastric bypass vs sleeve gastrectomy for type 2 diabetes mellitus’, Archives of Surgery, 146(2): 143-148.

94



Levy P, Fried M, Santini F, Finer N 2007, ‘The comparative effects of bariatric surgery on weight and type 2 diabetes’, Obesity Surgery, 17: 1248-1256.

Maggard MA, Shugarman LR, Suttorp M, Maglione M, Sugarman HJ, Livingston EH et al 2005, ‘Meta-analysis: surgical treatment of obesity’, Annals of Internal Medicine, 142:547-559.

Maggard MA, Yermilov I, Li Z, Maglione M, Newberry S, Suttorp M et al 2008, ‘Pregnancy and fertility following bariatric surgery: a systematic review’, Journal of the American Medical Association, 19: 2286-2296.

Maklin S, Malmivaara A, Victorzon M, Koivukangas V, Linna M, Sintonen H 2009, (poster abstract) ‘Cost-utility of bariatric surgery in the treatment for morbid obesity in Finland’, Value in Health, 12(7): A381.

Manterola C, Pineda V, Vial M, Losada H, Munoz S 2005, ‘Surgery for morbid obesity: selection of operation based on evidence from literature review’, Obesity Surgery, 15: 106-113.

McEwen LN, Coelho RB, Baumann LM, Bilik D, Nota-Kirby B, Herman WH 2010, ‘The cost, quality of life Impact, and cost–utility of bariatric surgery in a managed care population’, Obesity Surgery, 20(7): 919–928.

McTigue KM, Harris R, Hemphill B, Lux L, Sutton S, Bunton A et al 2003, ‘Screening and interventions for obesity in adults: summary of the evidence for the U.S. Preventive Services Task Force’, Annals of Internal Medicine, 139: 933-949.

Mechanick JI, Kushner RF, Sugerman HJ, Gonzalez-Campoy JM, Collazo-Clavell ML, Guven S et al 2008, ‘American Association of Clinical Endocrinologist, The Obesity Society, and American Society for Metabolic and Bariatric Surgery medical guidelines for clinical practice for the perioperative nutritional, metabolic, and nonsurgical support of the bariatric surgery patient’, Endocrine Practice, 14(S1): 1-83.

Medical Services Advisory Committee (MSAC) 2003, ‘Laparoscopic adjustable gastric banding for morbid obesity’, Canberra: Medical Services Advisory Committee 2003.

Minshall M, Swan T, Slusarek B, Ikramuddin S 2009a, (poster abstract) ‘Cost-effectiveness of the Roux-en-Y gastric bypass surgery compared with medical management for treatment of Type 2 Diabetes Mellitus (T2DM) in Spain, Italy, and Sweden’, Obesity Surgery, 19 (8): 1003-1004.

Minshall M, Swan T, Slusarek B, Ikramuddin S 2009b, (poster abstract) ‘Cost-effectiveness of the Roux-en-Y gastric bypass surgery compared with medical management for treatment of Type 2 Diabetes Mellitus (T2DM) in The UK, France, and Germany’, Obesity Surgery, 19(8): 1003.

Monk Jr, Nagib ND, Stehr W 2004, ‘Pharmaceutical savings after gastric bypass surgery’, Obesity Surgery, 14 (1): 13-15.

Narbro K, Agren G, Jonsson E, Naslund I, Sjostrom L, Peltonen M 2002, ‘Pharmaceutical costs in obese individuals: comparison with a randomly selected population sample

95



and long-term changes after conventional and surgical treatment - The SOS Intervention Study’, Archives of Internal Medicine, 162: 2061-2069.

National Health and Medical Research Council (2009) NHMRC levels of evidence and grades for recommendations for developers of guidelines, National Health and Medical Research Council (NHMRC), http://www.nhmrc.gov.au/_files_nhmrc/file/guidelines/evidence_statement_form.pdf, accessed 12 December 2010.

National Institutes of Health 1996, ‘Gastrointestinal surgery for severe obesity: NIH Consensus Development Conference March 25-27 1991’, Nutrition, 12(6): 397-402.

National Institutes of Health 1998, ‘Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: the evidence report’, National Heart, Lung, and Blood Institute in cooperation with the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health Publication, No. 98-4083, http://www.nhlbi.nih.gov/guidelines/obesity/ob_gdlns.htm, accessed 2 November 2010, accessed 2 November 2010.

National Institute for Health and Clinical Excellence (NICE) 2006, ‘Obesity – guidance on the prevention, identification, assessment and management of overweight and obesity in adults and children’, NICE clinical guideline 43.

Navaneethan SD, Yehnert H, Moustarah F, Schreiber MJ, Schauer PR, Beddhu S 2009, ‘Weight loss interventions in chronic kidney disease: a systematic review and meta-analysis’, Clinical Journal of the American Journal of Nephrology, 4:1565-1574.

Nguyen NT, Hinojosa M, Fayad, C, Varela E, Wilson SE 2007, ‘Use and outcomes of laparoscopic versus open gastric bypass at academic medical centers’, Journal of the American College of Surgeons; 205 (2): 248-255.

O’Brien PE, Sawyer SM, Laurie C, Brown WA, Skinner S, Veit F et al 2010, ‘Laparoscopic adjustable gastric banding in severely obese adolescents’, Journal of the American Medical Association, 303(6): 519-526.

O’Mahony R, Szomstein S, Rosenthal R 2006, ‘Laparoscopic gastric bypass for morbid obesity’, European Surgical Research, 38/6: 399-404.

Oude LH, Baur L, Jansen H, Shrewsbury VA, O’Malley C, Stolk RP, Summerbell CD 2009, ‘Interventions for treating obesity in children (review)’, Cochrane Database of Systematic Reviews, Issue 1, Art. No.: CD001872.

Padwal R, Brocks D, Sharma AM 2010, ‘A systematic review of drug absorption following bariatric surgery and its theoretical implications’, Obesity reviews, 11: 41-50.

Paxton JH and Matthews JB 2005 ‘The cost effectiveness of laparoscopic versus open gastric bypass surgery’, Obesity Surgery, 15(1): 24-34.

Picot J, Jones J, Colquitt JL, Gospodarevskaya E, Loveman E, Baxter L et al 2009, ‘The clinical effectiveness and cost-effectiveness of bariatric (weight loss) surgery for obesity: A systematic review and economic evaluation’, Health Technology Assessment, 13(41).

96

http://www.nhlbi.nih.gov/guidelines/obesity/ob_gdlns.htm



Pratt JSA, Lenders CM, Dionne EA, Hoppin AG, Hsu GLK, Inge TH, et al 2009, ‘Best practice updates for pediatric/adolescent weight loss surgery’, Obesity, 17: 901-910.

Preventative Health Taskforce (2009) Australia: the healthiest country by 2020, http://www.health.gov.au/internet/preventativehealth/publishing.nsf/Content/E233F8695823F16CCA2574DD00818E64/$File/obesity-jul09.pdf, accessed 2 July 2010.

Public Health Resource Unit (2006) Critical Appraisal Skills Programme (CASP) - making sense of evidence: 10 questions to help you make sense of reviews, Public Health Resource Unit, http://calder.med.miami.edu/portals/ebmfiles/UM%20CASP%20Systematic%20Reviews%20Assessment%20Tool.pdf, accessed 12 December 2010

Rubino F, Kaplan LM, Schauer PR, Cummings DE 2010, ‘The diabetes surgery summit consensus conference: recommendations for the evaluation and use of gastrointestinal surgery to treat type 2 diabetes mellitus’, Annals of Surgery, 251(3): 399-405.

Salem L, Devlin A, Sullivan SD, Flum DR 2008, ‘Cost-effectiveness analysis of laparoscopic gastric bypass, adjustable gastric banding, and nonoperative weight loss interventions’, Surgery for Obesity and Related Diseases, 4(1): 26-32.

Sauerland S, Angrisani L, Belachew M, Chevallier JM, Favretti F, Finer N et al 2005, ‘Obesity surgery: evidence-based guidelines of the European Association for Endoscopic Surgery (E.A.E.S.)’, Surgical Endoscopy, 19: 200-221.

Schneider WL 2000, ‘Laparoscopic adjustable gastric banding for clinically severe (morbid) obesity’, Alberta Heritage Foundation for Medical Research, Canada, HTA 7: Series B.

Scottish Intercollegiate Guidelines Network (2008) SIGN 50. A guideline developers’ handbook. Methodology checklist 1: Systematic reviews and meta-analyses, Scottish Intercollegiate Guidelines Network, http://www.sign.ac.uk/guidelines/fulltext/50/checklist1.html, accessed 12 December 2010

Shekelle PG, Morton SC, Maglione MA, Suttorp M, Tu W, Li Z et al 2004, ‘Pharmacological and surgical treatment of obesity’, Evidence Report/Technology Assessment No. 103.

Shi X, Karmali S, Sharma AM, Birch DW 2010, ‘A review of laparoscopic sleeve gastrectomy for morbid obesity’, Obesity Surgery, 20: 1171-1177.

Siddiqui A, Livingston E, Huerta S 2006, ‘A comparison of open and laparoscopic Roux-en-Y gastric bypass surgery for morbid and super obesity: A decision-analysis model’, American Journal of Surgery, 192(5): e1-e7.

Sjostrom L, Narbro K, Sjostrom D, Karason K, Larsson B, Wedel H et al 2007, ‘Effects of bariatric surgery on mortality in Swedish Obese Subjects’, The New England Journal of Medicine, 357(8): 741-52.

97

http://www.sign.ac.uk/guidelines/fulltext/50/checklist1.html

http://calder.med.miami.edu/portals/ebmfiles/UM%20CASP%20Systematic%20Reviews%20Assessment%20Tool.pdf

http://calder.med.miami.edu/portals/ebmfiles/UM%20CASP%20Systematic%20Reviews%20Assessment%20Tool.pdf

http://www.health.gov.au/internet/preventativehealth/publishing.nsf/Content/E233F8695823F16CCA2574DD00818E64/$File/obesity-jul09.pdf

http://www.health.gov.au/internet/preventativehealth/publishing.nsf/Content/E233F8695823F16CCA2574DD00818E64/$File/obesity-jul09.pdf



Snow V, Barry P, Fitterman N, Qaseem A, Weiss K 2005, ‘Pharmacologic and surgical management of obesity in primary care: a clinical practice guideline from the American College of Physicians’, Annals of Internal Medicine, 142(7): 525-532.

Society of American Gastrointestinal and Endoscopic Surgeons (SAGES) 2008, ‘SAGES guideline for clinical application of laparoscopic bariatric surgery’, Practice/Clinical Guidelines published on: 06/2008 by SAGES, http://www.sages.org/publication/id/30/ accessed 2 November 2010.

Sturm R (2007) ‘Increases in morbid obesity in the USA: 2000-2005’ Public Health, 121(7):492-496.

Tice JA, Karliner L, Walsh J, Petersen AJ, Feldman MD, ‘Gastric banding or bypass? A systematic review comparing the two most popular bariatric procedures’, The American Journal of Medicine, 121: 885-893.

Torgerson JS, Sjostrom L 2001, ‘The Swedish Obese Subjects (SOS) study – rationale and results’, International Journal of Obesity, 25(S1): S2-4.

Treadwell JR, Sun F, Schoelles K 2008, ‘Systematic review and meta-analysis of bariatric surgery for pediatric obesity’, Annals of Surgery, 248:763-776.

Tsiros MD, Sinn N, Coates AM, Howe PRC, Buckley JD 2008, ‘Treatment of adolescent overweight and obesity’, European Journal of Pediatrics, 167: 9-16.

United Nations 2003, ‘Handbook on reproductive health indicators’, United Nation Economic and Social Commission for Asia and the Pacific, http://www.unescap.org/esid/psis/publications/handbookhealth/handbook.pdf, accessed 17 November 2010.

US Department of Health and Human Services 2010, ‘Longitudinal Assessment of Bariatric Surgery (LABS)’, National Institutes of Health, http://win.niddk.nih.gov/publications/PDFs/LABSFactSheet(final).pdf accessed 14 December 2010.

van Gemert WG, Adang EMM, Kop M, Vos G, Greve JWM, Soeters PB 1999, ‘A prospective cost-effectiveness analysis of vertical banded gastroplasty for the treatment of morbid obesity’, Obesity Surgery, 9(5): 484-491.

van Mastrigt GAPG, van Dielen FMH, Severens JL, Voss GBWE, Greve JW 2006, ‘One-year cost-effectiveness of surgical treatment of morbid obesity: Vertical banded gastroplasty versus Lap-Band’, Obesity Surgery, 16(1): 75-84.

Walls HL, Wolfe R, Haby MM, Magliano DJ, de Courten M, Reid CM et al 2010, ‘Trends in BMI of urban Australian adults, 1980-2000’, Public Health Nutrition, 13(5): 631-638.

Wang RT, Koretz RL, Yee HF 2003, ‘Is weight reduction an effective therapy for nonalcoholic fatty liver? A systematic review’, American Journal of Medicine, 115: 554-559.

Wax JR, Pinette MG, Cartin A, Blackstone J 2007, ‘Female reproductive issues following bariatric surgery’, Obstetrical and Gynecological Survey 62(9): 595-604.

98

http://win.niddk.nih.gov/publications/PDFs/LABSFactSheet(final).pdf

http://www.unescap.org/esid/psis/publications/handbookhealth/handbook.pdf

http://www.sages.org/publication/id/30/



Whitlock EP, O’Connor EA, Williams SB, Beil TL, Lutz KW 2008, ‘Effectiveness of weight management programs in children and adolescents’, Agency for Healthcare Research and Quality, Evidence Report/Technology Assessment No. 170.

World Health Organization (2000) Obesity: preventing and managing the global epidemic: report of a WHO consultation, http://whqlibdoc.who.int/trs/WHO_TRS_894.pdf, accessed 2 July 2010.

- (2010) BMI classification, http://apps.who.int/bmi/index.jsp?introPage=intro_3.html, accessed 7 September 2010.

99

http://apps.who.int/bmi/index.jsp?introPage=intro_3.html

http://whqlibdoc.who.int/trs/WHO_TRS_894.pdf



Appendix A: Final literature review protocolSearch Strategy

Embase.com (Embase+MEDLINE) was searched for published articles using the search terms for the disease under evaluation – the search was not limited by date, but was limited to articles published in the English language. A summary of the search of Embase.com is presented in Table A.1.

Table A.1– Embase.com search, <1966 to 1 September 2010 (*)

Embase Session Results

No. Query Results#58 #57 AND [english]/lim 469#57 #16 OR #44 OR #56 516#56 #12 AND #55 358#55 #45 OR #46 OR #47 OR #48 OR #49 OR #50 OR #51 OR #52 OR #53 OR #54 670409#54 (clinical NEXT/2 (protocols OR protocol)):ab,ti 3228#53 'clinical pathway':ab,ti OR 'clinical pathways':ab,ti 1686#52 'best practice':ab,ti OR 'best practices':ab,ti 7101#51 guideline*:ab,ti OR consensus:ab,ti 238555#50 'evidence based practice'/de 10815#49 'consensus'/de 9832#48 'gold standard'/de 6019#47 'professional standard'/de 17550#46 'standard'/de 279111#45 'practice guideline'/exp 226445#44 #22 OR #39 OR #43 153#43 #12 AND #42 10#42 #40 OR #41 29459#41 (pooled NEXT/4 analys?s):ab,ti 29459#40 'pooled analysis':de 10#39 #12 AND #38 101#38 #32 OR #37 127513#37 #33 AND #36 97084#36 #34 OR #35 405017#35 qualitative:ti OR literature:ti OR evidence:ti OR 'evidence based':ti 247628#34 systematic:ti OR critical:ti OR methodologic:ti OR quantitative:ti 167131#33 synthesis:ti OR overview:ti OR review:ti OR survey:ti 599158#32 #28 AND #31 42908#31 #29 OR #30 1825619#30 consensus:ti OR literature:ti OR overview:ti 122324#29 review:it,ti OR guideline:it,ti OR guidelines:it,ti 1788094

100



No. Query Results#28 #23 OR #27 75469#27 #25 AND #26 65150#26 handsearch*:ab,ti OR search*:ab,ti 186312#25 #23 OR #24 542640

#24 hand:ab,ti OR manual:ab,ti OR electronic:ab,ti OR bibliograph*:ab,ti OR database:ab,ti OR databases:ab,ti 517044

#23 cochrane:ab,ti OR medline:ab,ti OR embase:ab,ti 48097#22 #12 AND #21 87#21 #17 OR #18 OR #19 OR #20 96178#20 cochrane:jt 10090

#19 quantitative*:ab OR systematic*:ab OR methodologic*:ab AND (review*:ab OR overview*:ab) 64056

#18 quantitative*:ti OR systematic*:ti OR methodologic*:ti AND (review*:ti OR overview*:ti) 18152

#17 'systematic review'/de 35818#16 #12 AND #15 51#15 #13 OR #14 65613

#14 'meta analysis':ab,ti OR 'meta analyses':ab,ti OR 'meta analytical':ab,ti OR metanaly*:ab,ti 36748

#13 'meta analysis'/de 49989#12 #3 AND #11 7192#11 #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 11888

#10

'stomach bypass':ab,ti OR 'gastric bypasses':ab,ti OR 'gastroileal bypass':ab,ti OR 'gastric reduction':ab,ti OR gastroplasty:ab,ti OR gastroplasties:ab,ti OR 'partial gastrectomy':ab,ti OR 'partial stomach resection':ab,ti OR 'stomach partial resection':ab,ti OR 'gastric bandings':ab,ti OR 'stomach banding':ab,ti OR 'gastric bands':ab,ti OR 'sleeve gastrectomies':ab,ti

3900

#9 'gastric bypass':de,ab,ti OR 'gastric band':de,ab,ti OR 'gastric banding':de,ab,ti OR 'sleeve gastrectomy':de,ab,ti 6741

#8 'banded gastroplasty':de OR 'band gastroplasty':de OR 'vertical gastroplasty':de 73

#7 'gastric banding'/de 2363#6 'partial gastrectomy'/de 950#5 'gastroplasty'/de 2423#4 'stomach bypass'/de 4903#3 #1 OR #2 209320

#2 obesity:ab,ti OR obese:ab,ti OR obesitas:ab,ti OR adipositas:ab,ti OR adiposity:ab,ti OR corpulency:ab,ti OR overweight:ab,ti OR 'adipose tissue hyperplasia':ab,ti OR 'fat overload syndrome':ab,ti

154078

#1 'obesity'/exp 175168* The search was conducted using Elsevier’s Embase Biomedical Answers Web site on 2 September 2010.

101



The Cochrane Library was searched for systematic reviews, controlled trials, economic evaluations and studies using the search terms for the disease under evaluation – the search was not limited by date and there were no database restrictions. A summary of the search of The Cochrane Library is presented in Table A.2 and Table A.3.

Table A.2– The Cochrane Library search, 2010 Issue 8 (*)

Current Search History

ID Search Hits

#1 MeSH descriptor Obesity explode all trees 5192

#2obesity OR obese OR obesitas OR adipositas OR adiposity OR corpulency OR overweight OR "adipose tissue hyperplasia" OR "fat overload syndrome"

9530

#3 (#1 OR #2) 9554

#4 MeSH descriptor Gastric Bypass , this term only 216

#5 MeSH descriptor Gastroplasty , this term only 175

#6 "banded gastroplasty" OR "band gastroplasty" OR "vertical gastroplasty" 64

#7 "gastric bypass" OR "gastric band" OR "gastric banding" OR "sleeve gastrectomy" 407

#8

"stomach bypass" OR "gastric bypasses" OR "Gastroileal Bypass" OR "gastric reduction" OR gastroplasty OR gastroplasties OR "partial gastrectomy" OR "partial stomach resection" OR "stomach partial resection" OR "gastric bandings" OR "stomach banding" OR "gastric bands" OR "sleeve gastrectomies"

321

#9 (#4 OR #5 OR #6 OR #7 OR #8) 560

#10 (#3 AND #9) 431

#11 Meta-Analysis:pt 423

#12 "meta analysis" OR "meta analyses" OR "meta analytical" or metanaly* 16469

#13 (#11 OR #12) 16469

#14 (#10 AND #13) 20

#15 Guideline:pt 28

#16 Practice Guideline:pt 20

#17 MeSH descriptor Consensus , this term only 32

#18 MeSH descriptor Evidence-Based Practice explode all trees 1489

#1 guideline* OR consensus 22899

102



9

#20 "best practice" or "best practices" 401

#21 "clinical pathway" OR "clinical pathways" 256

#22 clinical NEAR/2 (protocols OR protocol) 2413

#23 (#15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22) 26280

#24 (#10 AND #23) 49

#25 (#14 OR #24) 56

* The search was conducted using Wiley Interscience on 2 September 2009.

