Most recent update: Friday, April 3, 2020 - 08:40

Bariatric News - Cookies & privacy policy

You are here

National Bariatric Surgery Registry

Outcomes from the first bariatric surgery report

The First National Bariatric Surgery Report is the first comprehensive, prospective, nationwide analysis of outcomes from bariatric surgery in the United Kingdom and Ireland. The report is the result of the data collected by the National Bariatric Surgery Registry (NBSR). The outcomes from the report were presented for the first time at the Royal College of Surgeons’ ‘Improving performance through outcomes’ seminar on 13th April 2011. Mr Richard Welbourn (Taunton, UK) who authored the dataset and is the lead author of the report began by outlining the stark reality of obesity epidemic in the UK.

“Currently one in four adults are obese and many more are overweight,” said Welbourn. “The problem with obesity is that it has the potential to lead to obesity associated diseases. This is threatening to overwhelm the National Health Service (NHS) and as a society we have completely failed to get to grips with the issues. Today, we are going to present the first outcomes data from the NBSR, which shows bariatric surgery resoles a patient’s obesity and their obesity associated diseases.”

A total of 84 surgeons from 86 hospitals entered 8,710 operations in the NBSR; 7,045 in the financial years 2009 and 2010. The data presented in the report include a breakdown of a patient’s BMI and co-morbidities before and after surgery, as well as excess weight loss. In addition, the report records the number and type of procedures (as well as the procedure funding source, NHS and/or private insurers), the in-hospital mortality rate after primary surgery and the surgical complication rate.

Richard Welbourn

“It was decided not to collect quality of life data as this is a very involved process, time consuming and can be a very subjective, and therefore, inaccurate assessment of their quality of life,” said Welbourn. He revealed that the NBSR collects data on 12 comorbidities, as opposed to some 33 comorbidities collected by the Bariatric Obesity Longitudinal Database (BOLD). “This is because the Database Committee wanted to strike a balance between collecting too much information and risk disenfranchising data collectors, and collecting too little information resulting in meaningless data,” Welbourn explained.

Procedure type

According to the report, more than 85% of the operations recorded were either Roux-en-Y gastric bypass (RYGB) or gastric banding. In total, there were 3,817 gastric bypass procedures, 2,132 gastric band operations and 588 sleeve gastrectomies. All but one primary gastric band operation, and over 90% of RYGB operations, were performed laparoscopically. Almost all primary sleeve gastrectomies were laparoscopic, as were more >90% of revisions.


The NBSR data reveal that the vast majority of patients are female and this is typical of many other series reported across the world. The report also reveals that there is a significant decrease in the proportion of female patients with increasing age, excluding patients in the 25 and under age bracket.

For example, the rate falls from 88.1% female patients in the 25-29 year olds to 62.5% in the >64 year-old group (there is also a significant difference in gender when comparing patients ≤44 years of age versus >44 years of age, p<0.001).

According to the authors, this suggests that women are more likely to come for surgery at an earlier age, even though studies in other surgical specialties have shown that women are sometimes at greater risk of adverse events following surgery. It is also interesting to ask that if obesity rates are the same for both sexes, why are women more willing to come for bariatric surgery more frequently than men?


The issue of funding is one of the most controversial issues surrounding bariatric surgery with many patients' elegibility for surgery in the UK being dependent on their geographic location, leading to a "postcode lottery". On average 68.8% of procedures were publicly funded; for the three most common procedures, this ranged from 43.1% for gastric banding procedures to 81.4% for RYGB and 84.7% for sleeve gastrectomy.

The data also shows a significant difference in the ratio of publicly funded to privately funded surgery is almost 4:1 for male patients as opposed to only 2:1 for female patients; (p<0.001). For the female population, there was a slight increase in the proportion of operations that are publicly funded with ncreasing age, whereas amongst the male patients there is a distinct difference in the rate of publicly funded surgery for those under the age of 35 and those 35 years of age and older.

Body mass index, operation and gender

Interestingly, the report also suggests that women come to surgery earlier in the disease process than men, irrespective of the type of surgery they eventually elect to undergo, as there are considerably more female patients in the smaller BMI groups than men for each procedure type. Moreover, it is apparent that irrespective of gender, patients having a RYGB procedure have, on average, a greater BMI than patients having a gastric band, and that sleeve gastrectomy patients, on average, have an even higher BMI (Figure 1).

