The risk prediction model for acute urine retention after perineal prostate biopsy based on the LASSO approach and Boruta feature selection.

Abstract

Objective: One known side effect of transperineal (TP) prostate biopsies is acute urine retention (AUR). We aimed to create and evaluate a predictive model for the post-paracentesis risk of acquiring AUR.

Methods: This study included 599 patients undergoing prostate biopsies (April 2020-July 2023) at the Second Affiliated Hospital of Nantong University, selected based on abnormal digital rectal examination and/or PSA (prostate-specificantigen) > 4 ng/mL. Acute urinary retention (AUR) was defined as the inability to void within 72 hours post-biopsy, requiring catheterization. Patients were randomly divided into training (419 cases) and test (180 cases) sets. Univariate logistic analysis and feature selection Boruta and LASSO (Least absolute shrinkage and selection operator) identified predictors, followed by multivariate logistic regression to develop a predictive nomogram for AUR. Internal validation used the test set, with model performance assessed via the c-index, ROC (Receiver Operating Characteristic) curve, calibration plot, and decision curve analysis. The nomogram demonstrated strong discrimination, calibration, and clinical utility for AUR risk prediction.

Results: In 86 patients (14.3%), AUR happened. An examination of multivariate logistic regression revealed six distinct risk variables for AUR. Based on these independent risk factors, a nomogram was constructed. The training and validation groups' c-indices showed the model's high accuracy and stability. The calibration curve demonstrates that the corrective effect of the training and verification groups is perfect, and the area under the receiver operating characteristic curve indicates great identification capacity. DCA (Decision Curve Analysis) curves, or decision curve analysis, demonstrated the model's significant net therapeutic effect.

Discussion: The nomogram model created in this work can offer a personalized and intuitive analysis of the risk of AUR and has intense discrimination and accuracy. It can help create efficient preventative measures and identify high-risk populations.