Final Appraisal Determination: Surgery to aid weight reduction for people with morbid obesity

For the purposes of this guidance, people are defined as having morbid obesity if they have a body mass index (BMI) either equal to or greater than 40 kg/m², or between 35 kg/m² and 40 kg/m² in the presence of significant co-morbid conditions that could be improved by weight loss.

People with morbid obesity who are considering surgery to aid weight reduction should discuss in detail with the clinician responsible for their treatment (that is, the hospital specialist and/or bariatric surgeon) the potential benefits and longer-term implications of surgery, as well as the associated risks, including complications and postoperative mortality.

Surgery is recommended as a treatment option for people with morbid obesity providing all of the following criteria are fulfilled:

• this type of surgery should be considered only for people who have been receiving intensive management in a specialised hospital obesity clinic
• individuals should be aged 18 years or over
• there should be evidence that all appropriate and available non-surgical measures have been adequately tried but have failed to maintain weight loss
• there should be no specific clinical or psychological contra-indications to this type of surgery
• individuals should be generally fit for anaesthesia and surgery
• individuals should understand the need for long-term follow-up.

Surgery should be undertaken only after comprehensive, multidisciplinary assessment. In addition, arrangements should be made for appropriate healthcare professionals to provide pre-operative and postoperative counselling and support to individuals being considered for surgery.

Given the uncertainty surrounding the evidence for the relative safety and effectiveness of different surgical interventions, it is not possible to distinguish between them on grounds of cost effectiveness. The choice of surgical intervention should therefore be made jointly by the individual and the clinician after considering the best available evidence, the facilities and equipment available, and the experience of the surgeon who would perform the operation.

It is important that services and skills to support surgery for people with morbid obesity are developed in a planned and coordinated way. Existing collective mechanisms for specialist commissioning groups may offer an appropriate way for the development of detailed implementation strategies. Each group, in discussion with other groups, should decide how best to develop and expand the service, and the number of sites from which it should be provided.

Obesity occurs when a person's calorie intake repeatedly exceeds the amount of energy expended. This relationship may be influenced by genetic, social, cultural, psychological, environmental and economic factors. Morbid obesity is internationally defined as a BMI greater than or equal to 40 kg/m², where BMI is calculated as body weight in kilograms divided by height in metres squared. For the purposes of this guidance, morbid obesity also includes those with a BMI between 35 and 40 kg/m² with significant co-morbidity that could be improved by weight loss.

In 1998, an estimated 0.6% of men and 1.9% of women in England and Wales had a BMI of 40 kg/m² or more - this is equivalent to 124,000 men and 412,700 women or 2500 people for a typical primary care trust population of 200,000. The prevalence of obesity is rising as the average BMI increases; between 1994 and 1998 the average BMI increased by 0.44 kg/m² for men and 0.57 kg/m² for women.

Obesity is associated with increased morbidity and mortality. It is a risk factor for cardiovascular disease, hypertension, type 2 diabetes, cancer, musculo-skeletal disease, reproductive disorders and respiratory disorders. In addition, people with a BMI greater than 35 kg/m² have a rate of mortality at any given age double that of someone with a healthy BMI (range 20-25 kg/m²).

Obesity is also associated with decreased quality of life. There is a social stigma attached to obesity and those affected often face prejudice and discrimination. Obesity has a negative impact on mobility, productiveness, employment and psychosocial functioning, with many obese people left feeling depressed, defensive and unable to live life to the full.

Obesity has considerable costs for society, both direct, in terms of healthcare, and indirect, in terms of earnings lost through mortality or sickness. Direct costs of obesity in England in 1998 have been estimated at £480 million, or 1.5% of NHS expenditure, and indirect costs at £2.1 billion.

In the NHS, obesity is usually managed within primary care with advice on weight control, diet, physical exercise and lifestyle. Other interventions such as drug therapy, referral to specialist weight-loss clinics, behavioural therapy and low-calorie and very-low-calorie diets may be considered.