Table A.3– Breakdown of database retrieval from The Cochrane Library, 2010 Issue 8

Database Results

Cochrane Database of Systematic Reviews 13^Database of Abstracts of Reviews of Effects (DARE) 8Cochrane Central Register of Controlled Trials (CENTRAL) 5Cochrane Methodology Register (CMR) 0Health Technology Assessment Database (HTA) 9NHS Economic Evaluation Database (NHSEED) 21Cochrane Groups 0

Total 43

103



Appendix B: Overview of the studies identified for the clinical literature reviewA data extraction template was used to summarise the quality of and the key information in each of the included studies. Information concerning research question, study population, interventions compared, safety and efficacy outcomes and conclusions and recommendations were extracted by one of three reviewers. When uncertainty arises, data extracted was checked by another reviewer and differences in data extraction were resolved by consensus, referring back to the original article. Individual studies included in the meta-analysis and systematic literature review were assigned a level of evidence according to designations of ‘levels of evidence’ shown in Table B.1. The two key components of the body of evidence for each guideline reviewed were graded using the matrix shown in Table B.2. Evaluation of the quality of meta-analyses, systematic reviews and guidelines are detailed in Table B.3, Table B.4, and Table B.5 respectively, while data extracted from each of the study are listed in Table B.6, Table B.7, and Table B.8 respectively.

Table B.1: Evidence hierarchy

Level Intervention

I A systematic review of level II studiesII A randomised controlled trialIII-1 A pseudorandomised controlled trial (alternate allocation or some other method)III-2 A comparative study with concurrent controls:

Non-randomised experimental trial 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 study Interrupted time series without a parallel control group

IV Case series with either post-test or pre-test/post test outcomesSource: NHMRC 2009

104



Table B.2: Body of evidence matrix for assessing guidelines

Component A B C DExcellent Good Satisfactory Poor

Evidence base Several level I or II studies with low risk of bias

One or two level II studies with low risk of bias or a SR/multiple level III studies with low risk of bias

Level III studies with low risk of bias, or level I or II studies with moderate risk of bias

Level IV studies, or level I to III studies with high risk of bias

Consistency All studies consistent

Most studies consistent and inconsistency may be explained

Some inconsistency reflecting genuine uncertainty around clinical question

Evidence is inconsistent

Clinical impact Very large Substantial Moderate Slight or restricted

Generalisability Population/s studied in body of evidence are the same as the target population for the guideline

Population/s studied in body of evidence are similar to the target population for the guideline

Population/s studied in body of evidence differ to target population for guideline but it is clinically sensible to apply this evidence to target population

Population/s studied in body of evidence differ to target population and hard to judge whether it is sensible to generalise to target population

Applicability Directly applicable to Australian healthcare context

Applicable to Australian healthcare context with few caveats

Probably applicable to Australian healthcare context with some caveats

Not applicable to Australian healthcare context

Source: NHMRC 2009

Table B.3: Assessment of evidence base on meta-analyses

Buchwald et al, 2007 Number of studies 361Level of evidence (NHMRC) Review of level II to level IV evidence

II=4, III=166 and IV=191Studies were in most cases homogeneous

Did the reviewers try to identify all relevant studies?

1990 – 30 April 2006Medline, Current Contents, Cochrane Library databases, key word search and manual reference checks of accepted papers in recent reviews

Did the reviewers show that they assessed the quality of the studies?

Yes. All accepted studies were assigned a level of evidence (Centre for Evidence-Based Medicine, Oxford, UK), and RCTs were rated for quality using the Jadad scoring method

105



If the study results were statistically combined (i.e. meta-analysis), was it reasonable to do so?

Yes. Restricted, maximum-likelihood random effects meta-analyses were performedCochran Q statistic used to test heterogeneityStudies providing data from larger databases (e.g. SAGES) or large population studies were excluded because of the probability of overlap but are reviewed qualitatively in the discussion

Can the results be applied/generalised to the Australian population?

Yes. Most studies were performed in North America (57.6%) or Europe (34.7%) with Aus/NZ contributing almost 2,000 patients to report on banding. 13.5% were multicenter studies.

Buchwald et al, 2009 Number of studies 621Level of evidence (NHMRC) Review of level II to level IV evidence (II=10, III=312 and IV=296).

Data on weight loss and diabetes resolution were heterogeneous for the entire dataset but less heterogeneous in the pure diabetic population. Results across studies were substantially consistent.


1990 – 30 April 2006Medline, Current Contents, Cochrane Library databases. Reviewed PubMed for the prior 6 months with no limits and Current Contents for the prior year + manual reference checks of accepted papers in recent reviews within the last 2 years.


Yes. All accepted studies were assigned a level of evidence (Centre for Evidence-Based Medicine, Oxford, UK), and RCTs were rated for quality using the Jadad scoring method


Yes. Restricted, maximum-likelihood random effects meta-analyses were performedCochran Q statistic used to test heterogeneityResults across studies were substantially consistent in all buy the smallest studies, thus, meta-analytic pooling of the results were appropriate.


Yes. Most studies were performed in Europe (44%), North America (43%). 11% were multicenter studies.

Garb et al, 2009 Number of studies 28Level of evidence (NHMRC) Level II to level IVDid the reviewers try to identify all relevant studies?

An electronic search of English language articles for bariatric surgery was conducted using selected Medical Subject Headings terms to identify studies reporting weight loss outcomes for LAGB or LGB procedures. Databases used for the search included Medline, SCOPUS, Proquest, and the Cochrane Library Database of Systematic Reviews for the period 2003–2007.


Selected aspects of study quality such as patient follow-up rates and reporting of surgical variables for the procedures studied were examined. Authors noted that although problems in study quality raised significant concerns regarding the validity of current weight loss estimates in this area, there was no evidence of publication bias.

106




Meta-analyses were performed to examine mean %EWL outcomes separately for LAGB and LGB. Composite estimates of effect size and their 95% confidence intervals were computed separately for each type of surgery using a random-effects model. The Q-test was used to test for significant heterogeneity of effect size among studies within each surgery type and for significant differences between surgery types. A funnel plot of effect size versus sample size was calculated to assess publication bias.


Yes. Although country of studies were not reported, the setting for the studies was mostly academic hospital (78.6%), with the remainder community hospital and private practice (each 10.7%).

Shekelle et al, 2004 Number of studies 167 studies considered. A total of 142 studies were reviewed.Level of evidence (NHMRC) Level II to IV (28 RCT/CCTs, 1 observational study and 113 case

series)Did the reviewers try to identify all relevant studies?

Unclear. Searched Medline and Embase, scanned reference lists of recent extensive reviews and contacted experts in the field


Yes. Had a group of distinguished clinicians in the field to guide quality assessment


Yes. A pooled mean surgical weight loss for each procedure group was estimated using a random effects model, and an associated 95% confidence interval was constructed. However, because of heterogeneity, the authors did not feel meta-analysis was justified and summarised data narratively.


Yes.

Treadwell et al, 2008Number of studies 18Level of evidence (NHMRC) UnclearDid the reviewers try to identify all relevant studies?

Yes. Searched 15 databases including PubMed and EMBASE, bibliographies from identified studies, reviews and gray literature. Last search conducted on 31 December 2007


Yes. Overall stability and strength of the evidence for weight loss and comorbidity resolution after bariatric surgery were rated using a formal rating system which incorporates the quality, quantity, consistency, robustness of the evidence, as well as the magnitude of observed effects.


Yes. Meta-analyses of the mean change in BMI were conducted using the random-effects methods of DerSimonian and Laird. Heterogeneity was measured with the I2 statistic, with I2 ≥ 50% defining substantial heterogeneity. Double counting of patients (if there were multiple reports from same surgical centre) was accounted for.


Yes. 11 of 18 studies were conducted in the USA; the other 7 were conducted in Israel (two studies), Italy (two studies), Australia, Austria, or Saudi Arabia. 14 studies reported data from a single surgical centre, whereas the other 4 were from 2 or more surgical centres

107



Table B.4: Assessment of evidence base on systematic reviews

Brethauer et al, 2009 Number of studies 36Level of evidence (NHMRC) Level 1 + other evidence (2 RCT, 1 non-RCT, 33 uncontrolled case

series)Did the reviewers try to identify all relevant studies?

Identified prospective and retrospective series from 1996 to 31 January 2009. Searched PubMed and bibliographies of selected reports.


No. Large range in number of patients and duration of follow up


Yes. Pooled estimate calculated. Weights for the inverse variance weighted means determined by the sample size.Chi-square test and Fisher’s exact test for the pooled data across studies to compare the complication and mortality rates


Yes. Australian studies included and 3 were multicenter trials

Chapman et al, 2004 Number of studies 121 (64 Laparoscopic gastric banding + 57 VBG/RYGB)Level of evidence (NHMRC) Review of level II-IV studiesDid the reviewers try to identify all relevant studies?

Yes.Medline, Embase: 1988 – Aug 2001; Current Contents: 1993 – Aug 2001; HealthStar: 1988 – June 2001; and Cochrane Library: 2001 issue 2


Yes. The studies were tabulated and methodologically evaluated, including appropriateness of study exclusion criteria, quality of reporting, and possible confounding


N/A. No meta-analyses were performed


Yes. Australian paper – Australian safety and efficacy register of new interventional procedures-surgical (ASERNIP-S) review group

Colquitt et al, 2009 Number of studies 26 (3 RCTs and 3 prospective cohort studies compared surgery

with non-surgical management, and 20 RCTs compared different bariatric procedures)

Level of evidence (NHMRC) Review of level II evidence (+level III)

108




Yes. Resources searched:The Cochrane Library (Issue 3/2008);MEDLINE (until 29/7/2008);EMBASE (until 29/7/2008);PsychINFO (until 29/7/2008);CINAHL (until 16/7/2008);Science and Social Sciences Citation Index (until 29/7/2008);British Nursing Index (until 6/8/2008).Databases of grey literatureWeb of Science Proceedings (until 29/7/2008);BIOSIS (until 5/8/2008);AMED (until 5/8/2008).Ongoing trialsNational Research Register (until 30/7/2008);UKCRN (until 30/7/2008);Clinical Trials.gov (until 30/7/2008);Controlled Clinical Trials (until 30/7/2008);Australia NZ Clinical Trial Register (until 30/7/2008).Other resourcesContacted to obtain additional references, unpublished trial, and any ongoing trials.Handsearches of the following journals:• International Journal of Obesity (1977 to 2000, volume 24, part 12);• Obesity Research (1993 to 2001, volume 9, part 2);• Obesity Surgery (1991 to 2001, volume 11, part 2);• American Journal of Clinical Nutrition (1966 to 2000, volume 72, part 6);• Proceedings of the Nutrition Society (1960 to 2000, volume 59, part 4);• Journal of Human Nutrition and Dietetics (1988 to 2001,volume 14, part 1);• Journal of the American Dietetic Association (1980 to 1990, volume 90, part 12).


Yes. RCTs was assessed using criteria from the Cochrane Handbook for Systematic Reviews of InterventionsThe quality of controlled clinical (non-randomised) trials and prospective cohort studies was assessed using criteria from NHS CRD Report 4The minimum duration of follow-up for inclusion in this review was 12 months.


N/A. A meta-analysis was not appropriate. Synthesis of the included studies was through narrative review.


Yes.

De Groot et al, 2009 Number of studies 32 (diet/lifestyle = 7, RYGB = 8, LAGB = 12, VBG = 8)Level of evidence (NHMRC) Level II to level III evidence. Case reports and expert opinions

were excluded.

109




Unclear. Searched PubMed, the Cochrane Library and EMBASEBibliographies of retrieved articles manually searched


Yes. Reviews were assessed according to the Cochrane library definitions.

Primary outcome: GORD measured by 24h pH monitoring, manometry, endoscopy and/or radiological techniques. Reduction in reflux symptoms evaluated using questionnaires

Secondary outcome: Weight lossIf the study results were statistically combined (i.e. meta-analysis), was it reasonable to do so?

N/A. Methodological differences make comparison of results difficult/impossible. Particularly, definitions of obesity, GORD and weight loss differed among studies


Unclear. Possibly. The countries from which the paper originated were not described. Methodological differences between studies make comparison of results difficult. Confounding factors include different types of advice regarding PPI us and dietary and lifestyle habits.

Douketis et al, 2005 Number of studies 44Level of evidence (NHMRC) Review of level II and II studies (randomized controlled or

nonrandomized clinical trials)Did the reviewers try to identify all relevant studies?

Yes. MEDLINE (1966- Sept 2003),HealthSTAR (1975- Sept 2003), and the Cochrane Controlled Trials Register (1990- Sept 2003)


Yes. Study quality assessed using pre-specified grading

system adapted from the Process Manual for Producing and Disseminating CTF Reviews, Canadian Task Force on Preventive Health Care (May 2002 draft), and from the United States Preventive Services Task Force Methods Manual (August 2000 draft).

Study quality was rated as fair or good in dietary/lifestyle and pharmacologic studies, and fair or poor in surgical studies

Eligible studies excluded if <100 subjects, follow-up <2y (a 1 year follow-up allowed from pharmacologic studies)


No. Meta-analysis was not appropriate for dietary and surgical studies, because these studies differed in their interventions. However, results of studies investigating the efficacy of pharmacologic therapy compared with placebo was pooled using fixed and random effects models (in cases of significant heterogeneity, P<0.10) and a weighted summary estimate of treatment effects. A chi-squared test was used to assess heterogeneity.


Unclear. Weight loss studies have methodologic limitations that restrict the applicability of findings to unselected obese people assessed in everyday clinical practice. These limitations include an inadequate study duration, large proportions of subjects lost to follow-up, a lack of an appropriate usual care group, and a lack of reporting of outcomes in high-risk subgroups.

110



Farrell et al, 2009 Number of studies Not specifiedLevel of evidence (NHMRC) Level II through to level IVDid the reviewers try to identify all relevant studies?

A broad search of the English language literature was performed in late 2007 using both electronic and physical means. The electronic search used the PubMed and Cochrane Library databases. Manual reference checks of published review articles were performed to supplement the aforementioned electronic searches.


Yes. The articles were reviewed by the authors according to the protocol developed by the SAGES Guidelines Committee for internal use and graded as to level of evidence


Unclear. The study notes that a meta-analysis was undertaken, but does not provide any further information.


Yes. US authors

Gentileschi et al, 2002 Number of studies 94Level of evidence (NHMRC) Level II through to level IVDid the reviewers try to identify all relevant studies?

Yes. Medline, Embase, the Cochrane Library up to May 2001. Evaluated the references of each report and hand-searched Obesity Surgery, Obesity Research and International Journal of Obesity from 1991 to 2001


Yes. Level of evidence was assessed using the Oxford Center for Evidence-Based Medicine and the results from surgical therapy were evaluated according to Reinhold classification.


N/A


Unclear

Glenny and O’Meara Number of studies 97 papers (15 for surgical interventions)Level of evidence (NHMRC) Level II through to level IVDid the reviewers try to identify all relevant studies?

Yes. Electronic databases (including Medline, embase and psychLIT) were searched (1966 – Jan 1997), bibliographies of existing reviews were examined and experts in field were consulted.


Yes. The methodological quality of many studies was poor. Details of sample size, participant randomization etc.


N/A. Where possible graphs are included to illustrate the mean (or median changes in parameters from baseline to final analysis of each intervention, displayed with 95% CI where data allows. No pooled data, only a summary of information from individual trials.


Yes. Most research carried out in North America.

Manterola et al, 2005 Number of studies 31Level of evidence (NHMRC) Mainly level IV

111




No. Medline, LiLACS, Cochrane Jan 1990 to Dec 2002; no review of references or abstracts


Unclear. Inclusion criteria and analysis of methodological quality using ad-hoc designed score based on design, sample size and methodological aspect


N/A. Results were entered into a spreadsheet and descriptive statistics were applied (but not MA)


No.

Schneider, 2000Number of studies 9Level of evidence (NHMRC) Level II to level IV (RCT=1, prospective comparative studies=2

and clinical series=6)Did the reviewers try to identify all relevant studies?

Yes. Searched Medline, Embase, Best Evidence, HTA, EED, DARE, Cochrane, HealthSTAR, CMA Practice Guidelines, National Guideline Clearinghouse and indices from Obesity Surgery journal (1993-August 1999)


Yes. Used Jovell and Navarro-Rubio levels of scientific evidence to assess quality of studies


N/A. Not statistically combined


Yes.

Tice et al, 2008 Number of studies 14Level of evidence (NHMRC) Mainly level IIIDid the reviewers try to identify all relevant studies?

Yes. MEDLINE (1966 – Jan 2007), Cochrane clinical trials database, Cochrane reviews database, Google Scholar, EMBASE, Database of Abstracts of Reviews of Effects, and manual searches of bibliographies of systematic reviews and key articles


Yes. Quality assessed according to GRADE criteria, had to compare both surgery types


N/A. Not statistically combined


Unclear. In most of the studies, the two surgical groups were not comparable. LAGB studies were based in Europe while RYGB studies were based in US. Unable to determine whether the observed differences in outcomes reflect differences in the respective health care systems and patient populations, or true differences between the procedures.

Department of Health and Ageing (DoHA), 2003a (Australia)*Number of studies 36 surgical studiesLevel of evidence (NHMRC) Level II to IVDid the reviewers try to identify all relevant studies?

Yes. A systematic review of the scientific literature published in Medline and supplementation by materials provided by the Working Party and cross-referencing relevant articles. Methodology based on that used by the US NIH to develop its Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults.

112




Yes. The level of evidence and grades of recommendations are adapted from the National Health and Medical Research Council levels of evidence for clinical interventions and the US National Institutes of Health clinical guidelines.


N/A. Reported mean and range of weight loss 1-2 years after surgical treatment.


Yes

Department of Health and Ageing (DoHA), 2003b (Australia)*Number of studies Unclear. Around 5 studiesLevel of evidence (NHMRC) Level IVDid the reviewers try to identify all relevant studies?

Yes. A systematic review of the scientific literature published in Medline and supplementation by materials provided by the Working Party and cross-referencing relevant articles. Methodology based on that used by the US NIH to develop its Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults.


Yes. The level of evidence and grades of recommendations are adapted from the National Health and Medical Research Council levels of evidence for clinical interventions and the US National Institutes of Health clinical guidelines.


N/A


Yes

Kelly et al, 2005 (USA)*Number of studies 100+ papers identified, 26 most relevant were reviewed.

Criteria for assessing relevance not detailedLevel of evidence (NHMRC) Level I through to level IVDid the reviewers try to identify all relevant studies?

No. Medline, Cochrane Library Jan 1980 to Apr 2004


Yes. Quality of evidence based on models used by US Preventive Services Task Force and American Diabetes Association


Not combined


Unclear. Unclear how the relevant studies were identified.

Kelly et al, 2009 (USA)*Number of studies 135+ papers identified, 65 most relevant were reviewed.


Unclear. Medline, Embase, Cochrane Library Apr 2004 – Jul 2007


Yes. Quality was assessed formally according to a criteria


Not combined

113




Unclear. Unclear how the relevant studies were identified.

McTigue et al, 2003 (USA)*Number of studies UnclearLevel of evidence (NHMRC) Level I through to level IVDid the reviewers try to identify all relevant studies?

