Sacrocolpopexy (SCP) with mesh interposition is one of the most effective surgical procedures for level 1 apical defects of pelvic organ prolapse and has been shown to have one of the highest long-term success rates. Recent enthusiasm has been gained for use of robotic-assisted laparoacopic surgery to perform various gynecologic surgery including sacrocolpopexy. With the expansion of new surgical techniques, evaluating and understanding of the associated surgical learning curve is also important while avoiding potential complications. Surgical learning curves have analyzed with multiple outcomes including operative time, intra / post-operative complication rate, length of hospitalization and etc. Akl et al. reported a decrease in mean operative time after 10 cases when performing robotic sacral colpopexy. And Geller et al. revealed an improvement in several key steps of robotic sacral colpopexy after 20 cases. Zanten et al. noted that surgery time stabilized after24-29 cases with use of risk-adjusted cumulative summation (CUSUM) analysis. The aim of this study was to describe the surgical technique of robotic-assisted sacrocolpopexy (RASCP) and evaluate its feasibility, safety, learning curve, and perioperative complications

This is a retrospective chart review. Women who underwent robotic assisted laparoscopic sacrocolpopexy at our institution from June 2018 and June 2022 by one surgeon with experience in laparoscopic sacrocolpopexy. Docking time was defined as the time required to position the robot and to dock the robotic arms to the corresponding port sites. Surgeon console time was the actual time the surgeon spent at the robotic console during the procedure. Total operative time spanned the time from incision to final closure. Demographic data, intraoperative parameters, as well as postoperative outcomes were analyzed.  
The learning curves were generated by plotting operative time with B-spline regression. The Proficiency was defined as the point where the slope changed from a steep to a more moderately changing slope. Efficiency was defined as the point at which the slope approaches 0. Operative time was further analyzed using 2 additional methods. The learning rate and learning plateau were estimated using the nonlinear regression curve, y =a +b(1/x), with x being the sequential number of cases and y being operative time in minutes. The learning plateau (the theoretically best
achievable score, which is obtained as x approaches infinity). And the learning rate as the number of cases needed to achieve 90% of the learning plateau (ie, x=[10b/a])
Cumulative sum control chart (CUSUM) analysis was performed to detect differences in the surgeon’s performance and to determine proficiency. Results were put into a graph, in which the X-axis represented the number of procedures, and the Y-axis represented the cumulative surgery time. CUSUM analysis was also performed to detect differences in total surgery time. CUSUM analysis was used as a self-assessment tool. The mean surgery time was calculated. This value was used as a reference value. When a surgery took more or less time than the mean surgery time, the graph would rise or fall with the absolute difference, respectively.
The procedures were scored sequentially, based on operation date and time. Data were divided into 2 groups of first-13th procedures and 14th–41st procedures to look for differences in patients’ risk profile. Intraoperative and postoperative complications were defined as any deviation from the ideal operative course. Because the main objective is surgeon’s performance, conversions because of adhesions, anesthetics, or malfunction of the robot were excluded.

Statistical analysis was performed with SPSS statistics software (version 25.0; IBM, Armonk, NY). Data were presented as number and percentage for categoric data, mean, standard deviation, or median and range for continuous data. All tests were considered significant at .05 level

A total of 41 cases of robot-assisted sacral colpopexy was analyzes after excluding one case of conversion due to severe pelvic adhesion. The mean age was 56 years (SD: 8.6) and body mass index was 24.1 kg/m2(SD:2.4). The proportion of postmenopausal woman was 68.3%. The proportions of previous prolapse surgery and previous abdomen-pelvic surgery were 17.1% and 17.1%.  The  proportions of POP-Q staging  2, 3 and  4  were 51.2%, 46.3% and 2.4%. The concomitant  surgeries were performed  as follows - robotic assisted total hysterectomy (90.2%), subtotal hysterectomy (9.8%), bilateral salpingo-ophorectomy(24.4%), anti-incontinence surgery(7.3%), repair of  rectocele (17.1%). The operative time was 226.8 minutes(SD: 71.6). The mean day of hospitalization was 5.7 days (SD:0.8). There were no intra-operative and short-term post-operative complications. Based on the scatterplots of operative time, proficiency was noted at 13 cases and efficiency at 20 cases. After 20 cases, the mean operative time was 192.7 minutes. Operative time was also analyzed using a nonlinear regression curve.

In Figure 1, learning curve was obtained learning plateau and the learning rate. Based on the curve fitting, the best model to represent the learning curve of the surgeon is y =193.9 + 313.3(1/x), where x is the surgery number and y is the operative time in minutes. Therefore, the learning plateau is achieved at y = 193.9 minutes. The learning rate for the procedure is approximately 16.2 cases (ie, after 16 procedures, the surgeon achieved an operative time within 90% of the learning rate). Data were divided into 2 groups of first-13th procedures and 14th–41st procedures to look for differences in patients’ risk profile based on the turning case of proficiency. There were no differences between two groups (data are not shown) except operative time. Using operative time as end point, operative time started to diminish after 10 cases and learning curve of 18 cases was identified by CUSUM analysis(Figure 2)

There is a difference in the operative time learning curve for robotic sacrocolpopexy depending on the used statistical analysis. By integrating these results, surgical proficiency is attained after 18 cases and operative time was stabilized after then. However, CUSUM analysis based on the complications or conversion was not available due to no case of complication and a case of conversion even in the small case series.