A Novel Risk Prediction Model to Triage Difficult Urethral Catheterizations

Abstract

Objective: To construct a risk prediction model to identify cases of difficult urethral catheterizations (DUC) in order to prevent complications from improper placement.

Materials and methods: Using a single-institution database of urologic consults for Foley catheterizations from June 2016 to January 2020, a model to predict DUC in male patients was constructed. DUC was defined as requiring the use of a guidewire, cystoscopy, urethral dilation, and/or suprapubic tube (SPT) placement, while a simple Foley was defined as an uncomplicated placement of a regular or coudé catheter. A final model to predict DUC was constructed using multivariable logistic regression and internally validated using bootstrap statistics.

Results: A total of 841 consults were identified, with 181 (21.5%) classified as a DUC. On multivariable regression, patient-specific factors as overweight BMI (OR: 1.71; P = .014), urethral stricture disease (OR: 7.38; P < .001), BPH surgery (OR: 2.47; P < .001), radical prostatectomy (OR: 4.32; P = .001), and genitourinary (GU) prosthetic implants (OR: 3.44; P = .046) were associated with DUC. Situational factors such as blood at the meatus (OR: 2.40; P < .001), and consulting team (eg, surgery OR: 4.82; P < .001) were also significant. Bootstrap analysis of the final model demonstrated good overall accuracy (predictive accuracy: 75%).

Conclusion: This model is a promising tool to help providers identify patients who likely require catheterization by a urologist and potentially reduce catheterization-related complications. The high rate of uncomplicated catheterizations also highlights the need for continuing education amongst healthcare professionals. External validation and application to the initial Foley encounter will shed light on its overall utility.