Figure 1: Primary operations - operation, gender and body mass index (n=6,032)

“This implies that gastric banding is a procedure that is deemed suitable for the patients in the early stages of their disease, whereas for patients with a greater body mass index, with more extensive excess-weight and, concomitantly, with more obesity-related comorbid conditions, a more definitive and long-term solution is indicated,” the report states.

Baseline disease status

The report also includes an assessment of pre-operative comorbidities: type 2 diabetes, hypertension, dyslipidaemia, cardiovascular atherosclerosis, sleep apnoea, asthma, functional status, back or leg pain from arthritis, GORD, liver disease, depression and polycystic ovarian syndrome. 

The report notes that more than 68% of all records have no missing comorbidity data at all, and just under 17% have only one field missing. “This is an astonishingly impressive achievement for a newly-developed registry, considering the number of operations for which data have been entered, and speaks of the commitment of the contributors to providing high-quality data, and to the acceptance of the NBSR as a valid and valuable dataset,” he said.

He believes that this is the first data from a national registry to show primary comorbidity data grouped according to BMI, and more than half the patients with a BMI of >60 have more than four comorbidities (Figure 2), representing a massive burden of disease for these patients.

Figure 2: Primary operations with complete morbidity data - numbr of comorbidities and body mass index (n=4,355)

Perhaps the most startling of the comorbidity data presented was that 70% of patients in the registry presented with poor functional status, defined as unable to walk three flights of stairs.

“To say to these patients that you need to lose weight by taking exercise is just not feasible – they cannot even climb a flight of stairs,” said Welbourn. He also revealed that male patients have more comorbidities than women, with 43% of males presenting with type 2 diabetes.

The NBSR shows that both men and women have high rates of poor functional status (75.0% and 67.6%) and arthritis (54.0% and 53.9%), and that the rates of back or leg pain due to arthritis and GORD occur at similar rates in both male and female patient populations. The rates of comorbidities significantly higher amongst female patients are depression and asthma, although the reasons for these apparently higher rates in women are not fully understood (Figure 3).

Figure 3: Primary operations - gender and rates of various comorbid conditions recorded in the database

One of the comorbidity questions in the database is only collected for female patients (polycystic ovarian syndrome). Therefore, the maximum number of missing comorbidity data-items for male patients is 11, whereas the maximum number of missing data-items for female patients is 12.

Post-operative outcomes

The report states that bariatric surgery is safe as the observed in-hospital mortality rate after primary surgery was 0.1% overall, with a surgical complication rate of 2.6%. Most patients (80%) were discharged by the third post-operative day.

The authors say that their results provide evidence that bariatric surgery is one of the most clinically effective, safe and cost-effective treatments available to the NHS: “There were only seven deaths recorded in the entire registry… [This is] a remarkable achievement, and reflects the safety of bariatric surgery.”

The zero mortality recorded for gastric banding and sleeve gastrectomy, coupled with the 0.22% mortality for gastric bypass (Table 1) compare favourably with the best published international data. 

Table 1: Post-operative mortality

Post-operative stay

As expected, nearly all gastric band patients stay one day or less in hospital. Despite being a major procedure for patients with severe obesity-related disease, 80% or more of bypass patients are discharged by day three. The authors claim that the short in-hospital stay is almost certainly due to the surgery being performed via a laparoscopic approach and surgeons being sufficiently confident in their operative technique.


The NBSR recorded patients' weight at first consultation. Surgeons then have the option of also entering the patient’s weight at the point immediately before surgery, which allows an estimate of weight loss prior to surgery. The authors believe that the report includes the first international data to show that Bariatric Care Teams have successfully used strategies to help a substantial number of patients to lose weight, in some cases a significant amount of weight, prior to surgery.

Overall, 50.5% of gastric banding patients and 55.3% of Roux-en-Y gastric bypass patients reported some weight loss before surgery. In contrast, 49.5% and 44.7% of patients respectively fail to lose any weight before surgery, and 9.0% and 11.5% respectively actually gain weight.

