Robotic Versus Open Pancreaticoduodenectomy: A Single-Center Analysis of Safety and Efficacy Using Inverse Probability of Treatment Weighting

Abstract

Background/Objectives: Robotic surgery has emerged as a promising alternative to traditional approaches, offering enhanced precision and ergonomics. However, its application to pancreaticoduodenectomy (PD) remains controversial due to technical complexity and a steep learning curve. This study aims to evaluate the perioperative outcomes of robotic versus open PD at a single high-volume center, using inverse probability of treatment weighting (IPTW) to control for confounding variables. Methods: A retrospective, single-center analysis was conducted on patients who underwent PD between January 2020 and December 2024. Perioperative and oncological outcomes were compared using IPTW, aiming to balance potential confounding variables. The primary endpoint was 90-day postoperative mortality; secondary endpoints included operative time, estimated blood loss, major postoperative complications (Clavien-Dindo grade ≥ IIIa), and length of hospital stay. Results: Seventy-four patients were included in the intention-to-treat analysis (open: n = 31; robotic: n = 43). After IPTW adjustment, baseline characteristics were balanced across the groups. Operative time was significantly longer in the robotic group (median 540 vs. 479 min, p = 0.009). No significant differences were observed in 90-day mortality (4.9% both groups, p = 0.998), postoperative complications ≥ IIIa, or histopathological outcomes, including R0 resection and lymph node yield. Trends toward reduced estimated blood loss and shorter hospital stay in the robotic group did not reach statistical significance. Conclusions: Robotic pancreaticoduodenectomy demonstrates comparable safety and oncological efficacy to the open approach, with longer operative times but potential perioperative advantages. These findings support the continued integration of robotic techniques even in medium-volume centers with appropriate expertise.