Eligibility for bariatric surgery among adults in England: analysis of a national cross-sectional survey

Abstract

Summary

Objectives:

This study aimed to determine the number eligible for bariatric surgery and their sociodemographic characteristics.

Design:

We used Health Survey for England 2006 data, representative of the non-institutionalized English population.

Setting:

The number of people eligible for bariatric surgery in England based on national guidance is unknown. The UK National Institute for Health and Clinical Excellence criteria for eligibility are those with body mass index (BMI) 35–40 kg/m² with at least one comorbidity potentially improved by losing weight or a BMI > 40 kg/m².

Participants:

Of 13,742 adult respondents (≥18 years), we excluded participants with invalid BMI (n = 2103), comorbidities (n = 2187) or sociodemographic variables (n = 27) data, for a final study sample of 9425 participants.

Main outcome measures:

The comorbidities examined were hypertension, type 2 diabetes, stroke, coronary heart disease and osteoarthritis. Sociodemographic variables assessed included age, sex, employment status, highest educational qualification, social class and smoking status.

Results:

5.4% (95% CI 5.0–5.9) of the non-institutionalized adult population in England could meet criteria for having bariatric surgery after accounting for survey weights. Those eligible were more likely than the general population to be women (60.1% vs. 39.9%, p<0.01), retired (22.4% vs. 12.8% p<0.01), and have no formal educational qualifications (35.7% vs. 21.3%, p<0.01).

Conclusions:

The number of adults potentially eligible for bariatric surgery in England (2,147,683 people based on these results and 2006 population estimates) far exceeds previous estimates of eligibility. In view of the sociodemographic characteristics of this group, careful resource allocation is required to ensure equitable access on the basis of need.

Introduction

Despite publication of the UK National Institute for Health and Clinical Excellence’s (NICE) guidelines for bariatric surgery in 2006, ¹ it is unclear how many people are potentially eligible for this procedure. NICE recommends bariatric surgery for patients with a body mass index (BMI) > 40 kg/m² or 35–40 kg/m² with a ‘significant disease that could be improved if they lost weight’. ¹ Once referred, patients enter a pathway requiring lifestyle change, specialist obesity management and follow-up. The guidelines are similar across Europe including from the Bariatric Scientific Collaborative Group ² and the Scottish Intercollegiate Guidelines Network, ³ as well as internationally from the National Institutes of Health (USA). ⁴

Bariatric surgery improves mortality and morbidity outcomes ^5,6 and can be cost-effective when appropriately targeted. ⁷ A Cochrane systematic review found bariatric surgery was more effective than conventional obesity treatment. ⁸ Sjöström et al. ⁵ found a 25 ± 11% weight loss at 10 years in those undergoing bariatric surgery with a 24% decrease in mortality compared with the control group. Bariatric surgery also improves diabetes-related mortality, cardiovascular disease and reduces cancer risk. ^9,10

Annual bariatric surgery activity has increased fivefold from 1996 to 2004. ^11,12 Yet despite evidence of cost-effectiveness, ⁷ bariatric surgery rates are estimated to be a third of the National Health Service (NHS) benchmark rate for bariatric services. ¹³ These estimates of the bariatric surgery needs of the population have relied on a number of assumptions relating to data, current practice and expert clinical opinion.

The purpose of this study was to determine how many people in England are eligible for bariatric surgery using criteria from national guidance and to examine the sociodemographic and comorbidity profile of the eligible population.

Methods

Sampling and data collection

We used nationally representative data from the Health Survey for England (HSE) 2006, ¹⁴ which focused on cardiovascular disease risk factors, to estimate the number eligible for bariatric surgery.

The HSE is a national annual household survey of Health and Lifestyles in England. The survey uses a multistage stratified probability design with the first stage based on postcode sectors. ¹⁵ Within each sector, a random sample of postal addresses was drawn.

In 2006, personal interviews were carried out to obtain personal, socioeconomic, household, health and service use data. Research nurses subsequently took measurements including height and weight at a follow-up visit.