Unclear. UPSTF’s 1996 review, MEDLINE, Cochrane Library Jan 1994 – Feb 2003


Yes. Eligibility criteria, quality assessed using USP-STF criteria


Unclear. Mean weight loss calculated and there was mention of pooling of data, however, method not clearly specified.


Yes. Studies evaluated include systematic reviews from the U.S. National Institutes of Health (NIH), the Canadian Task Force on Preventive Health Care (CTFPHC), the University of York for the U.K. National Health Service (NHS), the U.S. National Task Force on the Prevention and Treatment of Obesity, and the British Medical Journal’s Clinical Evidence

Pratt et al, 2009 (USA)*Number of studies 1,085+ papers indentified, 186 most relevant were reviewed.


Unclear. Searched PubMed, MEDLINE, and Cochrane (April 2004 – May 2007)


Unclear. The system used to grade quality of evidence was described in another related, prior study


Not combined


Yes. Included adolescent weight loss studies from the US, Italy, Austria and Israel. Referred to a study by Australian authors on the topic of weight loss surgery and fatty liver disease.

Note: *Guidelines reviewed along with other systematic reviews and meta-analyses.

Table B.5: Assessment of evidence base on clinical guidelines

Apovian et al, 2005 (USA)Evidence base D – Eight pertinent case series published between 1980 and

2004, were identified and reviewed. These data were supplemented with expert opinions and literature on WLS in adults.

Applicability to Australian patient population

B- Applicable to Australian adolescents

August et al, 2008 (USA)Evidence base A/B – contains one meta-analysisApplicability to Australian patient population

B – American research applicable to Australian adolescents

Baur et al, 2010 (Australia and New Zealand) Evidence base B – no randomised controlled trial or controlled clinical trials on

adolescent bariatric surgery at the time of writing this guideline

114




A – Position paper from the Australian and New Zealand Association of Paediatric Surgeons, the Obesity Surgery Society of Australia and New Zealand and the Paediatrics and Child Health Division of The Royal Australasian College of Physicians

Buchwald, 2005 (USA) Evidence base D- Consensus Statement on the state of bariatric surgery for

morbid obesity has been prepared by a panel of broadly based and experienced experts based on presentations by investigators working in areas relevant to current questions in this field during a 1½-day public session.


B – applicable to Australian population

Fried et al, 2007 (European countries) Evidence base A - The panel’s recommendations are supported by the best

available evidence, which includes all evidence levels (Randomized controlled trials - RCTs, systematic reviews of cohort studies, observational “outcomes” studies, and expert opinion).


B – European guidelines applicable to Australian population

Laville et al, 2005 (France)Evidence base D – expert opinion. “The recommendations were developed by

the national associations of Obesity, Nutrition and Diabetes: the Association Française d'Études et de Recherches sur l'Obésité (AFERO), member of the EASO and IASO; the Association de Langue Française pour l'Étude du Diabète et des Maladies Métaboliques (ALFEDIAM); the Société Française de Nutrition (SFN); and the Société Française de Chirurgie de l’Obésité (SOFCO).


B/C – French research applicable to Australian population

Mechanick et al, 2008 (USA)Evidence base Range of evidence, A to DApplicability to Australian patient population

B – US guidelines applicable to Australian population

National Institutes of Health (NIH), 1998 (USA)Evidence base A - These guidelines are based on a systematic review of the

published scientific literature found in MEDLINE from January 1980 to September 1997 of topics identified by the panel as key to extrapolating the data related to the obesity evidence model.


B – United States guidelines applicable to Australian population

National Institutes of Health (NIH), 1996 (USA)Evidence base D –expert opinion. To resolve questions relating to surgery for

severe obesity, the National Institute of Diabetes and Digestive and Kidney Disease and the Office of Medical Applications of Research of the NIH convened a consensus panel representing the professional fields of surgery, general medicine, gastroenterology, nutrition, epidemiology, psychiatry, endocrinology, and including representatives from medical literature and the public, considered the evidence and agreed on answers to the questions that follow.

115




B –United States findings applicable to Australian population

National Institute for Health and Clinical Excellence (NICE), 2006 (UK)Evidence base A – Contains systematic reviews. Due to time constraints, full

systematic reviews were not undertaken. However, the evidence reviews were undertaken using systematic, transparent approaches. Plus expert opinion.


B – UK guidelines applicable to Australian population

Sauerland et al, 2005 (European countries) Evidence base A – contains systematic reviewApplicability to Australian patient population

B – European guidelines applicable to Australian population.

Snow et al, 2005 (USA) Evidence base A- contains systematic reviewApplicability to Australian patient population

B – US guidelines applicable to Australian population.

Society of American Gastrointestinal and Endoscopic Surgeons (SAGES), 2008 (USA) Evidence base A Applicability to Australian patient population

B

Table B.6: Data extracted from meta-analyses

Buchwald et al, 2007

Research question ≤30 day and >30 to 2 year mortality in patients who underwent bariatric surgery.

Patient population Mortality for subgroups Males vs. females Elderly Superobese

Interventions compared Mortality by procedure Lap gastric banding Lap and open gastroplasty Lap and open gastric bypass Lap and open BPD/DS

Mortality by procedure type Restrictive Restrictive/malabsorptive Malabsorptive

116



Safety outcomes reported(point estimates and statistical ranges)

Total mortality at <= 30 days: 0.28% (0.22%-0.34%)OpenRestrictive: 0.30% (0.15%-0.46%)Restricted/malabsorptive: 0.41% (0.24%-0.58%)Malabsorptive: 0.76% (0.29%-1.23%)LaparoscopicRestrictive: 0.07% (0.02%-0.12%)Restricted/malabsorptive:0.16% (0.09%-0.23%)Malabsorptive: 1.11% (0.00%-2.70%)

Males:Female 4.74:0.13>=65 years: 0.34% (0.00%-1.29%)Super obese (BMI>=50): 1.25% (0.56%-1.94%)

Total mortality at >30 days to 2 years: 0.35% (0.12%-0.58%)>=65 years: 0.0% (0.00%-1.63%)Super obese: 0.81% (0.00%-2.42%)

Mortality rate rend downward with more recent publication year, with the exception of late death in 2002-03

Efficacy outcomes reported(point estimates and statistical ranges)

N/A

Key conclusions and recommendations

The mortality for primary bariatric surgery procedures varies, with the lowest associated with LAGB and the highest with open BPD/DS and revisions.

The restrictive operatives have the lowest mortality, followed by the restrictive/malabsorptive gastric bypass operation and the malabsorptive BPD/DS has the highest.

Overall, relatively low mortality associated with bariatric surgery

Other key points No cross comparisons of mortality rates among different procedures (such as open vs. laparoscopic) were possible owing to the lack of comparative studies.

Buchwald et al, 2009

Research question To determine the impact of bariatric surgery on type 2 diabetes in association with the procedure performed and the weight reduction achieved.

Patient population Not specifiedBariatric surgery

Interventions compared All patients, gastric banding, gastroplasty, gastric bypass, BPD/DSOpen/laparoscopic


Not reported

117




%EWLAll patients*: 55.92% (54.06%, 57.78%)At <2 years: 53.82% (51.27%, 56.37%)At >=2 years: 59.00% (56.40%, 61.60%)*at time point for which data are available for at least 50% of study patients

Gastric banding:46.17 (43.14, 49.19)43.85 (40.25, 47.46)48.98 (44.00, 53.96)

Gastroplasty:55.53 (51.33, 59.73)54.58 (46.70, 62.46)56.48 (52.47, 60.49)

Gastric bypass:59.53 (56.47, 62.59)58.03 (54.25, 61.81)63.25 (58.39, 68.10)

BPD/DS:63.61 (57.52, 69.70)56.04 (47.91, 64.17)73.72 (69.02, 78.42)

Diabetic patient: 64.4%At <2 years: 67.1%At ≥2 years: 58.0%

Weight loss greatest for BPD/DS>gastric bypass>LAGB

Diabetes mellitus:Overall resolution: 78.1%Overall improvement/resolution: 86.6%Diabetes resolution was greatest for patients undergoing % resolved: BPD/DS (95.1%)>gastric bypass (80.3%)>gastroplasty (79.7%)> LAGB (56.7%)% resolved < 2 years: BPD/DS (94.0% resolved)>gastric bypass (81.6%)>gastroplasty (81.4%)> LAGB (55.0%)% resolved >= 2 years: BPD/DS (95.9% resolved)>gastroplasty (77.5%)> gastric bypass (70.9%)> LAGB (58.3%)Pure diabetic population: DM Resolution: 79.3%Improvement/resolution: 98.9%


The clinical and laboratory manifestations of type 2 diabetes are resolved or improved in the greater majority of patients after bariatric surgery; these responses are more pronounced in procedures associated with a greater %EWL and is maintained for 2 years or more.

Other key points N/AGarb et al, 2009

Research question There is a paucity of long-term clinical surveillance data for bariatric surgery beyond 1-year follow up. In response to this issues, the authors conducted a meta-analysis of the recent bariatric surgery clinical literature to examine reported weight loss outcomes for LAGB and LGB procedures.

Patient population Not specifiedInterventions compared LAGB and laparoscopic gastric bypass surgery.

118




Not reported


The composite estimate for %EWL was 49.4% (95% CI 44.9 to 54.0) for LAGB and 62.6 (95% CI 58.6 to 66.6) for LGB. The difference in effect sizes between the two types of surgery was statistically significant.

Composite %EWL broken down by time since surgery for LAGB showed an improving degree of EWL over time for both LAGB and LGB. Specifically, for LAGB, this was 42.6% at 1 year, 50.3% at 2 years, and 55.2% at >3 years since surgery. For LGB, this was 61.5% at 1 year, 69.7% at 2 years, and 71.2% at >3 years since surgery. Median attrition rates for LAGB studies and LGB studies at 1-year follow-up were 17.0% (range 0–77.7%) for LAGB and 0.0% (0–65.9%) for LGB patients. There was a marked loss of bariatric surgery patients to follow-up for both surgery types beyond the 1-year follow-up point.

The 24-month attrition rate was 49.8% (range 0–92.3%) for LAGB and 75.2% (0–95.8%) for LGB, and >3 year follow-up attrition rate was 82.6% (25.9–93.3%) for LAGB and 89.0% (49.2–90.4%) for LGB.

A composite %EWL of 49.4% for LAGB (versus the estimate of 47.5% by Buchwald et al.), and 62.6% for LGB (versus 61.6% for Buchwald et al.)

We found %EWL outcomes for LGB significantly superior to those for LAGB at all three time points examined (1, 2, and >3 years).


This meta-analysis confirms the superiority of LGB to LAGB in%EWL found in earlier studies. The results identified a composite %EWL of 49.4% for LAGB and 62.6% for LGB surgery.

Other key points This meta-analysis of bariatric surgery weight loss outcomes for 28 studies reported in the literature from 2003 to 2007 involved 7,383 LAGB and LGB patients. The studies reviewed were observational or interventional trials that were predominantly retrospective in design (72.4%) and carried out at academic hospital centres (79.4%) with the remainder carried out at community hospitals or in private practice (20.6%).

Shekelle et al, 2004

Research question What is the safety and efficacy of surgical therapies, such as stomach stapling and bypass surgeries, as interventions for children and adolescents with morbid obesity?

Patient population Gastroplasty (including horizontal banded gastroplasty, gastric portioning and gastrogastrostomy), jejunal-ileal bypass, BPD/DS, gastric bypass, adjustable and non-adjustable band, VBG

Adults ≥ 18 years of age with morbid obesity (BMI >40 or BMI >35 with serious comorbid disease

Interventions compared Pharmacological and surgical treatments

119




No clear pattern of differential mortality between various procedures

No clear pattern in terms of higher or lower early death rates in randomised trials vs. case series

Early mortality following bariatric surgery less than 1% Adverse events other than mortality are reported with

great variability amongst studiesEfficacy outcomes reported(point estimates and statistical ranges)

combined data show that RYGB patients reported about 10 kg more weight loss than patients treated with VBG, at both 12 and 36 months

In two RCTs, the weight lost using VBG, compared to LAGB, was 14 kg more at 12 months follow-up but only about 3 kg more at 36 months follow-up. No difference in net weight loss was seen in the pooled results from all studies combined.

No significant differences found between open and laparoscopic RYGB in terms of weight loss (>30 kg for both at 12 months). This is supported by the ‘all studies’ pooled analysis at both 12 months and out to 30 months

Surgical treatment results in greater weight loss than medical treatments in obese individuals (BMI ≥40), resulting in 20-30 kg of weight loss maintained up to 8 years, accompanied by significant improvements in several comorbidities vs. 2-5 kg in pharmaceutical studies (although direct comparison cannot be made due to the different patient populations)

For BMI 35-40, data strongly support the superiority of surgical therapy, but cannot be considered conclusive.


There is limited evidence supporting greater long-term weight loss (maintained at least to eight years) with surgery than with conventional treatments for severe obesity

Surgery is associated with adverse effects and the possibility of postoperative mortality.

Data are too limited to draw any conclusions regarding differences in efficacy or safety among surgical procedures

Other key points For patients with BMI between 35-40kg/m2, data strongly supports superiority of surgical therapy (but does not have concurrent comparison group)

Gastric bypass produces superior weight loss compared to gastroplasty procedures

Treadwell et al, 2008

Research question Review all published evidence pertaining specifically to bariatric surgery in paediatric patients in terms of weight loss, improve medical comorbidities, enhance quality of life and extend survival

Patient population Patients ≤ 21 years of age

120



Interventions compared Bariatric surgery performed in the US – LAGB, RYGB, VBG, banded bypass and BPD


LAGB No in-hospital or postoperative death reported Reoperations performed on 8% of patients to correct

various complications e.g. band slippage, gastric dilation, intragastric band migration, psychologic intolerance of band, hiatal hernia, cholecystitis and tubing crack

Band slippage most frequent (3% of cases) 8/352 cases of iron deficiency and 5/352 cases of mild

hair loss reported No studies reported on impact of surgery on growth or

developmentGastric bypass

No in-hospital death reported 1 patient died 9 months post surgery from severe

Clostridium difficile colitis, severe diarrhoea, an extended period of profound hypovolemia and multiple organ failure

3 patients died from causes unlikely to be directly related to bariatric surgery (at 2, 4 and 6 years)

Postoperative complications – shock, pulmonary embolism, severe malnutrition, immediate postoperative bleeding and gastrointestinal obstruction

Unclear as to whether physical malnutrition affected adolescents achieving their expected growth

121




LAGBReduction in BMI

Follow-up period after LAGB: 1-3 years % of patients included in each study 58%-100% 95% confidence interval of random-effects summary

statistic: -13.7 to -10.6 BMI units Significant weight loss post surgery persisted through

all sensitivity analysesResolution of comorbidities

Follow-up period: 1.3 -2.9 years Diabetes: 2/4 reported resolution rates of 100% and

80% Hypertension: ¾ reported resolution rates of 50%,

100% and 100%Gastric BypassReduction in BMI

Follow-up: 1-6.3 years % of patients included 61%-100% 95% confidence interval: -17.8 to -22.3 BMI units post

surgery Significant weight loss persisted through all sensitivity

analysesResolution of comorbidities

Follow-up: 5 months to 2.7 years Hypertension: ¾ studies reported resolution rates of

50%, 82% and 100%Key conclusions and recommendations

Bariatric surgery in paediatric patients result in sustained and clinically significant weight loss, but also has potential for serious complications

Limitations: lack of reporting of long-term data on sufficient number of participants, of comorbidity burden and resolution and compliance with postsurgical recommendations

Other key points For paediatric patients, three issues relating to bariatric surgery: informed consent, interference with physical growth/maturation and compliance with postsurgical diets

Limited evidence on quality of life improvements post surgery or extended survival

Paediatric patients who undergo bariatric surgery usually have previously had unsuccessful weight loss with nonsurgical methods

Table B.7: Data extracted from systematic reviews

Brethauer et al, 2009

122



Research question To evaluate the current evidence regarding weight loss, complication rates, postoperative mortality, and co-morbidity improvement after SG

Patient population AdultInterventions compared SG (high-risk patients/staged approach vs. primary group)

Of the 36 studies, open = 1 laparoscopic = 35Safety outcomes reported(point estimates and statistical ranges)

Complications and operative mortalityPostoperative complication rate ranged from 0%-23.8%For studies with >100 patients: 0%-15.3%Overall mortality rate 0.19%


Weight lossOverall mean %EWL = 55.4% (33%-85%)Mean post-operative BMI decreased from 51.2 to 37.1 kg/m2

Follow-up = 3-60 monthsCo-morbidity reductionDM resolved/improved in greater than 70% of patientsSignificant improvements in other components of the metabolic syndrome (i.e., hypertension, hyperlipidemia), sleep apnoea and joint painFollow-up = 1-5 years


From the current evidence, LSG is an effective weight loss procedure that can be performed safely as a first stage or primary procedure. LSG results in excellent weight loss and comorbidity reduction that exceeds, or is comparable to, that of other accepted bariatric procedures.

Other key points Long-term data are limited, but the 3- and 5- year follow-up data have demonstrated the durability of the SG procedure

Chapman et al, 2004

Research question To compare the safety and efficacy of LAGB with VBG and gastric bypass

Patient population Morbidly obese patients (BMI >35 kg/m2)Interventions compared LAGB, VBG, RYGB

123




Mortality rates – see table 1 below Short-term deaths

LAGB: 0.05% (0.01-0.11)RYGB: 0.50% (0.36-0.64)VBG: 0.31% (0.11-0.52)

The relative risks for short-term deaths from LAGB vs VBG 0.16 (0.04-0.61) P=0.0001LAGB vs RYGB 0.10 (0.03-0.33) P=0.007

Long-term deathsLAGB: 0.17%RYGB: 0.49% VBG: 0.45%

Relative long-term risks and confidence intervals were not calculated due to variation in long-term follow-up times between studies.

Complications (overall morbidity rates)LAGB: median 11.3% (range 0%-68.0%)VBG: median 23.6% (range 0%-93.3%)RYGB: median 27.4% (range 0%-76.7%)This analysis makes no distinction between the types of morbidity associated with each procedure.Might reflect varying sensitivities of different authors to what constitutes a reportable complication, or it might reflect a broad variability in surgical technique or experience.Specific morbiditiesMost common types of complications:

LAGB – pouch dilatation (4%) and displacement of the band (1.6%)

VBG – incisional hernia (5.1%), wound infection (3.9%), staple line disruption (2.9%), seromas and hematomas (2.6%), and stenosis (2.0%) or other pouch problems (1.8%)

RYGB – incisional hernias (8.9%), nutrient deficiencies, anemia or anorexia (6%), stenosis of the pouch outlet (4.8%), wound infection (4.5%), marginal ulcer (4.1%), and staple line disruption (2.4%).

Rates are underestimates as not every study that reported complications reported all complications.Vomiting and food intolerance (rates):LAGB: 0-60%VBG: 0.8%-76.5%RYGB: 4.7%-68.8%

124




Weight loss (% of excess weight lost at 4 years)LAGB: 44%-68%VBG: 40%-77%RYGB: 50%-67%Conversion ratesLAGB: 0%-25%5.3% was the highest rate from the larger series (>100 patients)Reoperation ratesLAGB is associated with a lower risk of reoperation (most reported <=8%) and the greatest risk pertains to VBG (reoperation rates among the larger series, were between 20-53%).Revision rateHighest rate of revision were associated with VBG (1.7-31%). RYBG is associated with very low revision rates (0.2%-10%). Revision rates for LAGB varied from 0.6% to 71%. In the study that reported 71%, significantly lower rates of reoperation rates were reported after technique modification. The highest rates of operative reversal are recorded in the smaller series of LABG and represented the initial experience with LAGB.Discharge (postoperative duration of hospital stay)LAGB: mean 1.2-11.8 days (range, 0-55)Open VBG: 2.9-11.4 days (range, unknown -90)Open RYGB:1.6-8.4 days (range, unknown -64)Psychosocial effectsOne study compared QoL outcomes for the three procedures examined. Overall, patient treated with RYBG reported significantly higher scores that those treated with either VBG or LAGB (no statistical differences between either of these groups) using the BAROS system. Another study reported 75% of ‘very satisfied’ patients among the RYGB group than the VBG group (54%). One of the historical studies found that at 2 year follow-up, patients treated with LAGB were significantly more likely to report a greater disparity between their current weight and their ideal weight, and scored more poorly on a range of measures + less positive evaluation of the surgery when compared to the RYGB group.Resolution of comorbiditiesOnly one study performed statistical analysis – found no significant differences between LAGB, VBG, and RYGB in terms of improvement in medical conditions.


