A multi-institutional study of 3,794 patients undergoing robotic-assisted laparoscopic radical prostatectomy to determine the surgical learning curve for positive margins and operating time.

Abstract

102 Background: The surgical learning curve for robotic assisted laparoscopic radical prostatectomy (RALP) is often cited as being shorter than for other surgical modalities. However, while this appears true with regards to surgical safety, the learning curve for more refined variables like positive surgical margin (PSM) rate and operative time (OT) is not well established. Our objective was to assess the surgical learning curve for RALP in terms of these parameters. Methods: We performed a retrospective cohort study of 3,794 patients who underwent RALP between Jan 2003 and Sep 2009 by three surgeons (DL, PW, AKT) from three centers (UPenn, Karolinska, Cornell). Mean overall PSM rates and mean overall OT were calculated for all three surgeons at intervals of 50 RALPs per surgeon, and learning curves for these means were fit using a loess method. R version 2.71 was used for all statistical analysis. Results: The learning curve for PSM rates for all patients demonstrated improvements that continued with greater surgeon experience, with over 1,600 cases required to get a PSM rate <10%. When only pT3 patients were evaluated, the learning curve started to plateau after 1,000-1,500 cases. Mean OT plateaued after 750 cases although with further surgical experience the OTs started to climb again. Conclusions: The learning curve for RALP is not as short as previously thought, and a large number of cases are needed to get PSM rates and OTs to a minimum. This suggests that RALP should be performed by high volume surgeons in order to optimize patient outcomes. [Table: see text]