Robot-assisted radical cystectomy (RARC) with intracorporeal neobladder - what is the effect of the learning curve on outcomes?

OBJECTIVE: To evaluate the effect of the learning curve on operative, postoperative, and pathological outcomes of the first 67 totally intracorporeal robot-assisted radical cystectomies (RARCs) with neobladders performed by two lead surgeons at Karolinska University Hospital. PATIENTS AND METHODS: Between December 2003 and October 2012, 67 patients (61 men and six women) underwent RARC with orthotopic urinary diversion by two main surgeons. Data were collected prospectively on patient demographics, peri- and postoperative outcomes including operation times, conversion rates, blood loss, complication rates, pathological data and length of stay (LOS) for these 67 consecutive patients. The two surgeons operated on 47 and 20 patients, respectively. The patients were divided into sequential groups of 10 in each individual surgeon's series and assessed for effect of the learning curve.
RESULTS: Patient demographics and clinical characteristics were similar in both surgeons' groups. The overall total operation times trended down in both surgeons' series from a median time of 565 min in the first group of 10 cases, to a median of 345 min in the last group for surgeon A (P < 0.001) and 413 to 385 min for surgeon B (not statistically significant). Risk of conversion to open surgery also decreased with a 30% conversion rate in the first group to zero in latter groups (P < 0.01). Overall complications decreased as the learning curve progressed from 70% in the first group to 30% in the later groups (P < 0.05), although major complications were not statistically different when compared between the groups. Patient demographics did not change over time. The mean estimated blood loss was unchanged across groups with increasing experience. The pathological staging, mean total lymph node yield and number of positive margins were also unchanged across groups. There was a decrease in LOS from a mean of 19 days in the first group to a mean (range) of 9 (4-78) days in the later groups, although the median LOS was unchanged and therefore not statistically significant.
CONCLUSIONS: Totally intracorporeal RARC with intracorporeal neobladder is a complex procedure, but it can be performed safely, with a structured approach, at a high-volume established robotic surgery centre without compromising perioperative and pathological outcomes during the learning curve for surgeons. An experienced robotic team and mentor can impact the learning curve of a new surgeon in the same centre resulting in decreased operation times early in their personal series, reducing conversion rates and complication rates.