Assessing the learning curve of robotic sacrocolpopexy.

Abstract

The aim was to evaluate the learning curve of robotic sacrocolpopexy, adjusted for surgical risk. The charts of 145 robotic sacrocolpopexies performed by urogynecologists at Mayo Clinic, Rochester, MN, USA, from 2007 to 2013, were reviewed. Outcomes of interest included operative time, intraoperative complications, and postoperative complications with a Clavien-Dindo grade 2 or higher. Risk-adjusted cumulative summation analysis was performed by comparing a calculated complication risk score with observed patient outcomes, and then cumulatively recalculating the rate of expected vs observed complications after each procedure. Proficiency was defined as the point at which the surgeon's complication rates were better than expected, given the patient's risk factors. The median operative time decreased significantly, from 5.3 to 3.6 h, during the 7-year period, and plateaued after the first 60 cases. A higher ASA classification was associated with an increased risk of intraoperative complications (p = 0.02), and a higher Charlson comorbidity index was associated with an increased risk of intraoperative or postoperative complications (p = 0.01). In risk-adjusted CUSUM analyses, accounting for these factors, and for body-mass index and vaginal parity, proficiency was identified at 55 cases for intraoperative complications and 84 cases for intraoperative or postoperative complications. Operative time plateaued after the first 60 cases, whereas complication rates continued to decrease beyond this. Proficiency, as determined by a risk-adjusted CUSUM analysis for complication rates, was achieved after approximately 84 cases. Evaluation of postoperative complications in addition to intraoperative complications, in a risk-adjusted model, is critical in depicting the surgical learning curve.