The advanced learning curve in robotic prostatectomy: a multi-institutional survey.

Several studies have attempted to define the learning curve associated with robot-assisted laparoscopic prostatectomy (RALP). These studies have focused on the acquisition of skills by novice robotic surgeons. It is unclear, however, if basic proficiency can be equated with satisfactory patient outcomes. We surveyed experienced robotic surgeons with high surgical volume in an attempt to define an "advanced" learning curve, relating to proficiency and outcomes with the robotic procedure. A questionnaire was designed to evaluate the learning curve of the RALP from basic to advanced techniques. High-volume, experienced surgeons were asked to complete this questionnaire on the basis of their personal experience with the RALP procedure. Nine institutions participated in the study accounting for a total case volume of 6,276. Median surgeon experience was 460 cases (range 325-1,500); median total operative and robotic time were 165 and 105 min, respectively. Median time to "basic proficiency" with the robot was 40 cases; proficiency in more challenging cases was approached after a median of 50 cases. Surgical outcomes were deemed satisfactory to the surgeon for continence, potency, and surgical margins after a median of 100, 200, and 300 procedures, respectively. These data confirm previous studies that basic proficiency with the robotic system occurs relatively quickly, after 25-40 cases. Obtaining "satisfactory outcomes" took substantially longer, from 100 to 300 cases. Satisfactory outcomes regarding surgical margins and potency took longer to obtain than continence, likely reflecting the relative complexity of cancer control and nerve-sparing compared with the vesico-urethral anastomosis.