Cumulative sum analysis (CUSUM) for evaluating learning curve (LC) of robotic-assisted laparoscopic partial nephrectomy (RALPN)

Robotic-assisted laparoscopic partial nephrectomy (RALPN) is becoming a standard treatment for localized renal tumors worldwide. Data on the learning curve (LC) of RALPN are still insufficient. In the present study, we have attempted to gain further insight in this area by evaluating the LC using cumulative summation analysis (CUSUM). A series of 127 robotic partial nephrectomies were performed by two surgeons at our center between January 2018 and December 2020. CUSUM analysis was used to evaluate LC for operative time (OT). The different phases of surgical experience were compared in terms of perioperative parameters and pathologic outcomes. In addition, multivariate linear regression analysis was used to confirm the results of the CUSUM analysis by adjusting the phases of surgical experience for the other confounding factors that may affect OT. The median age of patients was 62 years, mean BMI was 28, and mean tumor size was 32 mm. Tumor complexity was classified as low, intermediate, and high risk according to the PADUA score in 44%, 38%, and 18%, respectively. The mean OT was 205 min, and trifecta was achieved in 72.4%. According to the CUSUM diagram, the LC of OT was divided into three phases: initial learning phase (18 cases), plateau phase (20 cases), and mastery phase (subsequent cases). The mean OT was 242, 208, and 190 min in the first, second, and third phases, respectively (P < 0.001). Surgeon experience phases were significantly associated with OT in multivariate analysis considering other preoperative and operative parameters. Surgical outcome was comparable between the three phases in terms of complications and achievement of trifecta; hospital stay was shorter in the mastery phase than in the first 2 phases (4 days vs 5 days, P = 0.02). The LC for RALPN is divided into 3 performance phases with CUSUM. Mastery of surgical technique was achieved after performing 38 cases. The initial learning phase of RALPN has no negative impact on surgical and oncologic outcomes .