Analysis of the learning curve for Retzius-sparing Robot-assisted radical prostatectomy for a single surgeon

Background: The learning curve for retzius sparing robotic radical prostatectomy is not fully understood. Objective This study attempts to identify the learning-curve across the first 130 cases of a single surgeon. Design, Setting, and Participants: All retzius sparing robotic radical prostatectomy cases performed by a single surgeon at a high-volume tertiary hospital between April 2019 and July 2022 were included. Outcome Measurements and Statistical Analysis: Outcome measures included positive surgical margin rate, complication rates and unplanned readmission to hospital, postoperative urinary continence, erectile function, and PSA measurement. Cases were divided chronologically into 3 groups and differences between groups assessed. Results and Limitation: 130 RS-RARP cases were identified. Differences were found between groups in several areas. Positive surgical margin rate fell between Group 1 (30.2%) and Group 3 (9.1%). Safety, post-operative continence, erectile function and PSA remained stable. Median patient age increased between group 1 (59yrs) and Group 3 (66.5yrs) (P = 0.04). Proportion of patients with stage > T2 increased between Group 1 (27.9%) and Group 2 (41.9%) (P = 0.036). Median console time increased between Group 1 (120 mins) and Group 2 (150 mins,) (P = 0.01). Median gland weight increased between Group 1 (28g) and Group 3 (35.5g) (P < 0.001). Conclusions The positive surgical margin rate improved over the learning curve, despite the complexity of cases increasing, reflected in older patients, larger prostates and higher stage disease. Safety and functional outcomes are excellent throughout. The learning curve might be facilitated by careful case selection favouring smaller prostates with less advanced disease. Patient Summary: We analysed the learning-curve for Retzius-sparing robotic radical prostatectomy. Across the first 130 cases positive surgical margin rate fell; safety and continence remained excellent. Selection of smaller and less advanced cases may facilitate learning.