Intraoperative performance errors predict patient outcomes in robot-assisted radical prostatectomy

To evaluate the association between objectively assessed intraoperative surgical performance and postoperative functional outcomes following robot-assisted radical prostatectomy (RARP).

Prospectively collected data from 100 patients undergoing bilateral nerve-sparing RARP by seven experienced robotic surgeons were analysed. To control for age-related confounding, a subgroup of 66 subjects was selected from the initial cohort using age-matching. These patients, all of whom were preoperatively potent, were identified from an audited outcomes registry with complete video recordings. In all, 14 trained surgeons trained to score the recordings using validated RARP-specific metrics for procedure steps, errors, critical errors, and nerve-sparing phase errors.

There were no significant pre- or postoperative differences in demographic or clinical variables between patients with good or poor outcomes. The mean inter-rater reliability of video assessments was high (0.91). There were no significant differences in the number of steps completed or critical errors made. Intraoperative performance in cases resulting in erectile dysfunction at 12 months was characterised by 43% more total errors and 67% more neurovascular bundle (NVB) dissection errors (P < 0.001) compared to cases with good functional outcomes. Age-matched analysis confirmed differences of 31-54% (P < 0.001). The number of intraoperative errors, particularly during NVB dissection, was strongly associated with clinical outcomes (P = 0.01-0.008).

The objectively assessed intraoperative performance of the surgeon strongly predicted the clinical outcome of the procedure at 1 year. The more errors the surgeon made the more likely the patient was to have erectile dysfunction at 12 months.