Considering the increased risk of morbidity associated with VBG, along with the high likelihood of surgical failure and the requirement for the procedure to be revised, LAGB or RYGB is preferred, the former for its safety (at least in the short term) and the latter for its efficacy. A caveat to this observation is the lack of comparable long-term data available for the LAGB.

Other key points Up to 2 years, the laparoscopic gastric band results in less weight loss than RYGB; from 2-4 years, there is insufficient evidence to conclude that RYGB remains more effective than LAGB

Colquitt et al, 2009

Research question To assess the effects of bariatric surgery for obesity on weight, comorbidities and quality of life

125



Patient population Adults fulfilling the standard definition of obese, i.e. people with a BMI of 30 or over.Young people who fulfil the definition of obesity for their age, sex and height.

Interventions compared Comparison of surgical procedures: gastric bypass, AGB, BPD, SG, VBGSurgical procedures versus usual care (no treatment or medical treatment)Open surgery compared with laparoscopic surgery for the same procedure

126




SURGERY VERSUS NO SURGERYMortality (perioperative and total)

LAGB adverse advents include: operative interventions (13%), laparoscopic revision (prolapse or posterior) (10%), 5mm port site infection (2.6%), and acute cholecystitis (2.6%). Loss to follow-up was higher in the non-surgical group (16%) compared to laparoscopic adjustable gastric banding group (2.6%)

SOS: within 90days there were 5 (0.25%) deaths in the surgery group and 2 in the control group. Perioperative complications at 13%. Post-op complications at 2.2%. Cumulative overall mortality during a period of up to 16 years (mean 10.9 years follow-up). The hazard ratio of the surgery group compared with the control group was 0.76 [(95%CI 0.59 to 0.99) P = 0.04)]. There were 5% deaths in the surgery group and 6.3% deaths in the control group.

SURGERY VERSUS SURGERY Gastric bypass versus VBG Mortality - One 1987 RCT trial reported no deaths in the VBG group but two deaths (10%) in the gastric bypass group.Complications - Operative time was significantly less with LVBG. Early postoperative complications were more common following laparoscopic gastric bypass than LVBG. Similar readmission for late complications between the laparoscopic procedures Two studies of open surgery report that after approximately three years, conversions to an alternative bariatric procedure occurred more often in the VBG groups but neither study tested this for statistical significance.LRYGB versus LAGBMortality - no deaths reported. Complications and additional operative procedure. Operative time: LAGB < LRYGBMean hospital stay: LAGB (2 days) < LRYGB (4 days)Reoperation: rates were similar, LAGB (15.2%) > LRYGB (12.5%) Early complications requiring reoperation: LRYGB > LAGB but the numbers were small and not tested for statistical significance. Laparoscopic gastric bypass versus laparoscopic SGComplications and additional operative procedures. No conversions to open surgery and no intraoperative and postoperative complications reported. VBG versus LAGB Mortality - deaths reported, but unrelated to surgery Complications and additional operative proceduresResolution of comorbidities appears to be similar. Due to data limitations, it is difficult to draw conclusions regarding complications and additional operative procedures. More reoperations were necessary following open VBG than open AGB but a statistical comparison was not reported. LAGB was associated with a statistically shorter operative time and hospital stay than LVBG but there were statistically more late complications and reoperations. Open VBG led to more infections. Late complications requiring further surgery were similar but a statistical comparison was not reported.LAGB versus laparoscopic isolated SGComplications and additional operative procedures –Rates of

127




SURGERY VERSUS NO SURGERYWeight change, fat content, fat distributionLAGB vs non-surgical in people with BMI 30 to 40+comoribiditiesStatistically significant benefit on measures of weight change for those receiving LAGB (at up to 2 years)Statistically significant loss in people diagnosed with type 2 diabetes at 2 years.% EWLLAGB: 62.5%Conventional: 4.3%Similar benefits were noted on measures of waist circumference and waist hip ratio LAGB vs non-surgical in people with BMI 30 to 35+comoribidities Statistically significant difference in the weight of participants at 12, 18 and 24 months. People in LAGB group consistently lost weight during the 2 year follow-up, those in the non-surgical group increased in weight, despite an initial loss of weight at 6 months. BMI decrease in the LAGB group: 33.7 to 26.4 (at baseline and at 2 yrs) (87.2% of excess weight) compared to 33.5 and 31.5 (21.8% of excess weight) in the non surgical group.At 2 year follow-up (n=3505)23.4% weight loss among gastric surgery group vs 0.1% gain in patients receiving conventional treatment.At 10 year follow-up (n=1276)16% weight loss in surgical group vs 1.5% gain in patients receiving conventional treatment.At 15 years follow-up (n=not reported)Gastric bypass: 27% (SD 12)VBG: 18% (SD 11)Gastric banding: 13% (SD 14)Quality of life

One RCT reported improvement in scores on all 8 domains of SF-36 for LAGB group and 3 domains (physical function, vitality and mental health) for the non-surgical therapy group. Statically significant greater improvements were reported for 5 domains (physical function, physical role, general health, vitality and emotional role) for LAGB compared to the non-surgical group.

The SOS 1997-2007: At baseline the patients in the surgery group had generally worse HRQoL than those in the conventional treatment group. At 2 years follow-up gastric surgery patients had significant improvements in all HRQoL measures compared to patients receiving conventional treatment. Changes were significantly related to the magnitude of the weight lost. Improvements in HRQoL peaked at 1 year after surgery, followed by a gradual decline between 1-6 years, and then stabilises between 6-10 years follow-up. HRQoL measures compared to baseline at 10 years were improved for the surgery group, but some had improved while others had worsened for the conventional group.

128



SURGERY VERSUS SURGERY Gastric bypass versus VBGSeven RCTs were included. Six trials were of uncertain risk of bias as many factors were not reported. On measures of weight, gastric bypass > VBG, particularly at later time points, in five of the seven trials (although a statistically significant difference was only reported in three of these trials). Only one study reported on quality of life which they found to be better for the gastric bypass group. LRYGB versus LAGBOn a variety of measures of weight, a small study showed that LRYGB was superior to LAGB. The risk of bias in this study is uncertain, although the risk of bias from incomplete outcome data for weight loss and comorbidities is likely to be low.LRYGB versus LSGIn a small RCT with an uncertain risk of bias, BMI and weight loss at 12 months follow-up were similar between LRYGB and SG. %EWL was greater with SG at 12 months.VBG versus AGBThree studies were included; one had a low risk of selection bias and two were of uncertain risk of bias for several items, although missing outcome data for weight loss were adequately addressed. Weight loss results were inconclusive. One study found that weight loss was initially greater with VBG, but weight regain meant that by three years patients with AGB had a lower mean weight, and this was still the case at five years (statistical significance not reported). The second study found lower BMI and greater %EWL following LVBG, but this was statistically significant only at year one and not at years two or three. The third study found statistically significant lower BMI and greater %EWL at one and two years following open VBG, and greater % excess BMI loss seven years after open VBG (statistical significance not reported). However, the impact of participants being converted to another procedure in this study is unclear. More patients who had undergone AGB reported being satisfied with the results at five years but this apparent superiority was not tested statistically. LAGB versus laparoscopic isolated SGParticipants undergoing laparoscopic isolated SG showed more improvement than participants undergoing LAGB in one study with an uncertain risk of bias.


A positive effect related to the surgical approach when it was compared to conservative management. Certain procedures produce greater weight loss, but a meta-analysis could not be performed as the studies included were limited and subjected to potential bias. The evidence on safety is even less clear. Due to limited evidence and poor quality of the trials, caution is required when interpreting comparative safety and effectiveness.

Other key pointsDe Groot et al, 2009

Research question To compare the effect of various weight reducing modalities on manifestations of GORD in obese patients

Patient population BMI>25 + GORD

129



Interventions compared Diet/lifestyle, RYGB, VBG, AGB, laparotomic and non-adjustable gastric banding techniques


Not reported


Diet/lifestyle:4/7 studies reported an improvement of GORDDifficult to achieve/maintain significant weight lossLimitations: one of the methodologically sound RCT had a population with a mean BMI of 23, meaning no clear overweight or obese individuals were includedRYGB:All, except one, found positive effect on GORD symptomsMean % EWL: 72% (range 68.8%-76%) variable follow-upComparative studies showed RYGB yield better results than gastric banding with regard to GORD reductionLimitation – mainly questionnaires were used to evaluate the effect on GORD rather than objective measurementsRestrictive - VBG and LAGB:Conflicting evidence on GORDAll studies reported weight reduction, mean BMI decreased from 39.8 (range 23.5-56) to 31.5 (21.8-42.0)

VBG – the positive effect of weight loss may be counteracted by a negative effect of the operation (e.g. acid accumulation on the VBG pouch)

Gastric banding – use of different inclusion/exclusion criteria in the various studies may have contributed to the conflicting results. The negative results were mainly in studies with more objective tests (such as 24h pH-metry, manometry and endoscopy. Similar to VBG, reduced volume of the gastric cardia with an increased intragastric pressure and/or a change in the anatomy at the gastro-oesophageal junction may be associated with increase of GORD. On the other hand, decrease in GORD symptoms can be explained by an increased length and/or increased pressure of the lower oesophageal sphincter.


Diet and lifestyle intervention appears to be beneficial with respect to GORD. Of all the surgical techniques evaluated in this review, RYGB seems to be the most promising in reducing GORD, whereas VBG appears to be ineffective. Gastric banding may improve or worsen GORD.

Other key points It is recommended that standardized and validated questionnaires are used and 24h pH-metry measurements at different time points are included to quantify objectively the effect of weight reduction on GORD

Douketis et al, 2005

130



Research question To review long term (≥2 y) studies investigating dietary/lifestyle, pharmacologic, and surgical weight loss methods to assess

Weight loss efficacy, defined by absolute weight loss and the proportion of subjects with ≥5% weight loss (clinically important weight loss, defined by obesity experts as a loss of 5-10% of baseline weight);

Effects of weight loss on CV risk factors; and Applicability of findings from studies to everyday

clinical practice.Patient population Overweight or obese adult (aged 18-65 years) with a BMI

>=25kg/m2

Interventions compared Dietary/lifestyle, pharmacologic and surgical weight loss methods.

Dietary/lifestyle regimen consisting of a very low-calorie diet (<1100kcal/day)

Pharmacologic drug therapy with orlistat or sibutramine. Drug therapy is combined, typically, with dietary therapy.

Surgical therapy consisting of a restrictive procedure (e.g. VBG, gastric banding) or a diversionary procedure (e.g. gastric bypass). Surgical therapy is usually combined with dietary therapy and/or a behaviour counselling/lifestyle modification program and, typically, is limited to people with class II obesity (BMI≥35) with severe comorbidity or class III obesity (BMI≥40)


Not reported

131




Dietary/lifestyle therapy (16 studies) <5 kg weight loss after 2-4 years a trend of decreased blood pressure across studies,

statistically significant in some but not other studies. Many (9/16) studies had high proportion (31-64%) of

subjects lost to follow-up. In all but 2 studies, outcomes were reported based on a ‘study-completers’ analysis, which may overestimate weight loss because study completers may have greater weight loss than non-completers.

Pharmacologic therapy (19 studies) 5-10kg weight loss after 1-2 years Weight loss of >=5% occurred in 40-60% of subjects Pharmacologic therapy improved lipid levels and

glycaemic and blood pressure control, but these effects were drug specific and greatest in subjects with elevated baseline levels.

All studies except one had a follow-up of 2 yr or less. The 2 yr studies had a high proportion of subjects lost to follow up (30-57%)

Surgical therapy (9 studies) 25-75kg weight loss after 2-4 years No studies reported the proportion of subjects with

≥5% weight loss, although this was probably high given the absolute weight loss

Only 2 studies assessed this, thereby precluding meaningful commentary


In terms of weight loss efficacy: dietary/lifestyle therapy provides <5kg after 2–4 y; pharmacologic therapy provides 5– 10 kg after 1–2 y; and surgical therapy provides 25–75 kg weight loss after 2–4 y.

Weight loss of ≥5% is not consistently associated with improvements in cardiovascular risk factors, and when improvements occur they are mainly in high-risk groups and appear to be intervention specific; and

Weight loss studies have methodological limitations that restrict their application to everyday clinical practice.

Other key points N/AFarrell et al, 2009

Research question This review is intended to guide surgeons applying laparoscopic techniques to the practice of bariatric surgery. The main objectives of this study were to analyse the impact of laparoscopic bariatric surgery on mortality, weight loss, and comorbidities; to review the indications for bariatric surgery; to review the ancillary services required for a bariatric practice; to compare gastric bypass, BPD, and LAGB; and to review the various reoperative options available

Patient population Not specified, but reference is made to US adults. Interventions compared RYGB; AGB; BPD with DS; LSG.

132




LBPD with DS The 30-day mortality of early laparoscopic BPD series

ranges from 2.6% to 7.6%. Major complications, which occur in up to 25% of cases,

may include early occurrence of anastomotic leak, duodenal stump leak, intraabdominal infection, hemorrhage, and venous thromboembolism, or later bowel obstruction, incarceration, or stricture.

Cholelithiasis postoperatively occurs in 6% of patients to 25%.

LRYGB The mortality rate after RYGB ranges from 0.3% in case

series to 1% in controlled trials, and the rate of preventable and nonpreventable adverse surgical events is 18.7%.

The mortality rate in a review of selected LRYGB series ranged from 0.5% to 1.1%.

The most frequently reported perioperative complications associated with LRYGB are wound infection (2.98%), anastomotic leak (2.05%), gastrointestinal tract hemorrhage (1.93%), bowel obstruction (1.73%), and pulmonary embolus (0.41%), whereas the most frequently reported late complications are stomal stenosis (4.73%), bowel obstruction (3.15%), and incisional hernia (0.47%).

LAGB Life threatening complications are less frequent with

LAGB than with LRYGB. Case series and systematic reviews put early mortality

rates after LAGB at 0.05–0.4% compared with 0.5–1.1% for LRGB 0.5–1.0% for open RGB, 1.1% for open BPD, and 2.5–7.6% for LBPD.

Few comparative data on relative morbidity rates Overall complications and major complications are less

common with LAGB than with LRYGB or LBPD in a single centre experience

133




LBPD with DS The BPD ± DS technique initiates dramatic weight loss

during the first 12 postoperative months, which then continues at a slower rate over the next 6 months.

Weight loss is durable up to at least 5 years postoperatively.

Available data suggest that the weight loss effect of BPD is greater and more durable than that of LAGB. Likewise, BPD may be superior to RYGB for patients with a BMI of 50 kg/m2 or greater.

95% of patients with a BMI less than 50 kg/m2, and 70% of those with a BMI exceeding 50 kg/m2 achieve more than 50% EBWL

Co-morbidities: Up to 98% of patients with obstructive sleep apnoea

experience resolution. The BPD procedure has a dramatic impact on comorbidities.

At least 90% of patients with type T2DM cease diabetic medications by 12–36 months.

Between 50% and 80% of hypertensive patients are cured, with another 10% experiencing improvement.

LRYGB Patients who undergo LRYGB typically experience an

EBWL of 60–70%, with 75% control of comorbidities. In general, these outcomes are better than those for banding procedures, which have an EBWL of 45–50% and less predictable improvement of comorbidities, but poorer than the outcomes for BPD ± DS, which has an EBWL of 70–80% with excellent control of comorbidities.

Open and LRYGB have similar efficacy. In prospective randomized trials, they show no significant differences in weight loss in up to 3 years of follow-up evaluation. Similar results have been reported in case series.

LAGBEBWL: Approximately 35% by 6 months, 40% by 12 months, and 50% by 24 months. This percentage appears to remain stable after 3–8 years. However, as many as 25% of LAGB patients fail to lose 50% of their excess body weight by 5 years.Co-morbidities:

T2DM is improved for about 90% of, and diabetic medications are eliminated for 64%.

The rate of obstructive sleep apnoea drops from 33% to 2% for LAGB patients.

134




Bariatric surgery is medically indicated for morbidly obese patients who fail to respond to dietary, behavioural, nutritional, and medical therapies, with clear evidence of efficacy and safety. Age- and BMI-based candidacy guidelines should not limit access for patients experiencing progressive or poorly controlled obesity-related comorbidities if the risk-versus-benefit analysis favours surgery. Laparoscopic RYGB, AGB, and BPD all have been proved effective.

Other key points Given the marked paucity of prospectively collected comparative data among the different bariatric operations, it remains impossible to make definitive recommendations for one procedure over another. Currently, decisions are driven by patient and surgeon preferences as well as by considerations regarding the degree and timing of necessary outcomes versus tolerance of risk and lifestyle change.

Gentileschi et al, 2002

Research question Evidence based analysis of the literature on open and laparoscopic surgery for morbid obesity

Patient population BMI>35Interventions compared Open vs. Laparoscopic

AGB, RYGB, VBG, BPD/DSSafety outcomes reported(point estimates and statistical ranges)

Laparoscopic bariatric surgery LAGB (61 studies):

Overall morbidity rate: 11.5% (range 0-60%)Overall mortality rate: 0.05%The most frequent complications were band slippage (3.80%) and pouch dilation (3.25%)Reintervention rate ranged from 1.7% to 66.7%

LRYGB (10 studies):Overall morbidity: 16%Overall mortality: 0.2%Overall conversion rate: 2.4%

LVBG compared to open VBG (10 studies):Longer operative time for LVBG but decrease in incidence of would infections and incisional hernias compared to open VBG

LBPD/DS (1 study):Morbidity: 15%30 day mortality: 2.5%Conversion to laparotomy: 2.5%

135




Open bariatric surgery (12 studies)Significant greater long-term weight loss after RYGB (level of evidence IB, grade A) compared to VBGExcellent weight loss after BPD (level of evidence 2A, grade B) and BPD/DS (level of evidence 3B, grade B) have been shownGood weight loss after open AGB has been shown in one randomised controlled trial with 1 year follow-up evaluation. Long term efficacy cannot be determined due to incomplete and poor evidence.Laparoscopic bariatric surgery

LAGB: Mean estimated WL ranged from 18% to 72%. Mean estimated WL at 4 yrs ranged from 44%-68% though the evidence with regard to long-term weight loss was rather thin (i.e. very few patients reached the 4 year follow up evaluation)

LRYBG:Estimated weight loss at 3 years: 77%

LVBG compared to open VBG:No difference in weight loss reported

LBPD/DS (1 study):Estimated weight loss: 58%


LRYGB has proved to be as safe as its open counterpart, although its long-term weight loss results are still lacking.

Other key points LAGB is less invasive and preferable to open AGB. The efficacy of both procedures cannot be determined currently because of poor evidence. LVBG is following the decreasing trend of open VBG as a result of RCT comparing VBG and RYGB. As for LBPD/DS, its feasibility has been proved, but very poor evidence is provided currently regarding its effectiveness.

Glenny and O’Meara

Research question To assess the effectiveness of intervention used in the prevention and treatment of obesity and the maintenance of weight loss.

Patient population Overweight and obese adults and children, or those considered at risk of developing the condition were included. People suffering eating disorders were excluded. Surgery and it’s outcomes were only discussed for adults (‘usually morbidly obese’ – generally BMI>40kg/m2).

Interventions compared Behavioural, dietary, exercise, pharmacological, surgical (including: gastric bypass; gastroplasty; jejunoileal bypass; gastrogastrostomy; and gastric balloons) and alternative therapies.