There is evidence to suggest that even modest weight loss is associated with substantial benefits, particularly in terms of the reduction of co-morbidities. Many non-surgical approaches result in an initial weight loss of 10-15%, with the majority of people regaining this weight within 1 year. Modest weight loss that is maintained for some time (around 5% at 12 months and 3% at 2-3 years) has been reported in a specialist clinic, with maintenance being enhanced by exercise and behavioural therapies. However, there is little or no evidence concerning the effectiveness of conventional treatment of obesity in the primary care setting in England and Wales. Additionally, much of the evidence available on weight loss using non-surgical treatments has been obtained from randomised controlled trials of obese people (BMI greater than or equal to 30 kg/m²), but not specifically of people with morbid obesity.

Surgery to aid weight reduction (bariatric surgery) may be considered when all other measures have failed. Key aims of this type of surgery are to achieve weight reduction and maintain any loss through restriction of intake and/or malabsorption of food. Bariatric surgery is not commonly undertaken in England and Wales; currently, little over 200 operations are performed annually, and many of these are privately funded.

There are two main types of surgical intervention - malabsorptive and restrictive. With malabsorptive surgery, parts of the gastrointestinal tract are bypassed so that the absorption of food is limited. With restrictive surgery, the size of the stomach is restricted so the person experiences the feeling of fullness with less food. Malabsorptive procedures include jejunoileal bypass, gastric bypass and biliopancreatic diversion, while restrictive procedures include gastroplasty and gastric banding (see Appendix D for descriptions of each procedure). Traditionally, gastric surgery has been carried out as an open procedure, but increasingly laparoscopic techniques are being used.

Detailed pre-operative assessment supported by established selection criteria are thought to be essential for achieving and maintaining outcomes. Failure to lose weight prior to surgery should not exclude patients from surgery.

Postoperative care is crucial. Care should be available to manage complications as they occur, and patients should receive dietary and, where appropriate, psychological advice to help them modify their eating habits to promote weight loss and to prevent complications such as vomiting, dumping syndrome and diarrhoea.

The average cost of the various surgical interventions, including pre-operative assessment and postoperative care in the first year, is in the region of £4500-£5300.

The Appraisal Committee considered evidence from a number of sources (see Appendix B).

Seventeen randomised clinical trials (RCTs) and one non-randomised clinical trial were found. Two RCTs and the non-randomised clinical trial compared surgical interventions with conventional treatment. The remaining 15 RCTs compared different types of surgery. The methodological quality of the studies varied.

In an RCT comparing horizontal gastroplasty (n = 27) with a very-low-calorie diet (n = 30), there was no significant difference in weight loss at 12 months (23 kg vs 18 kg, p > 0.05), although at 24 months, patients who had undergone gastroplasty had lost significantly more weight (32 kg vs 9 kg, p < 0.05). In the Danish Obesity Project (DOP) trial comparing jejunoileostomy (n = 130) with medical management (n = 66), of patients remaining in the study those treated surgically (n = 101) had lost significantly more weight at 24 months (42.9 kg vs 5.9 kg, p < 0.001) than the 33 people treated medically. In the Swedish Obese Subjects (SOS) cohort study comparing surgery (vertical banded gastroplasty, gastric banding and gastric bypass, n = 1210) with conventional treatment (n = 1099), patients treated surgically had lost significantly more weight after 2 years than patients managed conventionally, whose weight remained unchanged (weight loss 23% vs 0%, p < 0.001). This weight loss was maintained at 8 years and was equivalent to a difference in weight loss of 20.7 kg between those patients treated surgically and those managed conventionally.

Four postoperative deaths were reported in the surgical arms of the trials (three from leakage of gastrointestinal contents and one resulting from a technical laparoscopic mistake), compared with two deaths in patients treated non-surgically (one following liver biopsy and one after gastric bypass surgery carried out 4 years later). Peri-operative complications were reported by all three studies and included subphrenic abscess (7%), pneumonia (4%), wound infection (4-6%), pulmonary complications (3-6%) and hepatic dysfunction (1.5%). No patients required re-operation in the gastroplasty trial, while 0.7% of jejunoileostomy patients had their procedure reversed in the DOP trial and 2.2% of surgical patients underwent re-operation in the SOS study.

