Does the learning curve in robotic rectal cancer surgery impact circumferential resection margin involvement and reoperation rates? A risk-adjusted cumulative sum analysis.

BACKGROUND The aim of this study was to evaluate the impact of surgeons' learning curve in robotic surgery for rectal cancer on circumferential resection margin (CRM) involvement and reoperation rates. METHODS Learning curve data were prospectively collected from four centers. Patients undergoing robotic proctectomy for resectable rectal cancer were included. CRM was involved when ≥1 mm. TME quality was classified as complete, nearly complete, or incomplete. T-test and Chi-square tests were used to compare continuous and categorical variables, respectively. Risk-adjusted cumulative sum (RA-CUSUM) analysis was utilized to evaluate the effect of the learning curve on primary endpoints. Univariate analysis of potential risk factors for CRM involvement and reoperation was performed. Factors with the p-value ≤0.2 were included in the multivariate logistic regression model for further RA-CUSUM analysis. RESULTS A total of 221 patients (80, 36, 62, and 43 patients operated on by surgeons 1, 2, 3, and 4, respectively) who underwent robotic surgery for rectal cancer during the surgeons' learning curves were included. CRM involvement rate was 0%, 11%, 3%, and 5% in surgeons 1, 2, 3, and 4, respectively. Reoperation rate was 3.7%, 8.3%, 4.8%, and 11.6%, respectively. RA-CUSUM analysis of CRM involvement (R2=0.9886) and reoperation (R2=0.9891) found a statistically significant decreasing trend in aggregate CUSUM values throughout the learning curve. CONCLUSIONS This study found a continued significant decrease in CRM involvement and reoperation rates throughout the learning curve in robotic rectal cancer surgery.