Why can’t you answer the important questions about Weight Loss Surgery?

20 Mar 2023 | Bariatric Surgery, Evidence Based Medicine

How many people regain weight after weight loss surgery?
What is the difference in long term outcomes between gastric bypass surgery and gastric sleeve surgery?
Is weight regain more common after gastric sleeve or gastric bypass?
What are the risk of long term complications after weight loss surgery?

These questions and others are really good and really important questions that people want to know before having weight loss surgery.  Unfortunately they are also really difficult questions to answer well, and although many people will have their opinions and theories, these can be misleading even when they come from reputable sources.  In this post I will explain why these questions are so difficult to answer, and reference back to it when presenting more reliable evidence answering these questions.  The two main limitations that impact on our ability to answer these questions are differences between people choosing different operations (or no operation), other than the operation itself, and the loss to follow-up that increases with time after surgery.

Firstly when trying to compare any two treatments (or a treatment with no treatment), it is important that the only difference between people having one treatment and another is the treatment itself.  In practice this is very difficult to achieve as there are many reasons a person may chose one treatment over another such as their age and general health, and these differences themselves may impact on the outcome such as weight loss.  The statistical name for this effect is confounding, and it explains why there is a correlation between ice cream consumption and water rescues.  More ice creams are eaten on hotter days, and more people go swimming on hotter days which leads to more water rescues, but eating ice cream does not cause people to get in distress in the water.

In weight loss surgery if people who chose a gastric sleeve are different to people who chose a gastric bypass then these differences, rather than the operation, can influence their outcome.  The only way to ensure this does not happen is to randomly assign people to the treatment, although it is hard to find people willing to let chance chose their operation.  Groups of people can also be matched using statistics if all other factors that can influence outcome are known and measured, but in practice this is very difficult to do well.  An example of this issue in weight loss surgery is the now historical lap band operation.  When the lap band was popular there were many surgeons publishing their patients’ outcomes after this surgery which were excellent.  However patients having lap bands tended to be younger and with less weight to lose, both of which meant their weight loss outcomes would look better.  It was not until two randomized studies comparing people undergoing lap bands versus gastric bypass were published that it was clear that weight loss following lap bands was much less.

Secondly there is the issue of follow-up.  Surgery happens at a single point in time, and although follow-up early after surgery is good, most people become less likely to follow-up with their surgical team, as more time passes.  This has been shown to be true for many different types of surgery, and in most cases is because people are well and have busy lives, and returning to their surgeon is not a priority.  In some cases though, a person’s outcome makes them more or less likely to follow-up.  For example someone with significant weight regain may be less likely to return for follow-up, or someone with a complication may be more likely to return for follow-up.  This ‘loss to follow-up’ is the biggest problem with studies on weight loss surgery, where long term follow-up results are really important.  Unfortunately the only real solution is to try to ensure that follow-up is as complete as possible.  This is one of the motivations behind the ANZMOSS bariatric surgery registry.

This has been a brief primer about what is known as evidence based medicine and how it applies to weight loss surgery.  Overall it is a huge topic but I would encourage anyone assessing any claims about weight loss surgery to ask themselves are the groups of people choosing two different treatments really the same, and how complete is the follow-up when looking at any long term outcomes.