In the SOS study, patients treated surgically had statistically significant improvements in all health-related quality-of-life measures at 2 years compared with patients treated conventionally (i.e. without surgery). In addition, at 8 years, the odds ratio of developing hypertension was 1.01 (95% confidence interval [CI] 0.61 to 1.67) and of developing diabetes was 0.16 (95% CI 0.07 to 0.36) for those patients treated surgically compared with those managed conventionally.

In summary, surgery for people with morbid obesity is associated with significant weight loss that is maintained for at least 8 years, whereas there is little sustained weight loss with conventional treatment in this group of patients. Surgery is also associated with improved quality of life and reduced co-morbidities. There are significant risks attached to surgery, although these are thought to be outweighed by the benefits.

Eight RCTs compared gastric bypass with gastroplasty. In general, gastric bypass resulted in significantly more weight loss than gastroplasty. Postoperative deaths were reported in 5 of 339 patients undergoing gastric bypass and in 1 of 456 patients undergoing gastroplasty. Revisions, re-operations and/or conversions were less common following gastric bypass than gastroplasty (0-39% vs 9-53%). Complications for gastric bypass patients included symptomatic ulcer disease (25%), intractable vomiting and stomal stenosis (25%), dumping syndrome (28%) and heartburn (59%). Complications of gastroplasty included superficial stomal erosions (5%) and heartburn (32%).

Two RCTs compared gastric bypass with jejunoileal bypass. Both studies found less weight loss with gastric bypass. However, 80% of patients undergoing jejunoileal bypass in one trial showed liver damage, with 7% suffering severe liver disease.

One RCT compared vertical banded gastroplasty with horizontal gastroplasty, and found significantly greater weight loss with vertical banded gastroplasty. One RCT compared vertical banded gastroplasty with adjustable gastric banding and found greater weight loss at 12 months with gastroplasty. However, at 5 years, patients who had undergone gastroplasty had regained some weight, while those who had undergone gastric banding continued to lose weight. At 5 years, 56% of gastroplasty patients were "satisfied" with the results of their operation, compared with 81% of gastric banding patients.

Two RCTs compared open with laparoscopic gastric bypass and neither found any significant differences in weight loss or complications between groups. While operative time was longer for laparoscopy, there were significant reductions in the proportion of patients requiring an intensive care unit stay, length of hospital stay, and time to return to activities of daily living and work. One RCT compared open with laparoscopic adjustable silicone gastric banding, and again found no significant differences in weight loss between groups. Open gastric banding was associated with higher rates of complications, more re-admissions and a longer length of hospital stay.

In summary, jejunoileal bypass is effective at achieving weight loss but is associated with serious complications, including liver disease. Gastric bypass appears to be more effective than gastroplasty in terms of achieving weight loss. However, it is a technically demanding operation, with potentially serious metabolic complications. Vertical banded gastroplasty appears to be more effective than horizontal gastroplasty. Gastric banding is associated with more long-term weight loss, fewer re-operations and greater patient satisfaction than vertical banded gastroplasty, although the difference in the results is not statistically significant. It is the least invasive of the procedures, involving no permanent alteration of the anatomy. Laparoscopic surgery is as effective as open surgery, but results in fewer complications and a reduced length of hospital stay.

Four published economic evaluations were found; two were from the USA, one was from the Netherlands and one was from Sweden. All evaluations were of poor quality. The Assessment Group developed its own model of cost-effectiveness, and one of the gastric band manufacturers included a cost-utility analysis.

Of the published evaluations, one study found that gastric bypass was more cost effective at producing and maintaining weight loss for follow-up of 2 to 6 years than a very-low-calorie diet ($250-$750 vs $100-$1600 per pound lost). A cost-utility analysis found that vertical banded gastroplasty dominated no treatment. One study found gastric bypass to be more costly than laparoscopic vertical banded gastroplasty despite shorter operating times ($16,700 vs $12,800). The Swedish study comparing gastric bypass, vertical banded gastroplasty and gastric banding with conventional treatment calculated the direct cost of surgical treatment at SKr165,000 (£11,000) per patient over 10 years. Health-related quality of life improved in the surgical intervention group but not in the control group.

