Previous studies evaluating the learning curve of robotic hysterectomy have used stabilization of operative times to determine surgical expertise. However, the true measure of a surgeon’s competence in performing robotic procedures should be intraoperative morbidity and patient outcomes rather than an ability to complete a procedure in a defined amount of time. The da Vinci Surgical System is a robotic procedure that has been used increasingly in gynecologic surgery. Several studies have demonstrated advantages of the robotic procedure over open or laparoscopic techniques. However, the few existing studies that have examined the learning curve associated with attaining proficiency in performing robotic hysterectomy have been limited because of small sample size and inconsistent reporting of complication rates.
The aim of this study was to examine the learning curve of robotic hysterectomy by examining intraoperative or postoperative complications and patient outcomes with increasing surgeon experience. All participants underwent robotic hysterectomy performed with the da Vinci system at Mayo Clinic, Rochester, Minn, from 2007 to 2009. Over the 36-month study period, operative experiences of 8 surgeons were compared for each 6-month period for outcomes of operative time, intraoperative or postoperative complications, and length of stay longer than 1 day. For learning curve analysis, separate standard and risk-adjusted cumulative summation curves were constructed to analyze changes in intraoperative complications (outcome 1) and any intraoperative or postoperative complications (outcome 2) for procedures performed by the 2 most experienced robotic surgeons (A and B). Using the curves for surgeons A an B, changes in surgical skill of the 6 least experienced robotic surgeons were assessed. Proficiency was defined as the point at which the cumulative summation curve for each surgeon crossed the acceptable control limit (H0), which was derived from published complication rates of abdominal hysterectomy. Cumulative summation parameters were based on a acceptable rate of 5.7% and an unacceptable rate of 11.4% for outcome 1 and an acceptable rate of 36% and an unacceptable rate of 50% for outcome 2.
Over the 3-year study period, operative time decreased significantly in 325 cases, from 3.5 hours during the first 6 months to 2.7 hours in the last 6 months. There was also a significant decrease in the proportion of patients with length of stay longer than 1 day (from 49.2% to 14.7%). Complications did not change significantly over time. The average number of attempts needed to cross the acceptable limit H0 was 91 for outcome 1 and 44 for outcome 2. The observed cumulative summation curves constructed using these estimates for each surgeon differed from the acceptable rates calculated from the curves of surgeons A and B.
These findings show that 36 months of experience with robotic hysterectomy is associated with decreased operative time and length of stay, but indicate that complications may not change with experience. The data suggest that cumulative summation analysis is an objective method to evaluate surgical proficiency and that this is achieved after performing approximately 91 robotic procedures.