Technique or technology? Evaluating leaks after gastric bypass

OBJECTIVE: To assess the relationship between technique and surgical devices on anastomotic and staple-line leaks after laparoscopic Roux-en-Y gastric bypass. BACKGROUND: Leaks after bariatric surgery remain a major source of morbidity and mortality. The association of surgical technique and devices with leaks after gastric bypass is poorly understood. SETTING: Multi-centered study that included teaching and non-teaching hospitals that participate in a statewide consortium for quality improvement using a payer-funded outcome registry. METHODS: We analyzed data from the Michigan Bariatric Surgery Collaborative and performed a case-control study comparing patients who sustained a leak with those who did not after primary laparoscopic Roux-en-Y gastric bypass. A total of 71 (.44%) patients with leaks were identified between January 2007 and December 2011. The leak group was matched 1:2 to a control group (nonleak) based on procedure type, age, body mass index, sex, and the year in which the procedure was performed. Technique-specific case characteristics and device-specific factors were assessed by reviewing operative notes from all primary bariatric procedures in our study population. RESULTS: The rate of leak decreased during the study period, and there was a significant downward trend (slope estimate: -.19961%, P = .0372). After performing multivariate analysis, the type of anastomosis (circular stapler, hand-sewn, or linear stapler) and stapler manufacturer were not associated with leaks. The use of buttressing material was associated with a higher rate of leaks (odds ratio: 8.79 [95% confidence interval: 2.49-31.01], P = .0007), whereas the use of fibrin sealant was associated with a lower rate of leaks (odds ratio .11 [95% confidence interval: .03-.41], P = .0013). These findings could not be explained by differences in measures of surgeon performance. CONCLUSION: Leak rates after laparoscopic gastric bypass have fallen in Michigan despite variations in technique and device utilization. Although the type of anastomosis and stapler manufacturer do not appear to be significantly associated with leaks, it appears that the use of buttressing material was more common in cases in which leaks occurred, whereas the use of fibrin sealant was not. Given the complex interplay of multiple variables that affect surgical outcomes, future studies justifying the benefits of operative devices should be evaluated prospectively in the context of surgeon technique and skill