The Assessment Group constructed an economic model comparing gastric bypass, vertical banded gastroplasty and adjustable gastric banding with non-surgical management. Utility gains from surgery were based on utility values categorised by BMI. Costs were based on laparoscopic procedures with conversion to open surgery based on results observed in trials. To be conservative, it was assumed that surgery does not affect life expectancy, that patients revert to their baseline BMI after 5 years and that diabetes was the only co-morbidity for which there is a cost-offset. The time horizon for the model was 20 years following surgery.

Based on these conservative assumptions and compared with non-surgical management, all three surgical interventions appeared to be cost-effective, with incremental cost-effectiveness ratios (ICERs) of £6289 (gastric bypass), £8527 (gastric banding) and £10,237 (gastroplasty) per quality-adjusted life year (QALY). The results were robust to a number of sensitivity analyses.

Given the paucity of and uncertainty in the data available, it is problematic to compare different surgical procedures. There are very small differences between surgical procedures in the model in terms of estimated total costs (just over £1000) and estimated QALYs gained (less than 0.2 per person), and consequently small changes to the estimates of costs and benefits can significantly change and even reverse the relative cost-effectiveness of the surgical procedures. It is therefore not possible on the basis of the evidence currently available to choose between types of surgery on the grounds of cost-effectiveness.

The submission from one of the gastric band manufacturers included a cost-utility analysis comparing gastric banding with treatment with the drug orlistat. The analysis showed that over a 10-year period, the costs of banding were very similar to those of orlistat, but that banding offered significantly greater benefits in terms of weight loss. The manufacturer argued that as it had a lower cost per QALY than orlistat, banding should be the treatment of choice.

In summary, surgery appears cost-effective compared with conventional treatment. Laparoscopic surgery is likely to be more cost-effective than open surgery due to a reduced length of hospital stay. However, it is not possible to choose between surgical interventions on the grounds of cost-effectiveness.

The Committee reviewed the evidence available on the clinical and cost-effectiveness of surgery for people with morbid obesity, and considered evidence from user representatives and clinical experts on the nature of the condition and the benefits of surgery. They took into account that people who would be eligible for surgery under the criteria in Section 1 will have tried a number of conservative approaches without success. They also considered the substantial costs to the NHS of treating the co-morbidities associated with obesity.

In developing its recommendations, the Committee considered the important issue of patient choice as well as the clinical evidence available. The Committee believed that the choice of surgical intervention should be based on an informed discussion with the clinician responsible for the patient's treatment, and should take into account the potential benefits and risks of surgery, the facilities and equipment available, and the experience of the surgeon who would perform the operation. The Committee also discussed the need for multidisciplinary support and the appropriate infrastructure to be in place before surgery is undertaken. Long-term follow-up was considered essential not only to monitor the patient's progress over time, but also to generate further data on the clinical and cost effectiveness of the surgery.

The Committee accepted that given the conservative assumptions in the Assessment Group's economic model and the relative robustness of the estimates of cost-effectiveness to changes in key parameters, the ICER for surgery compared with non-surgical management is almost certainly less than £10,000 per QALY. However, the Committee was not able to distinguish between different surgical interventions on the grounds of cost-effectiveness.

The Committee discussed the criteria, which form Section 1.3 of the guidance, that should be applied to people with morbid obesity before undergoing surgery. The criteria have been derived by the Committee from either the evidence base or discussions with consultees to the appraisal process. The criteria relate to the advisability of surgery in general for this group of people and the consideration that surgery should be regarded as an option for patients only when other intensive treatments have been unsuccessful in maintaining weight loss.

The use of surgery as a last-line therapy

There is currently little evidence on the long-term consequences and impact on quality of life of surgery for people with morbid obesity in whom all other conservative measures have failed to maintain weight loss. There are also few economic evaluations comparing different surgical interventions. High priority should therefore be given to collecting long-term data (as consecutive case-series) on the costs and outcomes associated with surgery for people with morbid obesity, as well as for those with BMI greater than 35 kg/m2 and co-morbidities.

