Six-year analysis of 30-day post-operative leaks for primary sleeve gastrectomy: a MBSAQIP database study
Narayan Osti,Ameer Aboud,Shamon Gumbs,Raja Sabbagh,Stephen Carryl,Sharique Nazir,Javier Andrade,Karina McArthur
DOI: https://doi.org/10.1007/s00464-024-11190-2
2024-09-01
Abstract:Background: Sleeve gastrectomy is the most performed bariatric surgery. Post-operative gastric sleeve leaks, although rare, are dreaded complications. This study aims to perform an updated investigation of the factors associated with sleeve leaks. Methods: This retrospective cohort study analyzed 692,554 cases from the MBSAQIP database (2016-2021) with CPT code 43,775 for primary sleeve gastrectomy. We excluded emergency operations, conversions/revisions, endoscopic interventions, patient with prior foregut surgery, and open operations. Multivariate logistic regression analysis (STATA version 15) was performed to identify factors associated with sleeve gastrectomy leaks. Results: Out of 692,554 patients, 600,910 (86.77%) patients underwent laparoscopic sleeve gastrectomy, and 91,644 (13.23%) patients underwent robotic sleeve gastrectomy. 1179 (0.17%) developed leaks within 30 days; 177(0.19%) were in the robotic group and 1002 (0.17%) in the laparoscopic group with no significant difference in leak rates between two groups on multivariate analysis. Black patients had lower odds of having leaks as compared to white patients (Odds Ratio (OR): 0.68 (0.56-0.82); p < 0.01). Hispanic patients had lower odds of having leak as compared to non-Hispanics. Factors associated with higher leak odds (p < 0.05) included hypertension, GERD, smoking, immunosuppression, increased operating time, and albumin < 3.5 g/dl. Higher odds of leaks were observed in years 2016-2019 vs 2020-2021 (OR: 1.44 (1.25-1.65), p < 0.01). Higher odds of leak in operations with general surgeons compared to bariatric surgeons was found (OR: 1.46 (1.04-2.02), p = 0.02); observed only on robotic group on subgroup analysis (OR: 2.2 (1.2-4.2), p = 0.02). Staple line reinforcement, oversewing, and performance of leak test showed no differences in leak rate. Bougie size and distance from pylorus were not associated with changes in leak rate. Conclusion: This study provides updated insights into the factors associated with sleeve leaks, reinforcing information gained from prior studies. A higher association of leak among general surgeons could represent a learning curve for new robotic general surgeons. The overall decreasing trend for gastric sleeve leak is encouraging and may be a sign of improved techniques.