Logo image
Back
IP66-25 PREDICTING SUCCESS AFTER ANTERIOR URETHRAL STRICTURE TREATMENT USING TRAUMA AND UROLOGIC RECONSTRUCTIVE NETWORK OF SURGEONS (TURNS) DATA
Abstract   Peer reviewed

IP66-25 PREDICTING SUCCESS AFTER ANTERIOR URETHRAL STRICTURE TREATMENT USING TRAUMA AND UROLOGIC RECONSTRUCTIVE NETWORK OF SURGEONS (TURNS) DATA

Kevin D. Li, Mi-Ok Kim, Hiren V. Patel, Nejd F. Alsikafi, Joshua A. Broghammer, Jill C. Buckley, Sean P. Elliot, Jeremy B. Myers, Andrew C. Peterson, Keith F. Rourke, …
The Journal of urology, Vol.215(5S), p.e1322
05/2026
DOI: 10.1097/01.JU.0001191672.88727.3a.25

View Online

Abstract

INTRODUCTION AND OBJECTIVES: Surgical success in anterior urethral stricture disease (aUSD) is difficult to predict. We developed models using the newly introduced LSE staging system for aUSD to predict multi-stage repair, non-glans meatus, and 5-year retreatment. METHODS: We analyzed 3,620 men undergoing urethroplasty at ten TURNS centers (2010-2019). Predictors included patient factors (age, BMI, smoking), disease characteristics (aUSD stage: length, etiology, location), prior procedures, and repair type. Logistic regression models were developed for each outcome and internally validated using 1,000 bootstrap resamples. Model performance was summarized with area under the receiver operating characteristic curve (AUC) for discrimination and calibration plots with mean absolute error (MAE). RESULTS: The cohort comprised 3,620 patients with a median age of 49 years and BMI of 29. Smoking prevalence was 10% active, 25% former, and 65% never-smokers. 55% had no prior urethrotomies. Stricture stages were 1A (25%), 1B (3.0%), 2A (24%), 2B (9.5%), 2C (3.4%), 3A (14%), 3B (7.4%), 4A (7.1%), 4B (1.7%), and 5 (4.5%). The model for predicting multistage repair demonstrated excellent discrimination (AUC 0.88, 95% CI 0.85–0.91) and near-perfect calibration (MAE 0.003). Models for non-glans meatus (AUC 0.91, MAE 0.007) and 5-year retreatment (AUC 0.77, MAE 0.014) also performed well. A web-based calculator is available to generate individualized probabilities of each outcome: kevndli. shinyapps. io/lseprediction. CONCLUSIONS: We developed clinically interpretable prediction tools that translate disease staging into actionable estimates of surgical success, allowing clinicians to counsel patients using data-driven probabilities and to benchmark outcomes across centers.

Details

Metrics

1 Record Views
Logo image