Implementation of a robotic pancreatoduodenectomy program: Navigating the learning curve with a liberal patient selection and conversion strategy.

Abstract

The robotic platform is being increasingly utilized to perform pancreatoduodenectomy, but implementation can be associated with a steep learning curve, and large number of cases is required to surmount it. We detail an approach for implementation of a robotic pancreatoduodenectomy program that incorporates a liberal patient selection and conversion strategy to achieve proficiency while maintaining outcomes. Consecutive patients undergoing pancreatoduodenectomy from January 2018 to June 2025 were identified. A robotic pancreatoduodenectomy program was implemented in October 2023. Difference-in-difference models with patients identified in the National Surgical Quality Improvement Program during the same period as a control cohort were used. A total of 205 patients underwent pancreatoduodenectomy, 127 in the preimplementation period and 78 in the postimplementation period. Of the 78 pancreatoduodenectomies performed in the postimplementation period, 62 (79.5%) were performed robotically with a conversion rate of 19.4%. Compared with the preimplementation cohort, the postimplementation cohort had similar complication and mortality rates but shorter median length of stay (5 days vs 8 days, P < .0001). On difference-in-difference analyses, the institutional cohort was associated with an increase in robotic use after program implementation (+74.8%, P < .001) compared with the National Surgical Quality Improvement Program control cohort. The institutional cohort was also associated with fewer pancreatic fistulas (-12.3%, P = .02), shorter length of stay (-2.0 days, P < .001), and similar 30-day major morbidity (-9.3%, P = .14) and readmission (+7.3%, P = .13), as well as mortality rates (-1.7%, P = .45) after program implementation. A robotic pancreatoduodenectomy program with a liberal patient selection and conversion strategy can be safely implemented while preserving overall outcomes compared with National Surgical Quality Improvement Program benchmarks.