For patients with a valid BMI, we examined these obesity-related comorbidities based on those in the NHS bariatric surgical service commissioning guide: ¹¹ hypertension, type 2 diabetes, stroke, coronary heart disease and osteoarthritis.

We also examined these sociodemographic variables: age, sex, employment status (employed, unemployed, retired and other), highest educational qualification; social class (managerial and professional, intermediate, routine and manual, and other), and smoking status.

The highest educational qualification was classified as up to O-level (school qualification for 14–16 year olds), up to A-level (college qualification for 17–18 year olds), and up to Degree level. Those who had not achieved any of the aforementioned qualifications were classified as having ‘no qualifications’.

Data analysis

We calculated the percentage of people with each obesity-related comorbidity in BMI groups <35 kg/m², 35–40 kg/m² and >40 kg/m² with 95% confidence intervals. In addition, we categorized the data into those with no comorbidities and those with at least one comorbidity in order to calculate those eligible for bariatric surgery. We also determined sociodemographic characteristics in the general adult population and in those eligible for bariatric surgery.

All data were weighted to the general population in England to overcome sampling errors. We used χ ² tests and calculated P values to compare frequencies. Statistical analyses were performed using Stata 11.1 (Stata Corporation, Texas, USA).

Results

Of 13,742 adult respondents in HSE 2006, we excluded participants with invalid data for BMI (N = 2103), comorbidities (N = 2187) or sociodemographic variables (N = 27), for a final study sample of 9425 participants.

Obesity-related comorbidities

Three hundred seventy-four (4.0%) of participants had a BMI 35–40 kg/m² with at least one comorbid condition and 179 (1.9%) of participants had a BMI > 40 kg/m² (Table 1). After taking into account survey weights, 5.4% (95% CI 5.0–5.9) of the non-institutionalized adult population in England fulfilled criteria for bariatric surgery.

Table 1.

Prevalence of obesity-related comorbidities in people aged greater than 18 years according to body mass index (BMI) in England*.

% BMI < 35 (CI)	% BMI 35–40 (CI)	% BMI > 40(CI)
Hypertension	19.3 (18.5–20.2)	42.1 (38.4–45.8)	51.6 (43.9–59.2)
Type 2 diabetes	3.2 (2.8–3.6)	8.8 (6.8–10.9)	16.7 (10.8–22.5)
Stroke	1.2 (1.0–1.5)	2.4 (1.3–3.4)	2.1 (0.0–4.2)
Coronary heart disease	3.2 (2.8–3.5)	5.9 (4.2–7.6)	8.4 (4.3–12.6)
Osteoarthritis	6.6 (6.1–7.2)	16.2 (13.6–18.8)	21.1 (15.2–26.9)
Comorbidities†
0	73.9 (73.0–74.8)	50.8 (47.0–54.6)	35.8 (28.3–43.3)
≥1	26.1 (25.2–27.1)	49.2 (45.4–53.0)	64.2 (56.7–71.7)
Weighted base (N)	8712	705	172
Unweighted base (N)	8525	721	179

^*Data from HSE 2006.

^†Matched with comorbidities stated in NICE commissioning guide.

The prevalence of obesity-related comorbidities was generally greater in those with a BMI ≥ 35 kg/m² than with a BMI < 35 kg/m². Similarly, the prevalence of comorbidity was higher with a BMI > 40 kg/m² compared to a BMI 35–40 kg/m² (64.2% vs. 49.2%), particularly for those with type 2 diabetes (16.7% vs. 8.8%).

The overall prevalence of comorbidities increased with BMI, almost doubling in the BMI 35–40 kg/m² group compared with the BMI < 35 kg/m² group (49.2% vs. 26.1%). At a BMI >40 kg/m², the proportion of people with ≥1 comorbidity increased further to 64.2%.

