A detailed analysis of the learning curve: Robotic hysterectomy and pelvic-aortic lymphadenectomy for endometrial cancer

Objective. To define the learning curve for robotic hysterectomy and pelvic-aortic lymphadenectomy for endometrial carcinoma. Methods. Patient demographics and segmental operative times on all patients at one institution who underwent robotic comprehensive surgical staging (hysterectomy, pelvic and aortic lymphadenectomy) for endometrial cancer were prospectively collected. Patients were arranged in order based on surgery date and outcomes were compared between quartiles (cases 1-20, 21-40, 41-60, and 61-79). Proficiency was defined as the point at which the slope of the curve becomes less steep for operative times. Efficiency was defined as the point at which the slope is zero. ANOVA or Fisher's exact test was used to compare the procedure times. Locally weighted regression generated smoothed lines that represent operative time over the sequence of the operations. Results. 79 patients were comprehensively staged robotically. While age, the percentage of patients with >= 2 co-morbidities, number of patients with previous laparotomy, EBL, LOS and lymph node counts do not differ between groups, the first 20 patients had a lower BMI compared to the next 20 (27 vs. 34 kg/m(2), P=0.009). Operative times decreased from the first 20 cases to next 20, but was not significantly changed over the next three quartiles. Each component of the procedure has a separate learning curve. Conclusions. Proficiency for robotic hysterectomy with pelvic-aortic lymphadenectomy for endometrial cancer is achieved after 20 cases; however, the number of procedures to gain efficiency varies for each portion of the case and continues to improve over time. (C) 2009 Elsevier Inc. All rights reserved.