A Matched Cohort Study Comparing Perioperative Outcomes in Robotic Versus Open Pancreaticoduodenectomy

Abstract

Introduction: Robotic pancreaticoduodenectomy (RPD) may offer technical, quality of life, or cost advantages over open pancreaticoduodenectomy (OPD). Methods: OPD and RPD performed from 2015-2020 were retrospectively matched 2:1 by age, sex, BMI and pathology. To account for matching, continuous variables were compared using generalized estimating-equations, and categorical variables using the Cochran-Mantel-Haenzel test. Subset analysis was conducted for pancreatic adenocarcinoma. Results: A matched cohort of 94 OPD and 47 RPD showed comparable demographics and pathology. Oncologic quality outcomes such as negative surgical margins, lymph nodes sampled, and operative time were equivalent (Table). RPD was associated with decreased estimated blood loss (EBL) (median 250mL vs 350mL, p=0.03), surgical site infections (9% vs 23%, p=0.04), abscess (6% vs 20%, p=0.04) and median length of stay (LOS) (6 days vs 8 days, p=0.05) for RPD compared to OPD. Readmission favored RPD (11% vs 27%, p=0.06), however, it did not reach significance in this limited cohort. There was no difference between the two groups with respect to pancreatic fistula, delayed gastric emptying, or hemorrhage. In subgroup analysis of pancreatic adenocarcinoma, LOS remained lower for RPD (6 vs 7.5, days, p=0.05), while oncologic quality measures were preserved. Conclusion: RPD is associated with lower EBL, postoperative infection, and LOS compared to OPD, with comparable oncologic quality. RPD is feasible and safe and should be considered in appropriately selected patients to minimize blood loss and postoperative complications. Given the improvement to perioperative complications and LOS, the potential for RPD to improve costs of care delivery should be investigated.