Are Age and Voided Volume Sufficient to Predict Detrusor Underactivity in Women? An External Validation Study

Detrusor underactivity (DU) is a significant cause of female lower urinary tract symptoms. We aimed to externally validate a non-invasive prediction model using age and voided volume (VV) in a tertiary center cohort.

This retrospective study included 450 consecutive women who underwent uroflowmetry and subsequently multichannel urodynamics (UDS) between 2014 and 2025. DU was defined by pressure-flow criteria as in the development study. Model performance was assessed via AUC-ROC, calibration plots, Brier scores, and Decision Curve Analysis (DCA).

The prevalence of DU in our cohort was 16.2% (73/450), compared to 10% in the original development study. The model demonstrated poor discrimination with an AUC of 0.575 (95% CI 0.50-0.65), a significant decrease from the original reported AUC of 0.82. At the suggested cut-off, sensitivity was 78%, but specificity was only 32%. Calibration was poor (slope = 0.134; intercept = -1.397), indicating a systematic overestimation of risk. The Brier Skill Score was negative (-0.154), and DCA confirmed that the model provided negligible net clinical benefit over the default "test-all" or "test-none" strategies.

Validation of the proposed model failed to demonstrate statistical reliability in our cohort. These findings suggest that the model lacks the necessary transportability to serve as a clinical screening tool for female DU, underscoring the continued need for invasive testing to ensure diagnostic accuracy.