Evaluating the learning curve for single-port radical prostatectomy: a multi-institutional cohort analysis

Single-port robot-assisted radical prostatectomy (SP-RARP) represents an evolution in minimally invasive surgery but introduces unique technical challenges. Existing evidence on SP-RARP learning curves is limited to single-center or single-surgeon experiences. This multicenter study aimed to define proficiency milestones for SP-RARP across multiple surgeons using robust statistical methods.

We retrospectively analyzed data from 327 patients who underwent SP-RARP performed by eight experienced multiport surgeons transitioning to the single-port platform across three U.S. centers (2019-2024). Learning curves for operative time (OT) and estimated blood loss (EBL) were evaluated using cumulative sum (CUSUM) analysis and piecewise linear regression. Piecewise multivariable logistic regression was used to adjust for confounders.

Of the 327 cases, 111 were transperitoneal (TP) and 216 extraperitoneal (EP). No clear proficiency plateau was observed for the TP approach, with persistent variability in OT and EBL. In contrast, EP cases demonstrated breakpoints at 55 cases for OT (95% CI: 53.8-55.6; R² = 0.91) and 97 cases for EBL (95% CI: 92.3-101.1; R² = 0.97). Multivariable analysis confirmed a significant association between case volume and OT (p = 0.048), but not for EBL (p = 0.12).

This is the first multicenter study to evaluate SP-RARP learning curves using CUSUM and piecewise regression. Clear proficiency thresholds were identified for the EP approach (OT at 55, EBL at 97 cases), though only OT was confirmed after adjustment. No definitive plateau was reached for the TP approach, suggesting different learning dynamics.