Robotic-Assisted vs Laparoscopic Bariatric Surgery and 30-Day Serious Adverse Events: A National MBSAQIP Analysis.

Abstract

Robotic platforms are widely adopted in bariatric surgery, but whether they improve short-term safety in already standardized, low-morbidity operations remains uncertain. We performed a retrospective cohort study of 949,507 adults undergoing minimally invasive sleeve gastrectomy, Roux-en-Y gastric bypass, or duodenal switch/single-anastomosis duodeno-ileal bypass with sleeve gastrectomy at MBSAQIP-accredited US centers from 2020 to 2024. Robotic-assisted laparoscopy (n=291,483) was compared with conventional laparoscopy (n=658,024). The primary endpoint was 30-day composite serious adverse events. Anatomy-stratified, conversion-aware multivariable logistic regression with robust standard errors and marginal standardization estimated adjusted absolute risk differences. After adjustment for patient demographics, comorbidities, operative year, and conversion status, robotic assistance was associated with higher 30-day serious adverse event risk across procedure types and operative contexts. Adjusted excess risk ranged from 2.4 events per 1,000 primary sleeve gastrectomy cases (95% CI 1.5-3.4) to 15.2 events per 1,000 primary duodenal switch/single-anastomosis duodeno-ileal bypass with sleeve gastrectomy (DS/SADI) cases (95% CI 5.3-25.1). Excess risk was higher in conversion than primary operations for sleeve gastrectomy and Roux-en-Y gastric bypass. Component analysis showed differences distributed across multiple event types, predominantly readmission, reoperation, and reintervention; mortality was rare and similar between platforms. Adjusted excess risk attenuated over time, but the platform-by-year interaction was not significant (p=0.20). In this national MBSAQIP cohort, robotic-assisted bariatric surgery was not associated with improved 30-day patient safety and was associated with small adjusted excess serious adverse event risk. Prospective studies incorporating surgeon- and center-level data are needed before short-term safety benefit is attributed to robotic adoption.