One study reported 1 death 3 days post RYGB and 2 more deaths with 2 years – cardiac arrests. Six papers reported re-operation due to weight loss failure. For gastroplasty re-operation ranges 12%-33%. One paper found 10 (50%) of RYGB experience dumping compared to no complaints in gastroplasty. GB may affect patients QoL and result in vitamin deficiency.

136




Typical weight loss one year after GB was 45-65 kg compared to 30-35 kg after gastroplasty.

One study, RYGB compare to VBG lost 25% more excess weight at 1 year.

Study two, 83% success for reoperation (isolated GB due to failure) compared to initial VBG and 58% for RYGB (p=0.003). Success = reduction to <50% of excess weight.

Third study showed 66% of patients achieved >50% loss of excess weight versus 44% for gastroplasty and 16% for gastrogastromy.


In general the weight loss associated with surgical interventions is greater and more sustained than that achieved by non-surgical methods. However surgery is associated with complications that may affect the patient’s quality of life.Jejunoileal bypass, VBG and gastric bypass all have been found to produce a significant weight loss. Out of the three procedures, gastric bypass appears to be the most effective of surgeries for weight loss maintenance.

Other key points N/AManterola et al, 2005

Research question Consider the evidence for the best bariatric surgical optionsPatient population Adult humans > 19 years with morbid obesity who had not

undergone prior bariatric surgeryInterventions compared Banding, gastroplasty, gastric bypass and BPD (laparoscopy and

open surgery)Safety outcomes reported(point estimates and statistical ranges)

Hospital stay periods- 3 to 7 days (longer for open techniques)

Rate of re-operations higher by 6% for laparoscopic operations (esp gastric banding)

Mortality minimal >0.5%, higher for laparoscopic techniques

Morbidity lower for laparoscopic techniques at 2-3 yrs (1 yr more for open series)


Open surgery superior over laparoscopic techniques at 12, 24 and 36 months for:

Decrease in BMI by >7% %EWL (difference of 6%) Reduction of co-morbidity (difference of 3%)

Subgroup analysis Amongst laparoscopic operations, BPD showed largest

decrease in BMI at 12 months Amongst laparotomy operations, BPD showed largest

decrease in BMI at 36 months, higher values of %EWL and co-morbidity reduction


Comparison of bariatric procedures difficult due to lack of uniformity of variables used for measuring results

Methodologies of studies were poorOther key points Level of evidence of studies were very poor Schneider, 2000

137



Research question Highlight evidence from published scientific literature regarding safety, efficacy and effectiveness of LAGB procedure

Patient population Not specified although defined clinically severe obesity as BMI >35 with attendant comorbidites, 45 kg overweight, or BMI >40

Interventions compared LAGBSafety outcomes reported(point estimates and statistical ranges)

In 2 studies – surgery duration averaged 90 minutes and hospital stay averaged 2 days

Conversion surgery reasons – left liver lobe hypertrophy, difficult and risky dissection, short instrument and incorrect band position, gastric perforation, peri-gastric dissection and bleeding from a retro-gastric vessel

Complications that occurred less than 12% of the time – aspiration pneumonia, band slippage, rotated access ports and infection of access port

Re-operation required in 4% of cases Range of time over which complications occurred – 3

months to 3 yearsEfficacy outcomes reported(point estimates and statistical ranges)

All studies reported decreases in BMI, weight loss and/or excess weight loss after LAGB surgery (fair to poor levels of evidence)


Advanced surgical skills along with patient compliance to follow up instructions are essential to achieve successful outcomes e.g. decrease BMI, high % excess weight loss and subsequent decreases in morbidity

Other key points Whether LAGB surgery can be offered to morbidly obese population outside a hospital setting cannot be determined (as studies took place within hospital setting)

Tice et al, 2008

Research question To evaluate the balance of patient-oriented clinical outcomes for LAGB and RYGB

Patient population AllInterventions compared LAGB and RYGBSafety outcomes reported(point estimates and statistical ranges)

Short term complications LAGB had shorter operative times by a median of 68

mins and hospitalisation length of stay about 2 days shorter

Fewer deaths in LAGB (0.06% vs 0.17%); but mortality low for both surgeries

Rates of conversion to open procedures, perforation, bleeding and anastomotic leaks low in both surgery types

Overall, reported difference in major early complications ranged 1.1%-6.3% in favour of LAGB


Weight lossMedian difference in excess body weight loss at 1 year = 25% (in favour of RYGB)Resolution of comorbiditiesAbsolute differences in resolution of comorbidities of 25% or more in favour of RYGB (no. needed to treat ≤ 4)

138




Observational evidence demonstrates greater weight loss and improvements in obesity related conditions with RYGB compared to LAGB

Randomized, controlled comparative trials with larger sample sizes are needed to determine whether there are subgroups of patients who may benefit from lower short term complication rates in LAGB

Other key points In the only randomized clinical trial: Excess body weight loss at 1 year: 51% for RYGB vs. 35%

for LAGB Patients failing to lose any weight at all: 4% for RYGB vs.

35% for LAGB Reoperation rates: 12% for RYGB vs. 15% for LAGB No deaths during follow up of 1 year

In the highest quality observational study: Weight loss at 1 year: 76% for RYGB vs. 48% for LAGB

(P<0.001) – results stable over 3 years Resolution of diabetes: 78% for RYGB vs. 50% for LAGB Reoperation rates: 19% for RYGP vs. 24% for LAGB No deaths reported in study

Department of Health and Ageing (DoHA), 2003a (Australia)*

Research question To provide detailed evidence-based guidance for assessing and managing overweight and obesity in Australia

Patient population Adults with a BMI >40 kg/m2 or a BMI >35 kg/m2 and serious medical co-morbidities

Interventions compared RYGB, VBG, LAGB and BPDSafety outcomes reported(point estimates and statistical ranges)

Gastric bypass Has a slightly higher risk of metabolic complications

related to malabsorption – such as vitamin B12 deficiency, and therefore should be reserved for heavier patients.

Gastroplasty Appears safer than gastric bypass Has the potential to cause dumping syndrome LAGB is less invasive, adjustable and more easily

reversed than VBG and gastric bypass, but is associated with a higher reoperation rate

BPD Patients may experience gastrointestinal side effects if

lifestyle requirements are not adhered to. BMI > 65 has been associated with increased morbidity

139




Gastric bypass Achieves significant and permanent weight loss in most

patients On average, weight loss one to two years is 36%, and

weight loss is well maintained over time Ineffective in 43% of super-obese patients 5.5 years

after the operation Laparoscopic RYGB is just as effective as open

procedures 70% of subjects found the results acceptable and

satisfactory.Gastroplasty

Is effective in inducing and maintaining long-term weight loss, however it is not as effective as gastric bypass and there is some weight regain over time.

On average, weight loss one to two years after gastroplasty is 32%, but after 3-8 years, this falls to 20%

VBG results in significant weight loss in super obese individuals, although they remain obese, with a BMI over 35

Laparoscopic VBG is just as effective as open procedures

On average, weight loss one to two years after LAGB is 24%, and this weight loss is well maintained over three to four years

Laparoscopic and open adjustable silicone gastric banding have been shown to be equally effective in terms of early (first-year) weight loss and post-operative complications.

BPD Good weight maintenance observed following the

procedure and may be considered for the most obese patients

On average, weight loss one to two years is 38%, and this weight loss is well maintained in the longer term

A study showed reversal of comorbidities in all patients therefore the procedure appears to be very effective for the treatment of morbid obesity

140




The weight loss induced by surgical intervention results in a marked reduction in some of the co-morbidities associated with obesity.

Whether intentional weight loss will return risk facts of baseline in a longer perspective (10-20 years) and reduce mortality is still too early to tell. However, there are indications that surgically induced weight loss has an impact on mortality.

The co-morbidities associated with obesity can, however, increase perioperative risk.

In patients with acceptable operative risks, mortality as a consequence of bariatric surgery is low, especially in skilled hands.

Bariatric surgery is often associated with impaired absorption of micronutrients, with requires lifelong monitoring and often folate or vitamin B supplementation.

Surgical intervention combined with permanent lifestyle change is the most effective therapy for weight reduction in terms of the extent and duration of weight loss.

Other key points The problem among specific groups and Aboriginal and Torres Strait Islander peoples in particular, has distinct characteristics that are currently less well understood.

Department of Health and Ageing (DoHA), 2003b (Australia)*

Research question There is limited evidence that gastric bypass or gastric restrictive surgery in obese adolescents induces a weight loss comparable to that shown in adult studies. There are, however, no established criteria for determining which subjects would benefit from such a procedure.

Patient population Older adolescentsInterventions compared VBG, RYGB, intra-gastric balloonsSafety outcomes reported(point estimates and statistical ranges)

There was a high prevalence of post-operative morbidity. Nearly one-third of subjects had a re-anastomosis.

Three deaths were reported. Postoperative morbidity was high and included

infection, poor wound healing, symptomatic cholelithiasis , and micronutrient deficiencies.


Not all subjects had major weight loss; in those whom the studies identified as successful, the mean weight loss was 50 kg, with reported losses up to 90 kg.

Subjects lost a mean of 60% of their weight above ideal body weight. Weight loss tended to plateau at 15 to 20 months, and some females experienced significant regain with pregnancy.


There is evidence that gastric restrictive or gastric bypass surgery induces a weight loss in adolescents, with a reduction in obesity-related co-morbidity that is comparable to that found in adult studies. The overall numbers are low, however, and long-term follow-up data are limited. Not every subject experiences significant weight loss, and there are no good data to suggest who will be successful. Post-operative morbidity is common.

141



Other key points Bariatric surgery might be considered as the last possible option in a severely obese adolescent with obesity-related comorbidity. Such a procedure should be undertaken only in an experienced surgical centre after extensive consultation, lengthy education of the patient and their family, and full psychological assessment. Continuing post-operative care inan experienced weight-management service would be mandatory

Kelly et al, 2005 (USA)*

Research question Establish evidence-based guidelines for best practices for surgical care in weight loss surgery (WLS)

Patient population BMI >40 kg/m2 or BMI >35kg/m2 with obesity-related medical comorbidities

Interventions compared Weight loss surgery: BPD (with or without DS), gastric bypass (include jejunoileal bypass and RYGB) and LAGB.


BPD with or without DS Associated with metabolic and nutritional

complicationsJejunoileal bypass

Unacceptable level of serious complicationsRYGB

Seems to offer the best balance of effectiveness vs. risk LRYGB is known to have a longer and more complex

learning curve than other laparoscopic techniques LRYGB improves short-term recovery from surgery and

has lower incidence of incisional hernias than open RYGB

LAGB lower average mortality rates than RYGB or

malabsorptive proceduresEfficacy outcomes reported(point estimates and statistical ranges)

BPD Effective in inducing weight loss, particularly in ‘super-

obese patients (BMI >50), but can cause significant complications

BPD/DS Effective in producing weight loss and to reverse

obesity-related comorbiditiesRYGB

Greater long-term weight loss than gastric partitioning alone or VBG

Open & laparoscopic RYGB produce similar short term weight loss and improvements in comorbid medical conditions

Long limb (>150cm) RYGB may produce superior short term weight loss in patients who are >200lb overweight or BMI≥50kg/m2 (however benefit of long limb decreases over time)

VBG Long-term have proven disappointing Inferior weight loss compared with RYGB

LAGB Produces variable short-term weight loss and

improvements in obesity-related comorbidities

142




Short-term adverse outcomes after LRYGB may be related, in part, to the more complex learning curve. The procedure needs to be performed by appropriately trained, qualified laparoscopic bariatric surgeon

LAGB should continue to be offered on a controlled basis at comprehensive weight loss centres that use appropriate patient selection criteria while more long-term data are accured

The role of VBG is limited and procedure has been largely supplanted by LAGB

A minimum of bimonthly interaction among members of the multidisciplinary team to review and discuss patient management and other pertinent topics

Updating of patient selection criteria to reflect current advances in technology and ongoing refinements in surgical techniques

High-volume surgeons (50 to 100 cases/yr) operating in properly equipped, high-volume weight loss centres (>100 cases/yr) with integrated and multidisciplinary treatment

Other key points Data on long term safety and metabolic side effects of BPD with DS limited, therefore procedure considered investigational

Placement of the LAGB in the pars flaccida path rather than retrogastric position may reduce incidence of postoperative complications

Kelly et al, 2009 (USA)*

Research question Update best practice guidelines for surgical care, with a continuing focus on long-term outcomes and patient safety

Patient population Not specifiedInterventions compared RYGB (open and laparoscopic), BPD and DS, LAGB, LSG, VBGSafety outcomes reported(point estimates and statistical ranges)

RYGB Benefits of LRYGB includes less postoperative pain,

shorter LOS, fewer postoperative abdominal wall complications, faster convalescence, reduction in rate of incisional hernia and wound infection.

Operative time may be longer, leading to increased risk of internal hernia

Long limb RYGB and VVLL-RYGB may increase risk of protein and micronutrient deficiencies

Banded RYGB may be subject to long term complications related to reintervention, reoperation and QoL; insufficient evidence to make recommendation

More data is needed on Long-term drawbacks of mini-gastric bypass

BPD BPD is likely to cause long-term nutritional and vitamin

deficiencies. DS appears to decrease incidence of severe protein malnutrition, marginal ulcer and dumping syndrome.

LSG Early reports find LSG safe and with relatively low

morbidity

143




RYGB Established long-term effectiveness for sustained

weight loss, reduction of comorbidities (including consistent and effective relief of GORD), and low risk for long-term nutritional sequelae

Weight loss similar between laparoscopic and open RYGB

Yet to determine whether LL-RYGB or VVLL-RYGB produce superior weight loss

BPD and DS Short-term data indicate that BPD is the most effective

surgery in terms of weight loss (>RYGB or LAGB) BPD and DS produce effective weight loss in patients

with BMI >50, weight loss may be superior to RYGB Possibly better resolution of comorbidities (diabetes

and hypercholesterolemia)LAGB

Short term data show promising outcomes with LAGB but long term studies raise questions on durability and reoperative rates

VBG associated with increased peri- and postoperative complications compared with LAGB; should not be used as a primary surgical treatment for obesity

LSG Early reports find LSG effective, with marked weight

loss and reduction in major obesity-related comorbidities


BPD and DS require lifelong patient follow-up Performance of LAGB should take place in accredited,

multidisciplinary settings by experienced surgeons with advanced laparoscopic skills, including those needed to revise LAGB to an alternative procedure

Monitoring of long-term data and continuation of current practice patterns, with yearly follow-up of patients is recommended

At minimum, WLS programs should be able to provide appropriate referrals to facilities that can provide advanced laparoscopic skills

LSG – several short term studies suggest safe and effective weight loss but long term data on safety and efficacy needed to recommend the approach

Obesity medicine specialists, nurse practitioners, physician assistants, residents, and bariatric nurse specialists can safely adjust bands under the supervision of a weight loss surgeon

VBG is associated with increased peri- and postoperative complications compared with LAGB, and should not be used as a primary surgical treatment for obesity

144



Other key points Revisional WLS can address unsatisfactory weight loss or complications after primary WLSl may also enhance weight loss and further improve comorbidities

Complications, LOS, and mortality are higher for revisional WLS, but can be safe and effective when performed by experienced weight loss surgeons

Combination procedures lead to greater EWL and resolution of comorbidities than restrictive procedures

Age may remain an independent risk factor for complications following WLS but procedures can be safe and effective in patients over 60

McTigue et al, 2003 (USA)*

Research question Examine evidence for screening and treating obesity in adultsPatient population Adults with BMI exceeding 40 or of 35 or more with associated

severe health complications and have not responded to other treatment methods

Interventions compared Obesity screening, behavioural/counselling interventions, pharmacotherapy interventions and surgical approaches (including gastric bypass, adjustable gastric bands, VBG)


Perioperative mortality rate 0% to 1.5% for both VBG, AGB and gastric bypass

VBG Reoperation: 20%-25% over 3-5 yrs wound infection: 8%-32% Less common events – gastric leak, stomal stenosis,

pouch dilatationsGastric bypass

Wound infection: 8%-20% Other complications include staple failure (15%),

vitamin B12 deficiency (40%), diarrhoea (13%) and gastrointestinal hemorrhage (3%)

Gastric banding Reoperation: 1%-20% Other complications include band dislocation, leakage,

or slippage Low rates of dysphagia, hunger, vomiting and

esophagitis One reported fewer surgical complications with

laparoscopic vs open procedures In terms of site of band placement, there are conflicting

data about the relative safety of esophagogastric vs. gastric placement

Up to 25% of surgical patients require surgery again over 5 years

RCTs did not report mortality

145




Weight loss post surgery – 10 to 159 kg over 12 to 48 months

Gastric bypass – mean reduction 45-65kg Gastroplasty – mean reduction 30-35kg CTFPHC (4 randomised trials and 1 prospective cohort

study) – mean weight loss 17-46kg after 2 to 5 yrsSwedish Obese Subjects (SOS)

Weight loss: 28kg (CI 26.9 to 29.1) for surgical patients vs 0.5kg (CI -0.2 to 1.2kg) for nonsurgical controls 2 years post surgery

Mean weight reduction – 21% (12% SD) for gastric banding; 23% (10% SD) for vertical banded gastroplasty; 33% (10% SD) for gastric bypass

8 yrs – average weight loss of 20kg (CI, 18 to 22kg) in 251 surgical patients and 0.7kg (CI, -0.8 to 2.2kg) in 232 controls

Comorbidities 90% follow up of 200 patients (50% glucose intolerant,

50% diabetic) – 91% had normal fasting glucose and glycosylated haemoglobin levels

Dramatic improvements in glucose metabolism, lipid profiles and blood pressure

Hypertension tended to recur within 3-10 yrsKey conclusions and recommendations

Surgical options clearly have the highest risk compared to counselling and pharmacological interventions

Only surgical options consistently result in substantial long-term weight reduction, however, they carry risk for complications (sometimes severe complications) and are expensive

In pooled samples, surgery led to death in less than 1% of patients, but up to 25% of patients may require reoperation over 5 years

Body size, health status and weight loss history may influence obesity treatment

Other key points Because of practical and ethical constraints to a true randomized, blinded, placebo-controlled trial of surgery for obesity, high quality evidence is limited. Systematic reviews of obesity therapy primarily examined randomised unblended trials comparing surgical techniques (nonsurgical controls not included).

None of the trials examined showed statistically significant different weight loss between groups, but all treatments promoted considerable loss.

Pratt et al, 2009 (USA)*

Research question Update evidence-based best practice guidelines for paediatric/adolescent weight loss surgery

Patient population Adolescents with BMI ≥ 35 and specific obesity-related co-morbidities and adolescents with BMI≥40

Interventions compared RYGB (gastric bypass), LAGB (adjustable gastric band), BPD, LSG

146




Early WLS may reduce obesity-related mortality and morbidity

Adolescents and children may be psychologically immature and risk of decreased compliance and long-term follow up increases

Adolescent girls particularly vulnerable to nutritional deficiencies and should receive special attention

Risk of getting pregnant increases after WLS AGB has not been approved by FDA for use in

adolescentsEfficacy outcomes reported(point estimates and statistical ranges)

T2DM, obstructive sleep apnoea, non-alcoholic fatty liver disease and non-alcoholic seatophepatitis, pseudotumor cerebri are strong indications for early weight loss surgery in adolescents

CVD and predictors of metabolic syndrome (e.g., high waist circumference etc) are not strong enough indications to recommend early WLS

WLS may bring important benefits to emotional health and quality of life in extremely overweight adolescents

Depression and eating disorders are not exclusion criteria for WLS


Recommend weight loss surgery selection criteria to include adolescents with BMI ≥ 35 and specific obesity-related co-morbidities for which there is clear evidence of important short-term morbidity (i.e., T2DM, severe steatohepatitis, pseudotumor cerebri, and moderate-to-sever obstructive sleep apnoea) and adolescents with extreme obesity (BMI ≥40)

RYGB considered safe and effective option for extremely obese adolescents as long as appropriate long term follow up is provided

BPD and DS procedures cannot be recommended in adolescents

SG should be considered investigational – existing data not sufficient to recommend

Other key points Patient selection When combination procedures are used in adolescents,

physical maturity should be documented (usually children 12yrs+)

Psychological maturity should be assessed prior to WLS BMI cut points in children and adolescents who meet

other criteria should be ≥35 with major co-morbidities and ≥40 with other co-morbidities

Children & adolescents should demonstrate ability to comply with treatment regimens and medical monitoring before WLS, especially children with mental disabilities

Staffing Ideal WLS team should include a minimum of 4-5

professionals and have at least one preoperative face-to-face meeting to prepare a treatment plan for each patient

Note: *Guidelines reviewed along with other systematic reviews and meta-analyses. Source: Deloitte Access Economics.