The average excess weight loss achieved before surgery in those who lost weight was 10.5%. Overall the increase in excess weight amongst those who gained weight was 7.2 %. It is hoped that future reports will be able to determine whether or not weight loss immediately before surgery has an impact upon surgical complication rates and/or long-term weight loss.

In the first two years following surgery. Excess weight loss was recorded as being greater following gastric bypass than after gastric banding procedures (Figure 4). The data also show that women lose more weight than men on average, and women continue to lose weight in the period one to two years after gastric bypass, whereas men’s excess weight loss stabilised at around 40-45% in the same period.

Figure 4: Selected primary operations - post operative excess weight loss

“Does this mean that bypass is better than banding?” asked Welbourn. “No, firstly because the patients are not randomised and secondly because we know from international data that weight loss on bands is slower.”

Improvement in diabetes

The report includes the first published registry data from the United Kingdom and Ireland on changes in the rates of the clinical indication of diabetes post-bariatric surgery.

The data show that there is a progressive and substantial increase over time in the number of patients reported as having no indication of type 2 diabetes (impaired glycaemia / impaired glucose tolerance at presentation; oral hypoglycaemics at presentation.; and insulin treatment at presentation) in follow-up (Figure 5).

Figure 5: Patients recorded as having an indication of diabetes prior to surgery and changes in rates of recorded diabetes indications and type of diabetes

From the procedures reporting follow up data, 379 patients had type 2 diabetes before surgery, while one year later that figure had fallen to 188, and after two years 86% of those with diabetes prior to surgery had no indication of the disease (i.e. were able to stop their medications). 

The longer a patient has diabetes the more the pancreas is damaged and the more damaged the pancreas the less likelihood a patient will return to a of ‘no indication’ status.

“If anyone suggests that the patient should diet, although this may improve glycaemic control unfortunately this does not resolve the diabetes but bariatric surgery can resolve diabetes,” Welbourn stated.

Outcomes by procedure – gastric banding

According to the data, patients undergoing a gastric banding operation have significantly fewer comorbid conditions than the rest of the patients in the registry. This is consistent with female patients choosing to have surgery at an earlier stage in their disease process, before they have developed more serious and extensive comorbidities.

Patients having a gastric banding procedure had fewer of each of the recorded obesity-related conditions, except for reflux disease and polycystic ovarian syndrome, compared to patients undergoing other procedures recorded in the NBSR. Despite this, patients having a gastric banding procedure still have a considerable burden of disease.

For instance, 60% were unable to climb three flights of stairs without resting and half had some form of limiting arthritis (Figure 6). There is a reduction in the rate of each comorbidity 12 months after gastric banding and two comorbid conditions show a statistically significant improvement in the same period: dyslipidaemia (p=0.037) and functional status (p<0.001).

Figure 6: Primary gastric banding procedures - comorbid conditions before and after surgery

For the patients with follow-up data dated at 12 months after their procedure, 15.9% have dyslipidaemia initially, reducing to a rate of 11.2% one year later. As far as functional status is concerned, the reported rate falls from 64.4% of patients unable to climb three flights of stairs to only 46.7% at 12 months after surgery.

The report states that this is an important measure as many of these patients who attain increased mobility will be able to have to fuller lives, including returning to work, exercise, reducing the fiscal burden on the healthcare and welfare sectors, and contributing to the public purse by becoming tax-payers.

The report does note some caution however, noting that “although the data show improvements for all the other named conditions as well, these results should be interpreted with some caution as this sub-population was relatively fit prior to surgery and there is some uncertainty around the reported rates (note that the 95% confidence intervals substantially overlap).”

The data also show the proportion of patients returning to a state of no indication of diabetes with gastric banding compared to all other operations (segmented according to BMI), and seem to show that the rate of change after gastric banding is lower than that for the other operations (eg. gastric bypass and sleeve gastrectomy).

It is clear that the duration of diabetes has an effect on the reported rate of clinical indications of diabetes, reinforcing the argument for early intervention for obese patients with or at risk of diabetes.

In the report's analysis there are relatively few patients, but, as might be expected, those with a shorter duration of diabetes were more likely to revert to a state of no indication of diabetes. Further analysis shows that the curve for patients who have had diabetes for <4years is significantly different to that for patients whose diabetes spanned >7 years (p=0.004); comparing the curves for <4 years and 4-7 years duration also reveals a significant difference (p=0.043); whereas, the differences between the curves for diabetes of 4-7 and >7 years duration did not attain statistical significance (p=0.443).