Sociodemographic factors

There were more women in the eligible group compared with the general population (60.1% vs. 50.0%, P<0.01) (Table 2). Those eligible had a significantly higher proportion with no qualification compared with the general population (35.7% vs. 21.3%). In addition, they were less likely to have studied up to degree level (22.0%) or A-level (7.9%) compared with the general population (34.6% and 16.4%, respectively). There were significantly less eligible patients in managerial/professional social class groups (26.7% vs. 34.6%) and significantly more working in routine/manual work (49.5% vs. 38.9%). Those in the eligible group were less likely to be current smokers compared with the general population (19.1% vs. 24.2%) but more likely to be ex-regular smokers (33.7% vs. 22.7%).

Table 2.

Sociodemographic characteristics of those eligible for bariatric surgery* compared with the general population.

General adult population (%)	Eligible population† (%)	P ‡
Age (years)
18–44	57.4	35.1	<0.01
45–64	32.8	48.2	<0.01
>65	9.8	16.8	<0.01
Sex
Women	50.0	60.1	<0.01
Men	50.0	39.9	<0.01
Employment status
Employed	66.4	49.1	<0.01
Unemployed	4.8	3.1	0.11
Retired	12.8	22.4	<0.01
Other	16.0	25.4	<0.01
Highest educational qualification
Degree	34.6	22.0	<0.01
A level	16.4	7.9	<0.01
O level	27.8	34.3	<0.01
No qualification	21.3	35.7	<0.01
Social class
Managerial and   professional	34.6	26.7	<0.01
Intermediate	21.9	20.4	0.40
Routine and   manual	38.8	49.5	<0.01
Other	4.7	3.5	0.24
Smoking status
Never	47.7	42.6	0.02
Ex-occasional	5.4	4.7	0.48
Ex-regular	22.7	33.7	<0.01
Current	24.2	19.1	<0.01

^*As per NICE guidelines.

^†Age > 18 years, and either a BMI 35–40 kg/m² with comorbidity or a BMI > 40 kg/m² regardless of comorbidity status.

‡P values refer to χ ² significance tests for a difference between the groups.

Discussion

Main finding of this study

Using representative data of the non-institutionalized population in England, we estimated that 5.4% of adults are potentially eligible for bariatric surgery. Based on population estimates for 2006, ¹⁶ this equates to 2,147,683 people. Those fulfilling the criteria for bariatric surgery were more likely to be women, retired, have lower educational qualifications and have lower socioeconomic status.

What this study adds

This is the first study to quantify the number of people eligible for bariatric surgery using data from a nationally representative survey in England.

Limitations of this study

There are some important limitations. First, of the 13,742 adult patients in the dataset, 2103 (15%) had missing BMI data although it is difficult to say in which direction this would affect the estimate. Second, many of the self-reported comorbid conditions rely on patient recall so some survey respondents with a BMI 35–40 kg/m² may have undiagnosed conditions that could qualify them for bariatric surgery. This could lead to an underestimation of those eligible for bariatric surgery. Third, extrapolating this nation-level data from a small sample size is likely to introduce a margin of error in these estimates. However, the data were weighted for non-response and was specifically designed to be representative.

The most common obesity-related comorbidities were assessed based on those in the NICE commissioning guide. ¹³ However, not all of these comorbidities could be included as HSE 2006 did not provide data about some comorbidities such as obstructive sleep apnoea. This could potentially result in an underestimation of the number eligible for surgery.

Due to a lack of available data, however, we were unable to exclude those who were unfit for surgery or anaesthesia, as well as those who were not committed to long-term follow-up. Failure to exclude these people, which is a part of the guidance from countries including the UK, means that the number of people eligible for bariatric surgery may be overestimated. Finally, we could not estimate how many of the individuals who are potentially eligible for bariatric surgery would actually wish to undergo such treatment.

What is already known on this topic

Few studies in the literature have looked specifically at the eligibility of patients for bariatric surgery according to guidelines. Livingston and Ko ¹⁷ examined the socioeconomic characteristics of those eligible for bariatric surgery compared to those actually having it.