147



Table B.8: Data extracted from clinical guidelines

Apovian et al, 2005 (USA)

Research question To establish evidence-based guidelines for best practice in paediatric/adolescent weight loss surgery

Patient population Adolescents 11 to 22 years old, BMI 42 to 55, follow up range 2 months to 10 years

Interventions compared RYGB, LAGB, BPD with DS, jejunoileal bypass and VBGSafety outcomes reported(point estimates and statistical ranges)

No statistical ranges reported. Various complications for each surgery are listed including: deaths, renal failure, sleep apnoea, wound infection, anaemia, cholelithiasis etc.


Post operative BMI down to 28 to 33. Weight loss between 62% and 87%.


Although weight loss surgery is an acceptable alternative for weight loss in severely obese adults, no conclusions have been made about the appropriateness of WLS for individuals <18 years of age.

A multidisciplinary team of paediatric specialists is needed for optimal preoperative decision making and postoperative management.

Research needs for the future should include prospective data collection and interpretation of long-term outcomes of adolescents undergoing WLS, especially for the newer, less invasive procedures such as the LABG.

Other key points Current literature on paediatric WLS is insufficient to determine how outcomes in adolescents may differ from those in adults who undergo WLS.

Recommend a biennial peer review process for all programs offering WLS to adolescents, with ongoing collaborative discussion, sharing of techniques, and updating of standards among all programs.

August et al, 2008 (USA)

Research question To formulate practice guidelines for the treatment and prevention of paediatric obesity.

Patient population Paediatrics – age not specifiedInterventions compared LAGB and RYGB. Also provides guidelines for lifestyle

modifications and pharmacotherapy.

148




Because of the high morbidity and mortality associated with jejunoileal bypass and the BPD with DS, these procedures cannot be recommended for use in children.

The LAGB procedure is considered safer than RYGB, but the FDA has not yet approved LAGB for use in adolescents.

The safety profile (moderate evidence) for LAGB after a follow-up period of 1 to 85 months revealed no operative or postoperative deaths; 26 of 328 patients required reoperation to correct complications (band slippage, intragastric migration, and port/tubing problems).

The safety profile (moderate evidence) for RYGB after a follow-up period of 2 wk to 6 yr revealed a combination of mild (slight malnutrition) and severe (pulmonary embolism, severe malnutrition, postoperative bleeding, and gastrointestinal obstruction) complications.


Conclusions reached in this assessment, mostly based onweak evidence, were that:

both RYGB and LAGB resulted in clinically significant weight loss (7% of body weight);

LAGB resolved the co-morbid conditions of diabetes and hypertension, whereas RYGB resolved hypertension (insufficient data to rate the resolution of other comorbidities)

149




Suggested that bariatric surgery be considered only under the following conditions:

The child has attained Tanner 4 or 5 pubertal development and final or near-final adult height.

The child has a BMI > 50 kg/m2 or has BMI above 40 kg/m2 and significant, severe co-morbidities.

Severe obesity and co-morbidities persist despite a formal program of lifestyle modification, with or without a trial of pharmacotherapy.

Psychological evaluation confirms the stability and competence of the family unit.

There is access to an experienced surgeon in a medical centre employing a team capable of long-term follow-up of the metabolic and psychosocial needs of the patient and family, and the institution is either participating in a study of the outcome of bariatric surgery or sharing data.

The patient demonstrates the ability to adhere to the principles of healthy dietary and activity habits.

Recommend against bariatric surgery for preadolescent children, for pregnant or breast-feeding adolescents, and for those planning to become pregnant within 2 yr of surgery; for any patient who has not mastered the principles of healthy dietary and activity habits; for any patient with an unresolved eating disorder, untreated psychiatric disorder, or Prader-Willi syndrome.

It must be clear to the patient and the family that bariatric surgery is an adjunct to a sincere commitment to alteration of lifestyle and behaviour rather than a cure. All obese children must first demonstrate their ability to adhere to a family-based dietary and lifestyle modification program.

Other key points Suggestion made in this paper for limited use of bariatric surgery places a relatively higher value on avoiding anatomical and functional changes in developing children, on avoiding unforeseen complications associated with lifelong exposure to these changes, and on avoiding the costs and perioperative complications of these procedures. It places a relatively lower value on the weight loss and amelioration of obesity-related complications associated with bariatric surgery.

Baur et al, 2010 (Australia and New Zealand)

Research question Development of a position paper on bariatric surgery on adolescents to guide decisions as to which adolescents should receive such surgery and how they should best be managed.

150



Patient population The small proportion of severely obese adolescents will require bariatric surgery within the context of an ongoing and coordinated multidisciplinary approach.Patient criteria for selection for bariatric surgery:

Minimum age of 15 years (14 years in exceptional circumstances)

Attainment of Tanner stage 4 or 5 pubertal development

Attainment of final or near-final adult height Severe obesity (BMI >40 kg/m2 and BMI >35 kg/m2 in

the presence of severe obesity-associated complications

Presence of the level of obesity despite involvement in a formal multidisciplinary and supervised program of lifestyle modification and pharmacotherapy. A minimum of six months of supervised multidisciplinary therapy should be provided prior to bariatric surgery being performed

The adolescent and family understand, and are motivated to participate in, the on-going treatment, lifestyle change and review following surgery

Informed consentRecommend against bariatric surgery for:

Adolescents under the age of 14 years Pregnant or breast-feeding adolescents Patients with significant cognitive disabilities Patient with untreated or untreatable psychiatric or

psychological disorder Patients with Prader-Willi syndrome and other similar

hyperphagic conditionsBinge eating disorder is not a contraindication to either LAGB or RYGB, however, surgical intervention should be coupled with psychological intervention.

Interventions compared LAGB and RYGBSafety outcomes reported(point estimates and statistical ranges)

LAGB No in-hospital or postoperative deaths were reported in

all eight studies Reoperations were performed in 8% of subjects, most

commonly for band slippage Eight subjects suffered iron deficiency Mortality rate of 0.05% compared to 0.5% for RYGB Revisional surgery in up to 10% of cases, however, the

majority of these surgeries can be performed laparoscopically

RYGB Four deaths were reported between 9 months to 6

years post surgery Most frequently reported complications relate to

protein-calorie malnutrition and micro-nutrient deficiency

151




LAGB 95% CI for weight loss between 1-3 years is -13.7 to -

10.6 BMI units. This compares favourably with non-surgical intervention programs

Diabetes was assessed in two studies with resolution rates of 100% and 80%.

Hypertension was assessed in three studies with resolution rates of 50%, 100% and 100%.

Decreased effectiveness is offset by the lower morbidity and mortality rates of LAGB compared to other procedures.

RYGB 95% CI for weight loss ranges from -17.8% to -22.3 BMI

units. Compare very favourably to non-surgical interventions in morbidly obese adolescents.

Resolution of hypertension ranged from 50% to 100% Obstructive sleep apnoea was assessed in two studies

with resolution rate of 100%Key conclusions and recommendations

Surgery is to be undertaken by an experienced bariatric surgeon (ideally with experience in the management of patients in the adolescent age group), which is affiliated with a team experienced in the assessment and long-term follow-up of the metabolic and psychosocial needs of the adolescent bariatric patient and family.

The institution should be either participating in study of the outcomes of bariatric surgery, or sharing such data in a proposed national registry of bariatric surgery and patient outcomes.

LAGB is the primary bariatric surgical procedure of choice for adolescents as it has good weight based outcomes, has a low complication rate and is potentially reversible.

Follow-up should be on a 4-6 weekly basis with long-term follow-up to extend beyond 10 years, and ideally for the whole of life

Other key points Complication rates are from published historical data; with subsequent improvements in devices and surgical techniques, complications rates my have improved.

Buchwald, 2005 (USA)

Research question Statement on the state of bariatric surgery for morbid obesityPatient population Adults (age not specified) with morbid obesity, also referred to

as “clinically severe obesity” or “extreme obesity,” defined as the criteria for bariatric surgery by the 1991 NIH Consensus Conference Statement on Gastrointestinal Surgery for Severe Obesity as a BMI >40 kg/m2 or a BMI >35 kg/m2 in the presence of high-risk co-morbid conditions.

152



Interventions compared Four operative procedures (in three classes of procedures), are currently in general use in the UnitedStates and worldwide:

gastric bypass with a standard, long-limb, or very long-limb Roux (restrictive and malabsorptive), alone or in combination with VBG;

LAGB (restrictive); VBG (restrictive); and BPD and DS (primarily malabsorptive).


Operative (30-day) mortality for gastric bypass when performed by skilled surgeons is about 0.5%. Operative morbidity (e.g., pulmonary emboli, anastomotic leak, bleeding, and wound infection) is about 5%.

Operative (30-day) mortality for LAGB when performed by skilled surgeons is about 0.1%. Operative morbidity is about 5%.

Operative mortality for VBG when performed by skilled surgeons is about 0.1%. Operative morbidity is about 5%.

Operative mortality for BPD and DS when performed by skilled surgeons is about 1%. Operative morbidity is about 5%.


Weight loss after a standard RYGB usually exceeds about 65% to 70% of the excess body weight (EBW) and about 35% of the BMI.

Weight loss after LAGB is about 50% of the EBW and about 25% of the BMI at 2 years.

Weight loss after VBG is about 50% to 60% of EBW, and about 25% to 30% of BMI.

Weight loss after BPD and DS is about 70% of the EBW and about 35% of the BMI.


Bariatric surgery is the most effective therapy available for morbid obesity and can result in improvement or complete resolution of obesity co-morbidities. Both open and laparoscopic bariatric operations are effective therapies for morbid obesity and represent complementary state-of-the-art procedures.

Other key points Bariatric surgery has been performed in morbidly obese adolescents for more than a decade. In these small series, surgical weight loss resulted in considerable improvement, if not complete resolution, of most obesity-related co-morbidities, supporting the position that bariatric surgery in adolescents is reasonable. Long-term efficacy, potential adverse consequences related to decreased absorption of nutrients, and degree of recidivism remains unknown.

BMI guidelines for adolescents should be identical to those advocated for adults. Deferring surgery to a higher BMI standard may increase operative mortality and morbidity, and possibly prevent reversal of co-morbid conditions.

Fried et al, 2007 (European countries)

153



Research question The aim of the Guidelines is to provide physicians, health-care policy makers, and health-care carriers and insurance companies with essential elements of good clinical practice in the treatment of morbid obesity.

Patient population In patients aged 18-60 years:1) with BMI ≥40 kg/m2;2) with BMI 35-40 kg/m2 with co-morbidity in which

surgically-induced weight loss is expected to improve the disorder.

In adolescents with severe obesity:bariatric surgery can be considered if the patient:

1) has a BMI >40 (or 99.5th percentile for respective age)2) and at least one co-morbidity, 3) Failed at least 6-12 months of organised weight-

reducing attempts4) Committed to medical and psychological evaluation and

postoperative treatment program5) Have access surgery affiliated with specialist paediatric

supportBariatric surgery above age 60 years: Considered on an individual basisContra-indications specific bariatric surgery:

1) Absence of periods of identified medical management;2) A patient who is unable to participate in prolonged

medical follow-up;3) Non-stabilized psychotic disorders, severe depression

and personality disorders, unless specifically advised by a psychiatrist experienced in obesity;

4) Alcohol abuse and/or drug dependencies;5) Diseases threatening life in the short-term;6) Patients who are unable to care for themselves and

have no long-term family or social support that will warrant such care.

Interventions compared Food limitation (restrictive) operations VBG SG adjustable and non-adjustable gastric banding Gastric bypass: proximal and long-limb Operations limiting absorption of nutrients and energy BPD Combined operations BPD-DS Distal gastric bypass (common limb ≤100 cm)

(interventions not compared, only noted)Safety outcomes reported(point estimates and statistical ranges)

Not statistically reported. The authors found surgical complexity and potential

surgical and long-term metabolic risks of procedures decrease in reverse order to weight loss and weight gain achieved with the various procedures

Laparoscopic technique should be considered as the first treatment choice in bariatric surgery, unless specific contraindications to a laparoscopic operation are present.

154




Not statistically reported. The expected average weight loss and weight maintenance increases with the following procedures: AGB, VBG, GBP, BPD-DS, BPD


At the moment, there are no sufficient evidence-based data to suggest how to assign a patient to any particular bariatric procedure

Bariatric procedures should be performed in interdisciplinary obesity management centres with appropriately trained staff and adequate equipment

Bariatric surgeon’s experience is a key issueOther key points This article also provides details on follow-up pathways after

surgery. Laville et al, 2005 (France)

Research question The French public health-care insurer asked the medical associations involved in obesity management to provide guidelines for obesity surgery

Patient population Surgical treatment is intended for subjects with major obesity that is having repercussions on health, and who have received detailed information and presents an acceptable surgical risk. The indications are:

morbid obesity, i.e. body mass index (BMI) >40 kg/m2 resistant to medical treatment and exposing subjects to serious complications, not controlled by a specific treatment;

obesity with a BMI between 35 and 40 kg/m2, if comorbidities exist which threaten life or the functional prognosis;

the procedure is not indicated in children and adolescents, except in exceptional circumstances determined in units specialized in nutrition and paediatrics with an expert analysis of the psychological situation; and

The indications in subjects >60 years of age must be considered very carefully, because the surgical risks and the dangers of weight loss are increased (loss of lean muscle mass and its consequences), whereas the risk of mortality linked to obesity is not increased in this age group.

Interventions compared There are two main types of operation practiced: one with simple gastric restriction (calibrated vertical gastroplasty and adjustable bands), the second favouring intestinal malabsorption (gastric bypass or isolated BPD).


Risk to life during the perioperative period: mortality figures vary from 0.1 to 0.5% (intervention not stated).


Due to the lack of comparative studies, it is not possible to choose between these two techniques based on cost/efficacy grounds. Procedures creating malabsorption produce greater weight loss than that obtained by gastric restriction alone.

155




To assess bariatric treatment, it is recommended to: record the surgical procedure in the medical

information system; ensure that good practice is carried out; produce registries and carry out an assessment of the

medico-economic factors and of satisfaction; and monitor the safety of the materials used.

Other key points To ensure high quality of care, it is recommended that: patients and doctors should be provided with

information on practising and monitoring this surgery; referral units should be identified which have the

multidisciplinary teams and equipment necessary for managing the most serious forms of obesity;

recommendations should be developed on anesthesia and intensive care for obese subjects;

medical imaging, lifting and transport equipment as well as beds should be adapted for patients whose corpulence is incompatible with existing standard models;

access to care by people in financial difficulty should be assisted and ensured.

Mechanick et al, 2008 (USA)

Research question Focus on the nonsurgical aspects of perioperative management of the bariatric surgery patient, with special emphasis on nutritional and metabolic support.

Patient population Candidates for bariatric surgery: BMI ≥ 40 or a BMI >35 with high-risk comorbid conditions such as life-threatening cardiopulmonary problems or uncontrolled T2DM. Other possible indications for patients with BMIs between 35 and 40 include obesity-induced physical problems interfering with lifestyle.

Currently, a consensus does not exist on the possible contraindications to bariatric surgery. Suggested contraindications include:

Extremely high operative risk (severe congestive heart failure or unstable angina)

Active substance abuse Major psychopathologic condition Patients who cannot comprehend the nature of the

surgical intervention and the lifelong measures required to maintain an acceptable level of health

Interventions compared LAGB, RYGB, BPD (with or without DS), staged bariatric surgical procedures and SG

156




LAGB Complication and mortality rates are lower than for

RYGB (0.1% vs 0.5%) safe among patients >55 years of age complications include band slippage, band erosion,

balloon failure, port malposition, band and port infections, and oesophageal dilatation

BPD may be associated with protein-calorie malabsorption may be associated with a variety of nutrient

deficiencies and metabolic derangements, such as iron deficiency anemia, deficiencies in the fat-soluble vitamins, and metabolic bone disease.

BPD-DS morbidity and mortality were increased in patients with

a preoperative BMI>65 also to have various postoperative nutritional and

metabolic complicationsEfficacy outcomes reported(point estimates and statistical ranges)

LAGB associated with substantially better maintenance of

weight loss associated with significantly less excess weight loss

than RYGB at 5 years associated with less loss of fat-free mass compared

with RYGB and BPD initial BMI<45 and the presence of postprandial satiety

postoperatively are associated with greater weight loss after LAGB

RYGB weight loss achieved is greater than that attained with

pure gastric restrictive proceduresBPD- DS

weight loss is comparable to that for patients with RYGB

Staged bariatric surgical procedures the first stage, a restrictive procedures such as a SG can

be associated with a 33%-45% loss of excess body weight at 1 year

SG one randomised, prospective trial has shown better

weight loss compared to LAGB at 3 yearsKey conclusions and recommendations

The purpose of bariatric surgery is to induce substantial, clinically important weight loss that is sufficient to reduce obesity-related medical complications to acceptable levels.

Procedures performed by more experienced surgeons were associated with much lower risk of death. Bariatric surgery is not uniformly a ‘low-risk’ procedure, and judicious patient selection and diligent perioperative care are imperative

Other key pointsNational Institutes of Health (NIH), 1998 (USA)

157



Research question To identify, evaluate, and summarize published information about the assessment and treatment of overweight and obesity;

To provide evidence-based guidelines for physicians, other health care practitioners, and health care organizations for the evaluation and treatment of overweight and obesity in adults; and

To identify areas for future research.Patient population Surgical interventions in adults with a BMI > 40 or a BMI >35

with comorbid conditionsInterventions compared Dietary therapy, physical activity, pharmacotherapy, behaviour

therapy, and surgery.Safety outcomes reported(point estimates and statistical ranges)

Not statistically reported. Extremely obese persons often do not benefit from the

more conservative treatments for weight loss and weight maintenance.

Obesity severely impairs quality of life, and these individuals are at higher risk for premature death.

The National Institutes of Health Consensus Development Conference consensus statement, “Gastrointestinal Surgery for Severe Obesity” concluded that the benefits outweigh the risks and that this more aggressive approach is reasonable in individuals who strongly desire substantial weight loss and have life-threatening comorbid conditions.


Not statically reported. Weight loss due to surgical intervention such as the

gastric bypass ranged from 50 kg to 100 kg over 6 months to 1 year.

Gastroplasty with diet had a favourable net outcome on weight loss after 2 years compared to diet alone.

VBG was more effective than horizontal- banded gastroplasty.

Gastric resection with a modest BPD without intestinal exclusion resulted in significantly greater weight loss than conventional RYGB; this long-limb modification of RYGB was shown to be safe and effective in patients who were 90 kg or more overweight and did not cause additional metabolic sequelae or diarrhoea.


Surgical intervention is an option for carefully selected patients with clinically severe obesity (a BMI >40 or a BMI >35 with comorbid conditions) when less invasive methods of weight loss have failed and the patient is at high risk for obesity-associated morbidity and mortality.

Other key points These guidelines also provide a detailed overview of overweight and obesity in the US including Health and Economic costs.

National Institutes of Health (NIH), 1996 (USA)

Research question Following 2 days of presentations by experts and discussion by the audience, a consensus panel weighed the evidence and prepared their consensus statement.

