Learning Curve for Robotic-Assisted Laparoscopic Colorectal Surgery of the Pelvis: Analysis of 50 Consecutive Cases

Objective:Robotic-assisted laparoscopic surgery is evolving as an important surgical approach in the field of colorectal surgery. Robotic techniques may help to overcome the limitations of laparoscopic and open surgery especially when operating in the narrow pelvis and neighboring critical structures. The aim of this study was to evaluate the learning curve of robotic-assisted laparoscopic procedures involving resections of the rectum and rectosigmoid.

Methods:Data were entered in a prospective database and later abstracted for analysis. Fifty consecutive patients underwent robotic-assisted laparoscopic surgery (RALS) between August 2008 and September 2009 (surgery types: abdominoperineal resection (APR), anterior rectosigmoidectomy (AR), lower anterior resection (LAR), and rectopexy (RP)) by a single surgeon (E.M.H.). Demographic data were tabulated including patient gender, age, body mass index (BMI), and American Society of Anesthesiologists (ASA) score. Intraoperative parameters including docking time (DT), surgeon console time (SCT), and total operative time (OT) were analyzed. The learning curves were analyzed using the cumulative sum (CUSUM) method.

Results:Surgery was performed on 22 (44%) female and 28 (56%) male patients. Twenty-six (52%) AR, 14 (28%) LAR, 6 (12%) APR, and 4 (8%) RP were performed. Mean age was 54.4±13.1 years (range: 24-82 years), BMI was 27.8±6.3 kg/m2 (range: 16-49.4 kg/m2), and median ASA was 2 (range: 2-4). Mean DT was 14.0±7.7 min (range: 6-45 min), SCT was 115.1±46.9 min (range: 40-210 min), and OT was 246.1±80.7 min (range: 90-540 min). DT and SCT accounted for 6.3% and 46.8% of OT, respectively.Given that the largest room for surgeon improvement occurs at the console, the SCT learning curve was analyzed in depth. The cumulative sum of SCT was plotted against the number of days after the initial robotic surgery. The SCT learning curve was best modeled as a second-order polynomial with equation CUSUMSCT in minutes = 0.01*days2 – 4.3*days – 4.4 (R=0.94). Upon inspection, the SCT learning curve was observed to consist of three unique phases, which were comprised of the initial 15, middle 10, and final 23 cases, respectively. The table below summarizes the line of best fit and patient characteristics for each phase of the SCT learning curve.

Conclusion:The three phases identified in this study represent characteristic stages of the surgeon’s learning curve for robotic colorectal procedures. Phase I represents the true learning curve phase which was found to be 15 cases. The phase II plateau represents increased competence with the robotic technology. Phase III was achieved after 25 cases and represents the mastery phase in which more challenging cases was performed.

Session: Poster

Program Number: P566

View Poster

« Return to SAGES 2010 abstract archive