Evaluation of the learning curve of robot-assisted laparoscopic ventral mesh rectopexy

Background The current standard treatment for external rectal prolapse and symptomatic high-grade internal rectal prolapse is surgical correction with minimally invasive ventral mesh rectopexy using either laparoscopy or robotic assistance. This study examines the number of procedures needed to complete the learning curve for robot-assisted ventral mesh rectopexy (RVMR) and reach adequate performance. Methods A retrospective analysis of all primary RVMR from 2011 to 2019 performed in a tertiary pelvic floor clinic by two colorectal surgeons (A and B) was performed. Both surgeons had previous experience with laparoscopic rectopexy, but no robotic experience. Skin-to-skin operating times (OT) were assessed using LC-CUSUM analyses. Intraoperative and postoperative complications were analyzed using CUSUM analyses. Results A total of 182 (surgeon A) and 91 (surgeon B) RVMRs were performed in total. There were no relevant differences in patient characteristics between the two surgeons. Median OT was 75 min (range 46-155; surgeon A) and 90 min (range 63-139; surgeon B). The learning curve regarding OT was completed after 36 procedures for surgeon A and 55 procedures for surgeon B. Both before and after completion of the learning curve, intraoperative and postoperative complication rates remained below a predefined acceptable level of performance. Conclusions 36 to 55 procedures are required to complete the learning curve for RVMR. The implementation of robotic surgery does not inflict any additional risks on patients at the beginning of a surgeon's learning curve