Patient population Not specified

158



Interventions compared Nonsurgical approaches to treatment of clinically severe obesity include various combinations of low- or very low-calorie diets, behavioural modification, exercise, and pharmacologic agents.Surgical: VBG and related techniques and RYGB


Not statistically reported. Immediate operative mortality rate for both VBG and RYGB is relatively low. On the other hand, morbidity in the early postoperative period, i.e., wound infections, dehiscence, leaks from staple line breakdown, stomal stenosis, marginal ulcers, various pulmonary problems, and deep thrombophlebitis in the aggregate, may be as high as 10% or more. In the later postoperative period, other problems may arise and may require reoperation. These are pouch and distal oesophageal dilation, persistent vomiting (with or without stomal obstruction), cholecystitis, or failure to lose weight. Moreover, mortality and morbidity rates with reoperation are higher than those of primary operations.In the long-term, micronutrient deficiencies, particularly of vitamin BI2, folate, and iron, are common after gastric bypass and must be sought and treated. Another potential result of this operation is the dumping syndrome, which is characterized by gastrointestinal distress and other symptoms. Occasionally, these symptoms may not respond to conservative measures and may be troublesome to the patient. Many data suggest that deficient nutrition in pregnancy comes with it a high risk of foetal damage or loss.


Not statically reported.The two major types of present operations for severe obesity are VBG and RYGB. The success rate for weight loss has been reported to be slightly higher with RYGB. Substantial weight loss generally occurs, with the weight nadir occurring in 18 to 24 months. Some regain of weight is common by 2 to 5 years after operation. A third operation, BPD, about which there are only limited data, also has been reported to produce weight loss but with a higher frequency of metabolic complications.


Decisions on what therapy to recommend to patients with clinically severe obesity should depend on their wishes for outcomes, on the physician's judgment of the urgency of the need for therapy, and on the physician's judgment of possible options for therapy and their probable efficacy.

Other key points One of the key problems in evaluating the current reports of case series in surgical therapy is the lack of standards for comparison. The present practice is to compare postoperative indicators of comorbidity to the same patient's own preoperative status. Although this approach may give some useful information on short-term effects of surgical therapy, it is insufficient for evaluation of long-term effects and of survival. An alternative approach for evaluating surgical therapy is to compare levels of morbidity and mortality in the surgical group with an appropriate comparison group. The establishment of a meaningful comparison group presents a challenge to future research.

National Institute for Health and Clinical Excellence (NICE), 2006 (UK)

159



Research question This is the first national guidance on the prevention, identification, assessment and management of overweight and obesity in adults and children in England and Wales. The guidance aims to:

stem the rising prevalence of obesity and diseases associated with it;

increase the effectiveness of interventions to prevent overweight and obesity; and

improve the care provided to adults and children with obesity, particularly in primary care.

Patient population Age not specified.Bariatric surgery is recommended as a treatment option for adults with obesity if all of the following criteria are fulfilled:

a BMI of 40 kg/m2 or more, or between 35 kg/m2 and 40 kg/m2 and other significant disease (for example, T2DM or high blood pressure) that could be improved with weight loss

all appropriate non-surgical measures have been tried but have failed to achieve or maintain adequate, clinically beneficial weight loss for at least 6 months

the person has been receiving or will receive intensive management in a specialist obesity service

the person is generally fit for anaesthesia and surgery commits to the need for long-term follow-up.

Children who: have achieved or nearly achieved physiological maturity have a BMI ≥ 40 kg/m2 with serious obesity-related

comorbidities or have a BMI of ≥ 50 kg/m2 with less severe comorbidities

Interventions compared Surgery versus non-surgical interventions LAGB versus gastric bypass (comparative studies). LAGB versus DS and BPD (comparative studies) DS-BPD versus gastric bypass (comparative studies) Laparoscopic gastric bypass versus open gastric bypass

(comparative studies)

160




ADULTS- In one comparative study, LAGB was associated with similar rates of early complications to LGBP, but with higher levels of late complication. Mortality was nil in both groups (n = 1).Reoperation rates were higher in the LAGB group compared with the LGBP group (26.2% vs 10.7% overall) (n = 1).Reoperation rates were similar in both laparoscopic and open GBP procedures, but late complications were more common in the open GBP group (24% vs 11%). From observational studies, reoperation rates were as follows, median (range):

LAGB: 6.5% (0.5% to 24%) LGB: 1.8% (0.03% to 9.8%) Open GB: 5% (2.8% to 12%) BPD-DS: 3.9% (2.7% to 6.3%)

From the observational studies, 2.3% of laparoscopic adjustable bands were removed, 0.06% of the LGBPs were reversed, and approximately 4.75% (median, range 3.8% to 6.8%, n = 4) of DS-BPDs were revised; the majority for lengthening of the common limb for nutritional problemsFrom observational studies, mortality rates were as follows, median (range):

LAGB: 0.0% (0 % to 0.6%) LGB: 0.4% (0% to 1.1%) Open GB: 0.5% (0% to 1.5%) BPD-DS: 0.5% (0% to 1.4%)

DS-BPD and RYGB have similar rates of wound infection (22% vs. 20%), postoperative anastomotic leaks (6% vs. 3%) and mortality (0.9% vs. 0.8%).Staged surgery is an appropriate surgical option for people with BMI > 50 kg/m2, but the evidence on weight loss and other outcomes remains limited.In the four trials comparing laparoscopic and open GBP conversion rates from laparoscopic to open surgery ranged from 2.5% to 23%.Complication rates (minor 7.6% vs. 11.8% and major 7.6% vs. 9.2%) were similar in both laparoscopic and open GBP procedures, and early complications were similar for both techniques, but complication rates after 30 days were lower in the laparoscopic group.CHILDREN – evidence suggests that severely obese children and adolescents who undergo bariatric surgery may develop micronutrient deficiencies and other postoperative complications.Adolescents and children who undergo bariatric surgery (more common in gastric bypass) may require revisional surgery, or may develop other late postoperative complications such as cholecystitis or hernias.Some older studies have reported deaths due to perioperative and postoperative complications. There are no reports of deaths in recent studies.One recent study reported band slippage, port infection and replacement of a leaking port in adolescents who underwent LAGB.

161




ADULTS – Evidence supports the use of surgery for weight loss in people for whom surgery is an appropriate optionOne RCT showed that at 12 months, the use of an exceptional diet, with intensive follow-up (approximately 30 contacts per year initially) and support (outpatient clinic visits and group meetings) can achieve similar results to surgery (–18.0 kg vs. –22.0 kg) in people with at least 60% excess weight, but these results are not maintained at 24 months.Surgery remains more effective than a non-surgical approach for people who are obese (BMI ≥ 38 kg/m2 for women, ≥ 34 for men) in the longer term (measured up to 10 years after surgery)%EWL

LRYGB >LAGB at all measured time points BPD-DS and RYGB show similar rates of % EWL at 12

and 24 monthsFrom observational studies At 24 months, median (range)

LAGB: 54.5% (38% to 87%) LGB: 69% (67% to 83%) Open GB: 65% (55% to 71%) BPD-DS: 71.5% (67% to 78%)

At 60 months, median (range) LAGB: 54% (44% to 66%) LGB: 82% Open GB: 57% (56% to 58%) BPD-DS: 69% (66% to 73%)

CHILDREN – There is no evidence on which surgical procedure is the most effective in achieving weight loss in adolescents.Evidence suggests that bariatric surgery should only be performed in obese adolescents who have systematically failed to manage weight for 6 months or more as determined by primary care provider.Evidence appears to suggest that an approximate change in BMI of –20 kg/m2 (after approximately 2 years) can occur in obese adolescents who underwent bariatric surgery.Evidence appears to suggest a varying median excess weight loss ranging from 15.9% at 6 months to 69% at 24 months for LAGB, and from 62% at 12 months to 87% at 2 years in adolescents who underwent gastric bypass.

162




ADULTS - Bariatric surgery is recommended as a treatment option for adults with obesity if all of the following criteria are fulfilled:

they have a BMI of 40 kg/m2 or more, or between 35 kg/m2 and 40 kg/m2 and other significant disease (for example, T2DM or high blood pressure) that could be improved if they lost weight.

all appropriate non-surgical measures have been tried but have failed to achieve or maintain adequate, clinically beneficial weight loss for at least 6 months

the person has been receiving or will receive intensive management in a specialist obesity service

the person is generally fit for anaesthesia and surgery the person commits to the need for long-term follow-

upBariatric surgery is also recommended as a first-line option (instead of lifestyle interventions or drug treatment) for adults with a BMI of more than 50 kg/m2 in whom surgical intervention is considered appropriate. CHILDREN - Surgical intervention is not generally recommended in children or young people. Bariatric surgery may be considered for young people only in exceptional circumstances, and if they have achieved or nearly achieved physiological maturity.Surgery for obesity should be undertaken only by a multidisciplinary team that can provide paediatric expertise in: preoperative assessment, including a risk–benefit analysis

that includes preventing complications of obesity, and specialist assessment for eating disorder(s)

information on the different procedures, including potential weight loss and associated risks

regular postoperative assessment, including specialist dietetic and surgical follow-up

management of comorbidities psychological support before and after surgery information on or access to plastic surgery (such as

apronectomy) where appropriate access to suitable equipment, including scales, theatre

tables, Zimmer frames, commodes, hoists, bed frames, pressure-relieving mattresses and seating suitable for patients undergoing bariatric surgery, and staff trained to use them.

surgical care and follow-up should be coordinated around the young person and their family’s needs and should comply with national core standards as defined in the Children’s NSFs for England and Wales.

all young people should have had a comprehensive psychological, education, family and social assessment before undergoing bariatric surgery.

a full medical evaluation including genetic screening or assessment should be made before surgery to exclude rare, treatable causes of the obesity.

163



Other key points The guidance supports the implementation of the ‘Choosing health’ White Paper in England, ‘Designed for life’ in Wales, the revised GP contract and the existing national service frameworks. It also supports the joint Department of Health, Department for Education and Skills and Department for Culture, Media and Sport target to halt the rise in obesity among children under 11 by 2010, and similar initiatives in Wales.

Sauerland et al, 2005 (European countries)

Research question The aim and focus of these guidelines cover key questions regarding effective and efficient surgical treatment of obesity, including patient selection, choice of surgical technique, management of complications and follow-up.

Patient population Adults – age not specifiedInterventions compared AGB, VBG, RYGB and BPDSafety outcomes reported(point estimates and statistical ranges)

Statistical ranges not reported Complications after LAGB include gastric erosion, band

slippage, pouch dilation, occlusion of the stoma, and port-related complications. Gastric erosion usually causes mild pain, various types of infections and prevents further weight loss.

After VBG, the range of complications includes stoma stenosis, pouch dilatation, band erosion and staple line disruption. Erosion or infection of the band at the pouch outlet should be treated by band removal. In severe cases, conversion to LAGB or other procedures may be necessary.

Stoma stenosis, gastric distension, anastomotic leakage, gastrojejunal ulcers and nutritional deficiencies may occur after RYGB. Stoma stenosis due to anastomotic strictures usually occurs during the first postoperative months.

The spectrum of complications after BPD is similar to RYGB. Complications have been found to be more likely in patients converted from other procedures to BPD.


In the long-term after BPD, patients typically loose between 65% and 75% of their excess body weight.

RYGB usually results in 60% to 70%, but the procedure is much better accepted in the US (about 70% of all procedures) as compared to Europe.

Postoperative weight reductions for VBG range between 55% and 65% nadir EWL.

Weight loss is less in LAGB compared to other procedures and usually reaches only 45% to 55%


AGB, VBG, RYGB and BPD are all effective in the treatment of morbid obesity, but differ in degree of weight loss and range of complications. The choice of procedure therefore should be tailored to the individual situation. There is evidence that a laparoscopic approach is advantageous for LAGB, VBG, and GB (and probably also for BPD).

164



Other key points Since obesity surgery has various competing aims, such as weight loss, adjustability, reversibility, and safety, it is difficult to draw universally valid conclusions about the optimal bariatric procedure. For all types of surgery, there is overwhelming evidence from case series on safety, efficacy, and effectiveness in terms of weight loss, but much less data are available on the comparative evaluation of different bariatric procedures. Therefore, the decision must be taken with the patient’s individual situation and the surgeon’s expertise in mind.

Snow et al, 2005 (USA)

Research question The intent of this guideline is to provide recommendations based on a review of the evidence on pharmacologic and surgical treatments of obesity.

Patient population These guidelines do not specify a population. However the Swedish Obese Subjects (SOS) study is referred to. Participant of the SOS study were obese adults with (BMI >34 kg/m2 for men and >38 kg/m2 for women).

Interventions compared RYGB, BPD, LAGB and VBGSafety outcomes reported(point estimates and statistical ranges)

Five randomized, controlled trials were identified thatcompared weight loss between or among surgical procedures and reported enough data for pooling

Early mortality rates for RYGB range from 0.3% (95% CI, 0.2% to 0.4%) for case series data to 1.0% (95% CI, 0.5% to 1.9%) in controlled trials

AGB had an associated early mortality rate of 0.4% (95% CI, 0.01% to 2.1%) for controlled trials and 0.02% (95% CI, 0% to 0.78%) for case series data. No statistically significant differences in mortality were seen among procedures.

Early mortality rates following bariatric surgery are 1% or less in published controlled trials and case series data (which come from a specific clinic or surgeon performing procedures on patients enrolled in a research study).


For participant of the Swedish Obese Subjects (SOS) study: at 8 years of follow-up, average weight loss was 20 kg for surgically treated patients and average weight did not change for medically treated patients. Patients treated with RYGB lost more weight than those treated with VBG or banding procedures.

For participants of five randomized, controlled trials the weight loss results support the conclusion that gastric bypass produces weight loss superior to that produced by gastroplasty procedures. In 2 other randomized, controlled trials, the weight lost using VBG compared with LAGB was 14 kg more at 12 months of follow-up but only about 3 kg more at 36 months of follow-up.

165




Surgery should be considered as a treatment option for patients with a BMI of 40 kg/m2 or greater who instituted but failed an adequate exercise and diet program (with or without adjunctive drug therapy) and who present with obesity-related comorbid conditions, such as hypertension, impaired glucose tolerance, diabetes mellitus, hyperlipidemia, and obstructive sleep apnoea. A doctor–patient discussion of surgical options should include the long-term side effects, such as possible need for reoperation, gall bladder disease, and malabsorption.

Other key points Patients should be referred to high-volume centres with surgeons experienced in bariatric surgery

Society of American Gastrointestinal and Endoscopic Surgeons (SAGES), 2008 (USA)

Research question To guide surgeons applying laparoscopic techniques to the practice of bariatric surgery

Patient population Adults (age not specified)Interventions compared

LBPD; LRYGB; and LAGB.Safety outcomes reported(point estimates and statistical ranges)

LBPD: dramatically impacts comorbidities. At least 90% of patients with T2DM will cease diabetic medications by 12–36 months. Of hypertensive patients 50–80% will be cured, with another 10% experiencing improvement. Up to 98% of patients with obstructive sleep apnoea symptoms will have resolution. The 30-day mortality of early LBPD series ranges from 2.6 to 7.6%. Major complications, which occur in up to 25% of cases, may include early occurrence of anastomotic leak, duodenal stump leak, intra-abdominal infection, hemorrhage, and venous thromboembolism, or later bowel obstruction, incarceration or stricture.LRYGB: The mortality rate after RYGB ranges from 0.3% in case series to 1.0% in controlled trials, and the rate of preventable and non-preventable adverse surgical events is 18.7% The mortality rate in a review of selected LRYGB series ranged from 0.5% to 1.1%. Safety of LRYGB has been compared to open RYGB, with laparoscopic patients having reduced incidence of iatrogenic splenectomy, wound infection, incisional hernia and perioperative mortality, but higher rates of bowel obstruction, intestinal hemorrhage, and stomal stenoses. The most frequently reported perioperative complications associated with LRYGB are wound infection (2.98%), anastomotic leak (2.05%), gastrointestinal tract hemorrhage (1.93%), bowel obstruction (1.73%), and pulmonary embolus (0.41%), while the most frequently reported late complications are stomal stenosis (4.73%), bowel obstruction (3.15%), and incisional hernia (0.47%).LAGB: case series and systematic reviews put early mortality rates after LAGB at 0.05–0.4%, compared with LRYGB at 0.5–1.1%, open RGB at 0.5–1.0%, open BPD at 1.1%, and LBPD at 2.5–7.6%. Regarding relative morbidity rates, comparative data are few. Overall complications and major complications are less common in LAGB than LRYGB or LBPD, in a single-centre experience.

166




LBPD initiates dramatic weight loss during the first 12 postoperative months, which continues at a slower rate over the next 6 months. Weight loss is durable up to at least 5 years postoperatively. 95% of patients with BMI>50 kg/m2, and 70% of those with BMI>50 kg/ m2, achieve greater than 50% EBWL. Weight may be regained over time, highlighting the importance of long-term follow-up.LRYGB patients who undergo LRYGB typically experience 60–70% EBWL, with >75% control of comorbidities. In general, these outcomes are better than banding procedures, which have 45–50% EBWL and less predictable improvement of comorbidities, but are less than BPD ± DS which has 70–80% EBWL with excellent control of comorbidities.LAGB is very effective at producing weight loss, with patients losing approximately 50% of their excess body weight. This weight loss occurs in a gradual manner, with approximately 35% EBWL by 6 months, 40% by 12 months, and 50% by 24 months. This percentage appears to remain stable after 3–8 years based on the few studies providing this length of follow-up. However, as many as 25% of LAGB patients fail to lose 50% of their excess body weight by 5 years. The short-term (< 12 months) weight loss of LAGB is inferior to RGB. This discrepancy is seen to continue, with a randomized controlled trial illustrating that EBWL at 5 years was 47.5% for AGB versus 66.6% for RYGB. Still, life-threatening complications are less frequent in LAGB as compared to LRYGB.


Given the marked paucity of prospectively collected comparative data between the different bariatric operations, it remains impossible to make definitive recommendations for one procedure over another. At the present time, decisions are driven by patient and surgeon preferences, as well as considerations regarding the degree and timing of necessary outcomes versus tolerance of risk and lifestyle change.

Other key points Adolescent bariatric surgery (age < 18 years) has been proven effective but should be performed in a specialty centre. Patient selection criteria should be the same as used for adult bariatric surgery.

Source: Deloitte Access Economics

167



Appendix C: Overview of the primary studies identified for the economic literature reviewA data extraction template was used to summarise the key information in each of the included economic studies. Information was extracted by one senior reviewer specialising in economic studies. Summary of data extracted from each of the economic study are listed in Table C.1.

