# Cracking the LUTS Code: A Pre-Urodynamic Tool for DU vs. BOO Diagnosis in Female Patients with Non-Neurogenic LUTS

## TL;DR

This study creates a non-invasive tool to help doctors tell the difference between two bladder conditions in women without using invasive tests.

## Contribution

A novel clinical prediction model is developed to distinguish detrusor underactivity from bladder outlet obstruction in women with lower urinary tract symptoms.

## Key findings

- The model achieved a C-index of 0.779, indicating moderate predictive accuracy.
- Key predictors included hesitancy, incomplete emptying, POP-Q staging, and UFL time to Qmax.
- At a probability cutoff of 0.3, the model had 86.8% sensitivity and 82.1% negative predictive value.

## Abstract

Background: Detrusor underactivity (DU) and bladder outlet obstruction (BOO) are common causes of voiding dysfunction in women with lower urinary tract symptoms (LUTS). However, differentiating between them remains challenging due to overlapping clinical presentations and a reliance on invasive urodynamic studies (UDS). This study aimed to develop a non-invasive, office-based clinical prediction model to distinguish DU from BOO in women with non-neurogenic LUTS. Methods: We conducted a retrospective analysis of 88 women who underwent pressure-flow studies at two outpatient clinics between 2012 and 2022. DU was defined using a projected isovolumetric pressure 1 (PIP1) < 30 cm H2O, and BOO was defined by a Female-Specific Bladder Outlet Obstruction Index (BOOIf) > 18. Clinical symptoms, uroflowmetry (UFL) parameters, and pelvic organ prolapse staging (POP-Q) were evaluated. A multivariate logistic regression model was constructed using a stepwise selection procedure. Results: Of the 88 patients, 38 (43.2%) were diagnosed with DU and 50 (56.8%) with BOO. Four predictors were retained in the final model: hesitancy (OR = 2.06, p = 0.18), incomplete emptying (OR = 3.52, p = 0.02), POP-Q < 3 (OR = 0.15, p = 0.02), and longer time to Qmax on UFL (OR = 1.05, p = 0.004). The model achieved a Harrell’s Concordance Index (C-index) of 0.779. Using a probability cutoff of 0.3, the model demonstrated a sensitivity of 86.8%, specificity of 46.0%, positive predictive value of 55.0%, and negative predictive value of 82.1%. Conclusions: We present a novel non-invasive prediction model incorporating clinical symptoms, UFL metrics, and pelvic exam findings that may aid in differentiating DU from BOO in women with LUTS.