With few exceptions, studies of surgery for obesity have been uncontrolled and short-term with no comparators. Carefully planned and executed randomised controlled trials comparing obesity therapies with the best alternative modern treatments and with currently standard non-surgical treatment are required and should be started soon. The studies should be prospective with comprehensive medical and cost-utility analyses.

Randomised controlled trials are required to compare different surgical techniques with regard to effectiveness, costs and quality of life of patients in both the short and long term.

The trials referred to in 5.2.1 and 5.2.2 should be powered to compare the clinical effectiveness and cost-effectiveness of surgical and non-surgical treatments, and of different surgical treatments, for people with and without co-morbidities, at different levels of BMI. Preliminary results should aim to inform the review of this guidance in 2005, but at least a further 5 years' follow-up would be necessary to generate the relevant health-economic data.

Resource impact for the NHS

The estimated cost to the NHS of implementing this guidance depends on a number of factors, all of which are subject to considerable uncertainty.

• The number of people in England and Wales who have a BMI of 40 kg/m2 or more is probably currently about 600,000. The number having a BMI between 35 and 40 kg/m2 and a serious co-morbidity is not known but it is assumed for the purposes of this Section also to be 600,000, giving an estimated total target group (people with morbid obesity) of 1.2 million.
• The rate of growth of this group is not known, but it is assumed to be 5% per year.
• The number of deaths in the next year among the 1.2 million in the existing target group is assumed to be 40,000.
• The proportion of the target group who might ever take up obesity surgery is not known, but it is assumed to be either 2% or 4%. (It may be higher, but this has not been modelled here.)
• The cost of each operation, over and above the cost of medical care, is assumed to be on average £5500.
• The rate at which the number of operations can be increased over time is not known. It has been assumed that, as only 200 operations are performed each year at present, capacity could rise by 300 operations to 500 in the first year, and by 500 additional operations in each year after that, up to a maximum of 4000 per year after 8 years. (This rate of increase is meant to be indicative rather than definitive.)
• Current surgical and non-surgical treatments are assumed to remain unchanged.

NHS hospitals that offer specialised services for people with morbid obesity, specialist commissioning groups and all clinicians who care for people with morbid obesity should review the current treatment options available for these people in line with the guidance set out in Section 1.

Specialists in hospitals that offer or want to offer surgery for people with morbid obesity should establish and maintain a database of these patients. Each patient record in the database should include evidence that the criteria for this type of surgery were met by the patient, the surgical procedure undertaken and the short-term and long-term outcomes, including complications, mortality, co-morbidity, long-term weight loss and impact on quality of life.

Local guidelines or care pathways on the care of people having surgery for morbid obesity should incorporate the guidance set out in Section 1.

Surgery to aid weight reduction is recommended as a treatment option for people with morbid obesity only when the following criteria are met.

The individual:

- has been receiving intensive management in a specialised hospital obesity clinic
- is aged 18 years or over
- has tried all appropriate and available non-surgical measures adequately but has not been able to maintain weight loss
- has no specific clinical or psychological contra-indications to this type of surgery
- is generally fit for anaesthesia and surgery
- understands the need for long-term follow-up.

Surgery to aid weight reduction for a person with morbid obesity is carried out only following comprehensive, multidisciplinary assessment, and only when arrangements are in place for appropriate healthcare professionals to provide pre-operative and postoperative counselling and support to the individual being considered for surgery.

An individual's informed consent for surgery reflects that the choice of surgical intervention was made jointly by the individual and the clinician responsible for treatment, following detailed discussion of the potential benefits and longer-term implications of surgery and the associated risks including postoperative mortality and complications. The discussion covers the best available evidence, the facilities and equipment available and the experience of the surgeon who would perform the operation.

The Institute has issued guidance on the use of orlistat and sibutramine for the treatment of obesity:

• National Institute for Clinical Excellence (2001) Guidance on the use of orlistat for the treatment of obesity in adults. NICE Technology Appraisal Guidance No. 22. London: National Institute for Clinical Excellence. Available from: www.nice.org.uk
• National Institute for Clinical Excellence (2001) Guidance on the use of sibutramine for the treatment of obesity in adults: NICE Technology Appraisal Guidance No. 31. London: National Institute for Clinical Excellence. Available from www.nice.org.uk

The review date for a technology appraisal refers to the month and year in which the Guidance Executive will consider any new evidence on the technology, in the form of an updated assessment report, and decide whether the technology should be referred to the Appraisal Committee for review.

