Surgical Performance Metrics for 1-Year Patient-Reported Outcomes After Radical Prostatectomy

Importance: There is a dearth of surgical performance measures that accurately predict long-term patient outcomes. Objective: To develop surgical performance measures collected at the time of surgery that accurately predict future outcomes. Design, Setting, and Participants: In this cohort study, surgical video of 157 patients undergoing robotic-assisted radical prostatectomy by 28 surgeons from 4 tertiary referral hospitals across the US and 1 hospital in Germany was analyzed. Four trained and blinded raters annotated video clips of the bilateral nerve-sparing step using standardized tools for identifying surgical gestures and assessing technical skills. Patients were prospectively enrolled from July 2016 to January 2023 and followed up for 12 months postoperatively. Data were analyzed from April to August 2024. Patients with clinically localized prostate cancer undergoing robotic-assisted radical prostatectomy were eligible. Inclusion criteria included adequate erectile function prior to surgery and access to complete surgical video. Interventions/Exposures: Robotic-assisted radical prostatectomy. Main Outcomes and Measures: Performance metrics were compared between patients who recovered erectile function and those who did not. Erectile function recovery was defined as achieving erections sufficient for intercourse measured using the Sexual Health Inventory for Men. Results: Fifty-three patients (34%) recovered erectile function at 12 months after surgery. The median age was 64 (IQR, 59-68) years and median body mass index was 28 (IQR, 26-30). In total, 80 957 surgical gestures were annotated and 2568 technical skills scores were evaluated. The impact of performance factors on erectile function recovery was evaluated using univariate logistic regression. Recovery was associated with a greater proportion of peel/push gestures (odds ratio [OR], 1.72; 95% CI, 1.24-2.42, per 0.1 increase; P = .001), lower proportion of energy gestures applied to the neurovascular bundle (OR, 0.35; 95% CI, 0.13-0.81, per 0.1 increase; P = .03), and less gestures grabbing the neurovascular bundle (OR, 0.02; 95% CI, 0.00-0.47, per 0.1 increase; P = .02). Erectile function recovery was associated with higher tissue handling skill scores (OR, 3.43; 95% CI, 1.23-10.90, P = .03). On multivariable regression the association between peel/push gestures and erectile function recovery remained significant (OR, 1.66; 95% CI, 1.18-2.39, per 0.1 increase; P = .005). Conclusions and Relevance: Surgical performance can be assessed from data collected during surgery and used to predict erectile function 12 months later. This was not previously feasible due to a lack of quantitative methods for assessing surgical performance. Combining surgical gestures and skills assessment demonstrates a novel opportunity for advancing surgical performance.