No Rush to Judgment for Bariatric Surgery

Abstract

It isunclearwhether current studieswill address critical questions about the long-term outcomes of bariatric surgery, including the sustainability ofweight loss and comorbidity control and long-term complication rates.1 These critical gaps in knowledge pose a significant problem for people considering a potential surgical option to treat severe obesity. Contributing to these deficits are the paucity of comparative trials, incomplete follow-up, a lack of standardized definitions for changes inhealth status (eg, diabetesmellitus remission), and the tendency to a rush to judgment in favor of surgical treatment options. Laparoscopic sleeve gastrectomy is a good case in point, evolving very quickly during the last several years into the dominant procedure in use2 despite a complete void of information about the longer-termeffects. Golomb et al3 try to address this by documenting 1-, 3-, and 5-year results in a cohort of 443 sleeve gastrectomy cases but clearly raise more issues than they can answer. They show that bothweight loss and type2diabetes remissiondegrade substantiallyover time: excess weight loss from 77% to 56% and complete remission of diabetes from 51% to 20%between 1 and 5 years. Those results are fromonly56people available for 5-year followupand with rates of loss to follow-up of 50%, on average, at 1 and 3 years.Onemust assume that people lost to follow-upmaydiffer in important ways. In addition, there are no standards to report comorbid health changes following bariatric surgery, which limits the ability to compare results across studies. Golomb and colleagues do an excellent job of outlining the study’s ownspecific definitions of prevalence, incidence, and remission of type 2 diabetes and other health outcomes, but theseareneither standardnor sharedbetweenmembersof the research community. How can these problems be addressed? Large, prospective observational studies such as the Longitudinal Assessment of Bariatric Surgery Consortium study report standardized definitions and do better with long-term retention,4 but the burden ofwork and subsequent high cost are problematic inahighlycompetitive fundingenvironment.The issuesofcost and feasibility also make a large randomized trial that could compare surgical procedures across heterogeneous populations impractical. There is hope for the use of large electronic databases to contribute to these knowledge gaps, but thehandling of large amounts ofmissing data is a critical feature that is often not well articulated by the authors or well understoodby the readers. Theanswerswill likelybegeneratedover time not only by a few of these large-scale efforts but also by thoughtful inference thatwill bemade throughpooled analyses of data like that from Golomb and colleagues and from many other disparate randomized and nonrandomized studies of bariatric surgery.5 It will take time, patience, and awillingness to avoid a rush to judgment. In the meantime, clinicians andprospective patientswill need to discuss andweigh the evidence in a dynamic exchange driven not always by final conclusions but by the most current available data.