Overcoming the Arduous Transition for Robotic Hepatopancreatobiliary Cases: A Multi-Procedure Learning Curve Study Utilizing CUSUM Analysis.

Abstract

Quick optimization and mastery of a new technique is an important part of procedural medicine, especially in the field of minimally invasive surgery. Complex surgeries such as robotic pancreaticoduodenectomies (RPD) and robotic distal pancreatectomies (RDP) have a steep learning curve; therefore, findings that can help expedite the burdensome learning process are extremely beneficial. This single-surgeon study aims to report the learning curves of RDP, RPD, and robotic Heller myotomy (RHM) and to review the results' implications for the current state of robotic hepatopancreaticobiliary (HPB) surgery. This is a retrospective case series of a prospectively maintained database at a non-university tertiary care center. Total of 175 patients underwent either RDP, RPD, or RHM with the surgeon (DRJ) from January 2014 to January 2020. Statistical significance of operating room time (ORT) was noted after 47 cases for RDP (p < 0.05), 51 cases for RPD (p < 0.0001), and 18 cases for RHM (p < 0.05). Mean ORT after the statistical mastery of the procedure for RDP, RPD, and RHM was 124, 232, 93min, respectively. No statistical significance was noted for estimated blood loss or length of stay. Robotic HPB procedures have significantly higher learning curves compared to non-HPB procedures, even for an experienced HPB surgeon with extensive laparoscopic experience. Our RPD curve, however, is quicker than the literature average. We suggest that this is because of the simultaneous implementation of HPB (RDP and RPD) and non-HPB robotic surgeries with a shorter learning curve-especially foregut procedures such as RHM-into an experienced surgeon's practice. This may accelerate the learning process without compromising patient safety and outcomes.