106 Background: Using previously developed prostatectomy incontinence nomogram (PIN) we sought to externally validate the nomogram that predicts probability of incontinence at 6-, 12-, and 24-months after robot assisted radical prostatectomy (RARP). Methods: Prospective data from 663 men with prostate cancer that underwent RARP from 2010 to 2014 at two comprehensive cancer centers and three large group practices was queried. The performance of the previously developed model was evaluated using calibration plots (predicted continence rates versus observed continence rates with 95% CI obtained using Jeffrey’s prior method) and receiver operating curves (ROC). Using Expanded Prostate Cancer Index Composite (EPIC-50) Urinary Function questionnaire, perfect continence was defined as 0 pads, social continence was defined as 1 or 2 pads, and incontinence was defined as ≥ 3 pads used after RARP. Results: The 6-, 12-, and 24- month social continence rates were 77%, 88%, and 93%, respectively. Similar to the 6- and 12-month model development cohort, the external validation cohort has modest predictability with a 6- and 12- month AUC of 0.61, and 0.62, respectively. The 24-month AUC of 0.62 in the external validation cohort is worse than what was reported in the development cohort (AUC 0.80). Conclusions: The externally validated prostatectomy incontinence nomogram is generalizable but has modest 6-, 12-, and 12-month predictability in risk of incontinence after RARP.
External validation of the Soeterik nomogram for side-specific extraprostatic extension after robot-assisted radical prostatectomy