Outcomes by procedure – gastric bypass

The gastric bypass procedure constituted 54.7% of all operations in the NBSR, with over 95% performed as a primary procedure. For less than 1% of patients the operation was a revision. The report also stresses that 3.6% of recorded gastric bypass procedures were carried out as a revision procedure where the primary surgery had been performed in another unit by another surgeon.

According to the report, patients undergoing gastric bypass had more obesity-related comorbid disease than for other procedures (male gastric bypass patients versus men undergoing other bariatric procedures: p=0.001; female gastric bypass patients versus women undergoing other bariatric procedures p<0.001). 

Interestingly, gastric bypass patients with a BMI of less than 55.0 have substantially more comorbidities than patients having other kinds of bariatric procedures. For patients witha BMI ≥55.0 the numbers of comorbid conditions are comparable across the different kinds of primary bariatric surgery.

The reports suggests that patients with significant disease are more likely to have gastric bypass operations, possibly in the belief that this procedure will more quickly produce improvements in their condition.

Although only 8.3% of gastric bypass operations were performed by open surgery (301 open versus 3,319 laparoscopic; 6 unspecified), analysis shows that this group of patients had significantly more comorbid conditions than the corresponding group of patients treated laparoscopically (p<0.001).

However, the NBSR does not record the reason that the patient and surgeon decided to opt for a particular operative approach, although one possible explanation for adopting the open approach is that surgeons find it preferable for high-risk patients.

The prevalence of each obesity-related comorbid disease is significantly greater in the patients undergoing gastric bypass surgery when compared to all the other patients in the registry, with the exception of gastro-oesophageal reflux disease for which the rates are almost identical in the two groups.

Therefore, this group of patients are clearly more systemically unwell and theoretically at higher risk of adverse outcomes following surgery, but they also have the more to gain in terms of improved wellbeing associated with successful weight loss after surgery.

The overall operative complication rate for gastric bypass was 3.4%, with the most-feared complication, leakage of one of the small bowel joins, occurring in only 0.6% of cases. The 30-day re-operation rate recorded for primary operations was 1.9%. The overall mortality rate for gastric bypass surgery was 0.22%, which the report states compares favourably to best international data.

Due to the small number of deaths comparisons of mortality rates across Obesity Surgery Mortality Risk Score (OSMRS) groups did not reach statistical significance.

For the patients with follow-up data dated 12 months after their procedure (Figure 7), there are some significant changes in morbidity rates including:

  • sleep apnoea rates decline by over 63%.

  • dyslipidaemia rates fall by over 61%.

  • the proportion of patients able to climb three flights of stairs improves dramatically, from 26.9% to 70.4%.

  • type 2 diabetes and GORD fall by over 56%.

  • for the remaining conditions analysed, the fall in the prevalence is in the range 30-42%.

Figure 7: Primary Roux-en-Y gastric bypass procedures - comorbid conditions before and after surgery

The gastric bypass operation has a dramatic effect on the reported rates of diabetes, with the majority of patients returning to a state where they exhibit no indication of diabetes. The level of improvement far exceeds that seen after other types of bariatric surgery.

Identifying the reasons for the rate and extent of this improvement after RYGB surgery will be an important area for further investigation that could have a profound implications for the treatment of type 2 diabetes within the NHS.

Outcomes by procedure – sleeve gastrectomy 

Sleeve gastrectomy comprised 8.4% of the operations recorded in the NBSR and the 588 patients recorded in the NBSR constitutes one of the largest patient series to date. The average age for a female patient undergoing a sleeve gastrectomy procedure was 45.0 years (n=376; SE=0.56 years), and for a male patient 45.3 years (n=167; SE=0.77 years).

Sleeve gastrectomy was performed more frequently for male publicly funded patients, probably because of the perceived high risk in patients with much comorbidity. A greater proportion of the sleeve gastrectomy patients were aged 50-64 than patients undergoing other bariatric procedures.

These differences between the age profiles for sleeve gastrectomy patients and other patients are not statistically significant (p=0.097), although the differences may become significant as the registry matures. 