Table C.1: Data extracted from economic studies

Study Country(s) of analysis, population

Interventions Modelling approach Study type, perspective

Data sources Price year, annual discount rate

Key results

Ackroyd et al 2006

France, Germany, UKBMI 35+ with T2DM

Gastric bypass, AGB, conventional management (CM, 1 year guided diet, 4 years watchful waiting)

Deterministic linear algorithm model, 5 year horizon using actual reported data. T2DM prevalence extrapolated for years 4-5Utility change per BMI unit assumed the same regardless of starting BMI

CUA, healthcare payer

Clinical data from HTA reports, recent RCTs and prospective series data. Author opinion for resource use. Local unit costs

2005 €/£, costs/ outcomes discounted at 3.5%

Incremental QALYs vs. CM = 1.03 (banding), 1.34 (bypass). Results also presented by BMI years and T2DM-free years. Incremental costs vs. CM = €5030 for bypass, €-3,586 for banding (Germany); €-5,877 for bypass, €-4,480 for banding (France); £2,033 for bypass, £1,984 for banding (UK). Surgeries dominated CM in France and Germany. ICURs for surgery vs. CM in the UK were £1,517 (bypass) and £1,929 (banding). In a worst case scenario banding no longer dominated in France (ICUR = €1,379) and UK ICURs increased to £2,599 (bypass) and £3,251 (banding)

168






Key results

Ananthapavan et al 2010

AustraliaSeverely obese adolescents (14-19 years, BMI 35+)

LAGB, conventional management

Case series data (n=28) extrapolated to Australian eligible patient population (n=4,120) over lifetime using BMI at 3 years post surgery (assumed to remain constant over lifetime) and future cost assumptions

CEA, ‘societal’ (healthcare payer plus patient time and travel costs)

Case series data (retrospective patient audit) for BMI change and costs at 3 years

2001 Australian dollars, costs/DALYs discounted at 3%

Cost per DALY saved = $4400 (95% CI $2900-$6120). Post-operative BMI loss would have to reduce to 5% of the modelled level before LAGB is no longer considered cost-effective (i.e. above $50,000 per DALY gained)

Anselmino et al 2009

Austria, Italy, SpainBMI 35+ with T2DM

Gastric bypass, AGB, conventional management (1 year guided diet, 4 years watchful waiting)

Same model as Ackroyd et al 2006


Same data as Ackroyd et al 2006, with author opinion for local resource use and local unit costs

2009 €, costs/ outcomes discounted at 3.5%

For clinical outcomes see Ackroyd et al 2006. Incremental costs vs. CM per 1000 pts were €-1.938m for bypass, €-2.942m for banding (Austria); €-1.670m for bypass, €-1.107m for banding (Italy); €3.570m for bypass, €1.497m for banding (Spain). Surgeries dominated CM in Austria and Italy. ICURs vs CM for Spain were €2664 (bypass), €1456 (banding). In a worst case scenario banding and bypass no longer dominated CM in Italy (ICURs = €638 and €94) and Spain ICURs increased to €4347 (bypass) and €3142 (banding)

169






Key results

Campbell et al 2008

USPeople with a BMI 35+

LAGB, LRYGB, no treatment

Lifetime Markov model. No treatment is assumed to confer no weight change

CUA, perspective not stated but appears to be the healthcare payer

Single head to head RCT of LAGB vs. LRYGB

2006 US dollars, costs/ outcomes discounted at 3%

Woman aged 40 years with BMI of 35–40: LAGB has lower average costs than LRYGB for the initial procedure (US$15,470 vs. US$23,160) and complications (US$3680 vs. US$11,930), but results in less weight loss. ICURs are US$13,990 (LAGB) and US$14,690 (LRYGB) vs. no treatment; US$16,540 for LRYGB vs. LAGB. Woman BMI 40–50: respective ICURs are US$4860, US$5150, US$5780. ICURs generally higher for men due to shorter life expectancies. Both procedures are effective and cost effective; the choice should be based on patient and provider preferences

Campbell et al 2010

USPeople aged 18-74 years with BMI 40+ or 35+ with comorbidities

LAGB, LRYGB, no treatment

Lifetime Markov model. No treatment assumed to confer no weight change

CUA, third party payer

Single head to head 5-year RCT of LAGB vs. LRYGB. Sensitivity analysis used a systematic review of 36 studies. Utilities linked to BMI via a single study. Other published data for costs and mortality risks by BMI category


ICURs for surgery vs. no treatment are all below US$27,000, and for LRYGB vs. LAGB are all below US$15,000. Surgery is more cost effective for women, higher BMI, and younger people. PSA shows all ICURs for surgery vs no surgery and LRYGB vs. LAGB to be below US$50,000. ICUR for LRYGB vs. LAGB is sensitive to model assumptions; long term quality studies are needed to more accurately calculate ICURs. For now, the choice of procedures should be based on patient/provider preferences

170






Key results

Clegg et al 2002

UKCohort with average age 40 years, average BMI 45, 90% female (full report focuses on those with BMI 40+ or 35+ with significant co-morbidities)

RYGB, VBG, AGB, non-surgical management

Markov model with 20 year time horizon. Little information on model structureChange in comorbidities confined to reversal of diabetes in base case; comorbidity data taken from the RCTs only showed a lasting impact on diabetes

CUA, UK National Health Service

Weight loss (%) each year and changes in comorbidities from systematic review of literature (no meta-analysis or combined data). Utilities by BMI from a NICE submission of orlistat. Surgeries micro-costed using RCT data and assumptions about associated resource use

2000 £, costs/QALYs discounted at 6%/1.5%

ICURs vs non-surgical intervention were £10,237 (VBG), £8,527 (band), £6,289 (RYGB). ICURs comparing surgeries were £6,176 (band vs. VBG), £742 (bypass vs. VBG), £256,856 (band vs. RYGB). Bypass appears the preferred option, though studies for different interventions have different durations; some future benefits of banding may not be captured whereas longer term efficacy is proven for RYGB. In seven worst case scenarios, the ICUR for RYGB vs. no surgery remained below £21,000.Surgery appears cost effective and all surgeries increase costs and QALY relative to non-surgical management

171






Key results

Craig & Tseng 2002

USSeverely obese, aged 35-55 years, BMI 40-50. Non smokers, no CVD, no major psychological problems, no drug addictions, unsuccessful with conservative interventions

Gastric bypass, no treatment

Decision tree. Revisional and reversal surgery includedSensitivity analyses were arbitrary. Underlying study for obesity costs of excluded cancer and musculoskeletal disease


Excess weight loss (%) derived from single study with 14 years follow up. Mortality and cost data from a published US study


ICURs ranged between US$5,400 (woman 55 years BMI 50) to US$35,600 (men 55 years BMI 40). Costs per LYG were substantially higher. Bypass is generally more cost effective for women and higher BMIs. In older less obese men, ICERs were sensitive to the cost and duration of surgery and disutility with obesity. Results were robust in women

172






Key results

Hoerger et al 2010

USBMI 35+ with T2DM

Bypass, banding, usual care for diabetes

Markov lifetime model (Centers for Disease Control-RTI International Diabetes Cost Effectiveness Model) with adaptations for bariatric surgery including its cost and operative mortality, quality of life (based on unit decline in BMI), and diabetes remission/relapse rates

CUA, perspective not stated but appears to be healthcare payer

Parameters used to adapt the model populated with data from published meta-analyses, long term study of surgery outcomes (SOS), Medstat claims data


ICURs for bypass vs no surgery were US$7,000 (newly diagnosed), US$12,000 (established T2DM). ICURs for banding vs. no surgery were US$11,000 (newly diagnosed), US$13,000 (established T2DM). Higher ICURs for established T2DM are mainly due to lower survival gains with surgery. PSA for newly diagnosed patients shows 95% of ICURs fall below US$23,000 (bypass) and US$30,000 (banding)

Ikramuddin et al 2009

UST2DM. Population not stated as obese but presumably were

RYGB, medical management

CORE diabetes model: Markov with Monte Carlo simulation, 35 year horizon

CUA, third party payer

Baseline characteristics and risk factors from US cohort were the same as in Minshall et al (2009a/b). Previous studies used for costs, utilities, and adverse events of surgery


Bariatric surgery conferred an additional 0.9 QALYs, 0.67 LYs, US$19,760 total costs. ICERs were US$29,676/LYG and US$21,973/QALY gained. ICERs mainly driven by time horizon (ICUR = US$122,001 at 10 years) and utility loss for high BMI

173






Key results

Jensen & Flum 2005

USMorbidly obese white women

Gastric bypass at age 40 years with BMI 40+ (cure), diet/exercise at age 18 years with BMI 35+ (prevention)

Lifetime Markov model from birth to death

CUA, societal (medical costs, patient time, caregiver time)

Indirect comparison of two studies. US study reporting life expectancy and obesity related costs by BMISelection and applicability of efficacy studies is questionable

2004 US dollars, discounting not reported

Bypass had an ICUR of US$7,126 (approximately US$4,600 additional cost and 0.61 additional QALYs). Worst case ICUR was below US$35,000

Keating 2009a AustraliaBMI 30-40, with recently diagnosed (<2 years) T2DM

LAGB, medical management

Within trial (2 year) analysis, Australian unit costs applied to resource use (intervention costs only)

CEA, healthcare system

Single RCT, Australian unit cost data

2006 Australian dollars, no discounting

Mean resource costs per patient were $13,383 (LAGB) and $3,396 (medical management). ICER was $16,600 per additional case of diabetes remitted

Keating 2009b Australia obesity (BMI 30-40) with recently diagnosed (<2 years) T2DM

LAGB, medical management

Markov lifetime analysis using 2-year RCT data (Keating et al 2009a) and an assumed annual diabetes relapse rate

CUA, healthcare system

Single RCT, Australian data for intervention costs, diabetes costs, and utilities for diabetes and the general population

2006 Australian dollars, costs/QALYs discounted at 3%

LAGB conferred an additional 0.7 LYs, 1.2 QALYs, 9.4 years in remission and a cost saving of $2,444. LAGB dominated medical management. Surgery costs recouped by other cost savings after 10 years. Results would improve if benefits of weight loss were captured in addition to diabetes remission

174






Key results

Maklin et al 2009

FinlandPopulation not stated but presumably the morbidly obese

Bariatric surgery (pools gastric bypass, SG, gastric banding), ordinary treatment

Markov model (health states: alive, re-operation, abdominoplasty, death) with 10 year horizonbased on extensive follow up data. Unclear how BMI and associated events were incorporated

CUA, healthcare provider

Large population survey on quality of life and resource use. Register data, wider literature, expert opinion

€, price year and discounting not reported

Bariatric surgery dominated ordinary treatment. Mean costs were €31,800 (surgery) and €44,800 (ordinary. Mean QALYs were 7.05 (surgery) and 6.51 (ordinary). Results robust to extensive sensitivity analyses

McEwen et al 2010

USManaged care population undergoing bariatric surgery

Bariatric surgery (33% LRYGB, 64% open RYGB, 3% other), no surgery

Prospective cohort study/before and after study, outcomes extrapolated to 2 years or lifetime


Data collected 12 months after surgery. Baseline (pre-surgery) EuroQol (EQ-5D) and VAS utilities collected from another cohort requesting surgeryAssumptions for how costs would change over lifetime

US dollars, price year not reported, costs/QALYs discounted at 3%

Incremental costs/QALYs/ICUR with surgery were US$13,626/0.28/US$48,662 (2 years), US$2,505/1.76/US$1425 (lifetime). Surgery is cost saving after 18 years

175






Key results

Minshall 2009a France, Germany, UKObese with T2DM


Lifetime horizon, model included chronic heart failure, death, peripheral neuropathy, background diabetic retinopathy, macular oedema, angina and myocardial infarction


Baseline characteristics, risk factors, and 2.3 year observational data from US cohort (n=204), other data from published sources

2009 £, 2008 €, costs/ outcomes discounted at 3.5% (UK), 5% (Germany), 3% (France)

Cost per LYG and ICUR for RYGB vs medical management were £4,128 and £2,922 (UK), €401 and €298 (France), cost saving (Germany). Primary drivers of the ICERs were HbA1c, lipid changes, and weight loss

Minshall 2009b

Italy, Spain, SwedenObese with T2DM


Lifetime horizon CUA, healthcare payer

See Minshall et al 2009a

2008 €/SEK, costs/ outcomes discounted at 6% (Spain), 3% (Italy and Sweden)

Cost per LYG and ICUR for RYGB vs medical management were €3,807 and €2,034 (Spain), €490 and €364 (Italy), SEK32,823 and SEK24,437 (Sweden). Primary drivers were reported by Minshall et al 2009a

MSAC 2003 AustraliaObese people, implied to be the same as the Clegg et al 2002 cohort. Full report focuses on BMI 35+

LAGB, VBG, open RYGB

Clegg et al 2002 analysis with following parameters updated: procedure times, re-operation costs, inpatient days. Costs of adjustments for LAGB were also included

CUA, private hospitals

Clegg et al 2002, more recent data identified for some parameters (see modelling approach). Estimated 5.8 adjustments per LAGB procedure

Australian dollars, price year not reported, no discounting required (procedure cost only without revisions or complications)

Procedure costs were $9,121 (LAGB), $8209 (RYGB), $5,456 (VBG). For LAGB vs. VBG, the maximum ICUR implied is $26,178 using Clegg et al’s (2002) incremental gain of 0.14 QALYs. ICUR for LAGB vs. RYGB not estimated due to concerns over which has superior benefits (RYGB shows greater weight loss and mortality)

176






Key results

Paxton and Matthews 2005

USPeople undergoing surgery for weight loss

LRYGB, open RYGB

Meta-analysis of relevant studies to derive surgical safety and long-term efficacy (%EWL at 3 years). Costs applied to complications

CEA, restricted societal (healthcare system and income losses)

Studies of the two interventions published from 1984-2004. US cost data for complications, and hospital stay and income lost due to complications

2004 US dollars, no discounting

Excess weight loss (%) similar but complication rates lower with LRYGB. Although procedure costs are higher there are savings of US$2,783. LYRGB is therefore dominant. Long term effects of both procedures may be similar because they both rely on the same surgical alterations to produce their effects

177






Key results

Picot et al 2009

UKCohort with average age 40 years, average BMI 45, 90% female, or BMI 37+ with T2DM. Full report focuses on morbidly obese defined as BMI 40+ or 35+ with significant comorbidities

Laparascopic gastric bypass, LAGB, non-surgical treatment (primarily monitoring)

Markov model with 20 year horizon, including T2DM, stroke, CHD, dead, remission of comorbidity, no comorbidity

CUA, NHS and Personal Social Services

Weight loss (%) in each year from systematic review (not meta-analysed or combined). Targeted search for studies reporting association between BMI and events. Framingham risk equations used for CHD/stroke. Resource use from Clegg et al 2002 and additional economic evaluations. Utility values from a single study showing utility change with BMI change

2008 £, costs/QALYs discounted at 3.5%

ICUR (optimistic/pessimistic): BMI 40+ = £1,897/3,863 (LAGB), £3,160/4,127 (bypass), sensitivity analysis all <£10,000. BMI 30-40 with T2DM: £1,367 (LAGB), sensitivity analysis all below £5000. BMI 30-35: £12,673 (LAGB), sensitivity analysis all below £34,000. Sensitivity analysis not reported for bypass. Extensive deterministic sensitivity analyses, PSA, and results at different time points are presented

178






Key results

Salem et al 2008

USMales and females with BMI 40-60 aged 35-55 years

LAGB, LRYGB, non operative interventions

Decision tree based on Craig & Tseng (2002). Lifetime analysis. LAGB patients could undergo band removal. Outcomes at three years include peri-operative mortality, lifetime with initial BMI, weight loss


Event probabilities and surgery outcomes were average values from a literature review of studies reporting data to 3 years. Utilities from Craig & Tseng (2002) survey. Lifetime medical costs and survival based on Framingham and National Health and Nutrition Examination Survey studies and 5 complications of obesity


ICUR: men 35 years BMI 40 = US$11,604 (LAGB), US$18,543 (LRYGB); women 35 years BMI 40 = US$8,878 (LAGB), US$14,680 (LRYGB). LAGB more cost effective than LRYGB for all base-cases (men and women, ages 35, 45 and 55, BMI 40, 50 and 60) and sensitivity analyses. ICUR for LAGB was most sensitive to weight loss, operation cost, and frequency of band removal. ICUR for LRYGB was most sensitive to operative mortality rate, weight loss and operation cost. All ICURs below US$25,000 for all base cases. LAGB is more cost effective despite conferring lower weight loss

179






Key results

Siddiqui et al 2006

USPatients with BMI 35+

LRYGB, open RYGB

Decision tree evaluating complications and mortality following the procedure and within one year (surgical complications only)

CEA, healthcare payer

Literature published since 1990 reporting mortality and morbidity in the year following surgery

2004 US dollars, discounting not required (one year analysis)

In three BMI groups, LRYGB was associated with greater success (no complications), lower mortality, lower costs, and dominated. Incremental success/mortality/cost was 4%/-0.3%/-US$4,002 (BMI 35-49); 5%/-0.4%/-US$5,658 (BMI 50-59); 4%/-0.5%/-US$9,001 (BMI 60+). Sensitivity analyses evaluate surgical complication rate thresholds at which the dominating procedure changes. Difference in weight loss is not an issue because different methods of the same procedure are being compared. For the super obese these procedures have not been compared in an RCT; the study extrapolates available data to compare these using a decision model

180






Key results

van Gemert 1999

NetherlandsClinically severe obesity BMI 40+

VBG, no VBG Before and (two years) after study (ITT analysis)QALY gain appears to be calculated using 0.25 additional utility for normal life expectancy (48 years) plus an additional 3.6 years

CUA, healthcare provider

21 study patients. QoL data (Nottingham Health Profile, VAS). Netherlands total healthcare costs with population attributable fraction for clinically severe obesity applied to life expectancy. Interviews on productivity before/after surgery. Life expectancy following surgery derived using DEALE method. Mortality from a single much older paper

US dollars, price year not reported, ‘all values’ discounted at 5%

VBG vs no VBG: 3.6 LYG, 12 QALYs gained (based on a 0.25 increase in the VAS). The VBG procedure cost (US$5,865) was outweighed by cost of illness savings: US$8,029/US$7,118 when clinically severe obesity prevalence was 0.25% /1%. VBG was dominant even without considering productivity gains

181






Key results

van Mastrigt 2006

NetherlandsBMI 40+, or 35-40 with significant comorbidity

LAGB, VBG Within trial analysis (one year), bootstrapping analysis

CUA, societal

Excess weight loss (EWL %) at 12 months, EuroQol (EQ-5D) utility at 3/6/12 months. Actual billing costs. Productivity loss and informal care costs calculations/data not reported

1999 €, discounting not required (one year analysis)

Total cost difference with LAGB was not significant (€-1,843; 95% CI: €-5,999, €1,765). EWL significantly higher with VBG (17.82%; 95% CI: 9.60%, 26.05%). QALYs were lower with VBG but difference was not significant (-0.05; 95% CI: -0.117, 0.016). ICER for VBG was €105.80 per additional 1% EWL. However, for QALY outcomes, LAGB dominates in most replications. Because the cost and QALY differences are not significant, the treatment choice should be based on efficacy, safety, clinical aspects, and long term cost effectiveness


182

Limitation of our workThis report is prepared solely for the use of Department of Health and Ageing. This report is not intended to and should not be used or relied upon by anyone else and we accept no duty of care to any other person or entity. The report has been prepared for the purpose of reviewing Medicare Benefits Schedule (MBS) Items for the surgical treatment of obesity. You should not refer to or use our name or the advice for any other purpose.

Deloitte Access Economics 184

Contact usDeloitte Access EconomicsACN: 49 633 116

Level 19 Sydney AvenueBarton ACT 2600PO Box 6334Kingston ACT 2604 Australia

Tel: +61 2 6175 2000Fax: +61 2 6175 2001

www.deloitte.com/au/economics

About DeloitteDeloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee, and its network of member firms, each of which is a legally separate and independent entity. Please see www.deloitte.com/au/about for a detailed description of the legal structure of Deloitte Touche Tohmatsu Limited and its member firms.Deloitte provides audit, tax, consulting, and financial advisory services to public and private clients spanning multiple industries. With a globally connected network of member firms in more than 150 countries, Deloitte brings world-class capabilities and deep local expertise to help clients succeed wherever they operate. Deloitte's approximately 170,000 professionals are committed to becoming the standard of excellence

About Deloitte AustraliaIn Australia, the member firm is the Australian partnership of Deloitte Touche Tohmatsu. As one of Australia’s leading professional services firms, Deloitte Touche Tohmatsu and its affiliates provide audit, tax, consulting, and financial advisory services through approximately 5,000 people across the country. Focused on the creation of value and growth, and known as an employer of choice for innovative human resources programs, we are dedicated to helping our clients and our people excel. For more information, please visit Deloitte’s web site at www.deloitte.com.au.Deloitte Access Economics is Australia’s pre-eminent economics advisory practice and a member of Deloitte's global economics group. The Directors and staff of Access Economics joined Deloitte in early 2011.Liability limited by a scheme approved under Professional Standards Legislation.Member of Deloitte Touche Tohmatsu Limited

© 2011 Deloitte Access Economics Pty Ltd

http://www.deloitte.com/au/economics

Documents

Report - Department of Health · Web viewGiven that bariatric surgery is a significant procedure with potential complications, the provision of bariatric surgery should be targeted