The guidelines for bariatric surgery in the USA from the National Institutes of Health ⁴ are similar to those in Europe. ^1,2 The 2000 National Health Interview Survey database found that 2.8% of the American population (5,324,123 people) was eligible for surgery. Of those eligible, a higher proportion were ‘black, poorly educated, or impoverished’. Among those eligible, 38% relied on Medicare or Medicaid to pay for the operation. In contrast, only 13% of those who actually had bariatric surgery in 2000 had the treatment funded by Medicare or Medicaid. Furthermore, fewer operations were performed on black people than expected.

A later national study based on data from America between 2005 and 2006 identified 22,151,116 people eligible for bariatric surgery using the National Institutes of Health criteria. ¹⁸ Those eligible were found to have lower incomes, education levels and less healthcare access as well as a greater proportion being from ethnic minorities.

Both of these studies support the finding of the current study that a greater proportion of those eligible for bariatric surgery are from groups with lower socioeconomic status.

In the UK, estimates for NHS needs calculated for the NICE bariatric surgery commissioning guide used IMS Disease Analyser, which extracts data from general practices, to provide a representative population sample. ¹³

For people with a BMI 35–39.9 kg/m² with comorbidities, they found 0.8% of the population eligible. However, in our study, using HSE data, the equivalent figure was five times greater at 4.0%. Our analysis was based on a systematically sampled community based survey of households whereas General Practitioner (GP) data rely on health-seeking behaviour of registered patients. Although in recent years GPs have improved their disease registers, the incomplete recording of conditions and measures such as BMI is likely to underestimate the prevalence of obesity.

The current NHS benchmark for a bariatric surgical service at 5 years is 0.01% of the population per year. ¹³ In 2007, this was more than threefold the estimated rate of NHS commissioned bariatric surgery. ¹³ The NICE topic-specific advisory group admit this is not the optimum rate of procedures required and that rates ‘may need to increase beyond this’.

Recent Hospital Episode Statistic data show surgery rates from 2003/4 to 2009/10 have risen year on year in England; with rates highest in women and those aged 40–54 years. ¹⁰ However, service delivery rates still fall significantly below the level needed to support all those who could potentially benefit.

With national guidance stating explicit criteria for assessing eligibility for bariatric surgery, and both this and other studies finding eligibility to be well above provision rates, we need to ask why this is not being implemented. Several factors may contribute. First, at the patient level, this study could be consistent with an inverse care law whereby those most in need of bariatric surgery are in socioeconomic groups who tend to make less use of healthcare services. ¹⁹ Second, patient awareness of bariatric surgery and patient commitment to complete lifestyle intervention programmes prior to surgery, may also represent barriers to uptake. ²⁰ Third, GPs may not adequately identify and refer patients eligible for treatment perhaps due to a lack of awareness of the potential benefits or local variation in guidance limiting access. Fourth, national bariatric service provision may be insufficient.

Implications for practice and further research

The current eligibility rate for bariatric surgery far exceeds the estimated service delivery which itself falls short of NHS benchmarks for bariatric services.

This has resource implications for the NHS as the cost of implementing NICE guidance, at least in the short term, would be huge. Further work is needed to understand the long-term cost-benefit of bariatric surgery. This highlights the important distinction between having clinical guidelines on paper, as opposed to the actual implementation of these guidelines in practice. Additionally, examining the future need for bariatric surgery, accounting for changing sociodemographic factors, will improve accuracy of service planning.

In England, with the emergence of GP-led commissioning consortia and new local guidance at this level, there is the risk that bariatric surgery access may become less uniform. Coupled with the sociodemographic characteristics of those meeting eligibility criteria, any service change needs to ensure equitable access to bariatric surgery on the basis of need. These issues will become increasingly important to address, across Europe and worldwide, as obesity rates continue to rise.

Conclusions

5.4% of the general adult population is eligible for bariatric surgery in England, far exceeding the current bariatric surgery uptake. Due to the limited capacity of health services to meet demand under existing criteria, greater investment into service provision may be required to meet a growing need. This would have significant resource implications. Since those eligible are more likely to be of a lower social class and have lower qualifications, such resources would need careful allocation to ensure equitable access on the basis of need.