The increased age of female patients undergoing a sleeve gastrectomy procedure means that there is potentially added risk of adverse events after the operation, simply because of the patients’ relatively greater age, but the data recorded in the NBSR show that they are at no greater risk of having an operative complication (p=0.851), nor of having a 30-day re-operation (p=0.461) than the patients undergoing gastric bypass surgery, the observed post-operative complication rates are also comparable (cardiovascular complication are almost identical at around 1%, p=0.937; other complications are 2.8% for both procedures, p=0.916), suggesting that this is a safe bariatric procedure in the hands of surgeons across the UK and Ireland.

Patients receiving a sleeve gastrectomy also had a higher frequency of obesity-related disease, suggesting that the patients undergoing this kind of bariatric procedure are a generally sicker population. These patients have relatively more coexisting disease at higher (>44.9) BMIs, suggesting something more complex increasing the comorbidity than simply increased BMI.

This observation holds even when comparing those patients having a primary sleeve gastrectomy procedure with the patients undergoing primary gastric bypass surgery (excepting the >59.9 group, where there is convergence across all procedure types, irrespective of gender), and the same general pattern is there in both the male and female patient groups.

The reported prevalence of three conditions is lower amongst patients undergoing a the sleeve gastrectomy procedure, although none of these differences are statistically significant:

  • arthritis.

  • GORD.

  • polycystic ovarian syndrome.

  • The rates of the remaining comorbid conditions are higher for sleeve gastrectomy patients (Figure 8). For six of the named conditions, the observed prevalence for the patients undergoing a sleeve gastrectomy is significantly higher:

  • type 2 diabetes.

  • hypertension.

  • dyslipidaemia.

  • sleep apnoea.

  • asthma.

  • poor functional status.

Figure 8: Primary sleeve gastrectomy procedures - comordib conditions before and after surgery

There are no significant differences in the rates of comorbid conditions between bypass and sleeve gastrectomy patients, although the latter have much higher rates of hypertension (p=0.001), sleep apnoea (p=0.008) and poor functional status (p<0.001). There is therefore a very high degree of comorbidity in these patients. 

The 30-day re-operation rate was 2.9%, which the authors claim is low for an operation performed predominantly for high-risk patients with substantial comorbidity. The risk of a re-operation within 30 days is significantly higher after revisional surgery (3.1% vs. 11.4%; p=0.019), which is consistent with operating on patients who have had prior surgery in the same area.

There are reductions in all rates of comorbidity for this group of patients, and for five of the conditions the fall is already statistically significant 12 months after surgery:

  • type 2 diabetes.

  • hypertension.

  • dyslipidaemia.

  • sleep apnoea.

  • functional status.

The report notes a vast improvement in functional status. Before surgery only 15.1% could climb three flights of stairs without resting; one year later 68.9% could do this:. Only bariatric surgery can produce such an improvement in functional status in these patients, the authors claim.

The 30-day re-operation rate was 2.9%, which is low for an operation performed predominantly for high-risk patients with substantial comorbidity. The risk of a re-operation within 30 days is significantly higher after revisional surgery (3.1% versus 11.4%; p=0.019), which is consistent with operating on patients who have had prior surgery in them same area.

Although as yet the number of sleeve gastrectomies performed as a revisional bariatric procedure are relatively low, there is a statistically significant increased risk of any complication developing and also for re-operations within 30 days.

“In conclusion, we are treating one third of one per cent of patient who are eligible for bariatric surgery. We know the from published reports that the cost effectiveness for bariatric surgery is between £2,000-£4,000 per quality of life year. This data shows that bariatric surgery is safe, greatly improves functional impairment and co-existing disease,” concluded Welbourn. “The report clearly demonstrates that bariatric surgery is an obvious strategy to reduce premature mortality from the major causes of death.”

About the NBSR

The NBSR is a web-based registry (built by Dendrite Clinical Systems) allowing surgeons and their delegates to enter data in a safe and secure manner from any hospital, provided they have appropriate access. To gain access to view, add new or edit existing data, each user has their own ID and password.

These are issued only to registered bariatric surgeons and their designated, named delegates. Each user can only see their own data, and not data belonging to any other surgeon. Access to the database as a whole is restricted and the database design is controlled by the NBSR Database Committee.