AI Model Predicts Indwelling Catheter Use After Spinal Cord Injury

Clinicians need tools to anticipate conversion to indwelling catheters in this population. Anticipating this outcome could support personalized bladder management. To help address this challenge, researchers have developed a machine learning–based prognostic stratification model.

Investigators at the Fourth Military Medical University in Xi’an, China, retrospectively analyzed 135 consecutive patients with early-stage spinal cord injury-related neurogenic bladder admitted between November 2017 and January 2026. They built interpretable machine-learning survival models to predict indwelling catheter time and the probability of indwelling catheter conversion within one year. Five survival models were developed and validated for this purpose.

Across the cohort, the median indwelling catheter time was 72 days. During follow-up, 102 patients (75.56%) experienced indwelling catheter conversion. Model performance was strongest for a SHapley Additive exPlanations (SHAP)-informed, five-variable Random Survival Forest (RSF) model, which achieved a validation C-index of 0.8742 and time-dependent area under the receiver operating characteristic curve values of 0.9352, 0.8834, and 0.8803 at three, six, and 12 months, respectively.

The SHAP analysis identified five dominant, early-obtainable predictors: H-reflex, lower extremity motor score, urinary tract infection, time from lesion to rehabilitation facility, and the spinal cord independence measure bladder score. Using an RSF-derived prognostic output from these variables, the team stratified patients into favorable, intermediate, and unfavorable groups. Three-month cumulative indwelling catheter conversion incidence was 97.30%, 50.23%, and 2.22% for these groups, respectively.

The findings were published online on August 23, 2026, in Neurourology and Urodynamics. The authors noted that, following successful external validation, the tool could support personalized bladder management by implementing prognosis‑stratified clinical pathways. According to the report, this approach offers an interpretable framework that aligns early clinical assessments with individualized risk of indwelling catheter conversion.