It is proposed that the guidance on this technology is reviewed in May 2005.

Professor David Barnett

Chairman, Appraisal Committee

May 2002

The Appraisal Committee is a standing advisory committee of the Institute. Its members are appointed for a 3-year term. A list of the Committee members appears below. The Appraisal Committee meets twice a month other than in December, when there are no meetings. The Committee membership is split into two branches, with the Chairman, Vice-chairman and a number of other members attending meetings of both branches. Each branch considers its own list of technologies and topics are not moved between the branches.

Committee members are asked to declare any interests in the technology to be appraised. If there is a conflict of interest, the member is excluded from participating further in that appraisal.

The minutes of each Appraisal Committee meeting, which include the names of the members who attended and their declaration of interests, are posted on the NICE website.

Dr Jane Adam

Radiologist, St. George's Hospital, London

Professor R L Akehurst

Dean, School of Health Related Research, Sheffield University

Dr Sunil Angris

General Practitioner, Waterhouses Medical Practice

Professor David Barnett (Chairman)

Professor of Clinical Pharmacology, University of Leicester

Professor Sir Colin Berry

Professor of Morbid Anatomy, St Bartholomew's and Royal London School of Medicine

Dr Sheila Bird

MRC Biostatistics Unit, Cambridge

Professor Carol Black

Consultant Physician, Royal Free Hospital & UCL, London

Professor John Brazier

Health Economist, University of Sheffield

Professor Martin Buxton

Director of Health Economics Research Group, Brunel University

Professor Bruce Campbell

Consultant Surgeon, Royal Devon & Exeter Hospital

Professor Mike Campbell

Statistician, Institute of General Practice & Primary Care, Sheffield

Dr Karl Claxton

Health Economist, University of York

Professor Sarah Cowley

Professor of Community Practice Development, Kings College, London

Dr Nicky Cullum

Reader in Health Studies, University of York

Professor Jack Dowie

Health Economist, London School of Hygiene & Tropical Medicine, London

Mr Chris Evennett

Chief Executive, Mid-Hampshire Primary Care Group

Dr Paul Ewings

Statistician, Taunton & Somerset NHS Trust

Professor Terry Feest

Clinical Director and Consultant Nephrologist, Richard Bright Renal Unit, and Chairman of the UK Renal Registry

Ms Jean Gaffin

Formerly Executive Director, National Council for Hospice and Specialist Palliative Care Service

Mrs Sue Gallagher

Chief Executive, Merton, Sutton and Wandsworth Health Authority

Dr Trevor Gibbs

Head, Global Clinical Safety & Pharmacovigilance, GlaxoSmithKline

Sally Gooch

Director of Nursing, Mid-Essex Hospital Services Trust

Mr John Goulston

Director of Finance, The Royal Free Hampstead NHS Trust

Professor Trisha Greenhalgh

Professor of Primary Health Care, University College London

Miss Linda Hands

Chief Executive, Barnet & Chase Farm Hospitals NHS Trust

Professor Philip Home

Professor of Diabetes Medicine, University of Newcastle

Dr Terry John

General Practitioner, The Firs, London

Dr Diane Ketley

Research into Practice Programme Leader, NHS Modernisation Agency

Dr Mayur Lakhani

General Practitioner, Highgate Surgery, Leicester, and Lecturer, University of Leicester

Ruth Lesirge

Patient Representative; Director, Mental Health Foundation

Dr George Levvy

Patient Representative; Chief Executive, Motor Neurone Disease Association

Dr Gill Morgan

CEO, North & East Devon Health Authority

Professor Miranda Mugford

Health Economist, University of East Anglia

Mr M Mughal

Consultant Surgeon, Chorley and South Ribble NHS Trust

Mr James Partridge

Chief Executive, Changing Faces

Siān Richards

General Manager, Cardiff Local Health Group

Professor Philip Routledge