Development and validation of a nomogram model for predicting bone cement leakage during vertebroplasty

To develop and validate a nomogram model for predicting the risk of bone cement leakage (BCL) during vertebroplasty, providing a visual tool to assist preoperative risk assessment and clinical decision-making.

A retrospective analysis was conducted on patients who underwent vertebroplasty at our hospital between January 2022 and June 2025. Patients were divided into two groups based on the occurrence of BCL. Univariate and multivariate Logistic regression analyses were performed to identify independent risk factors, which were then used to construct a combined predictive model. Based on the regression results, a nomogram was established. The discriminative ability of the model was evaluated using the receiver operating characteristic (ROC) curve. Calibration was assessed using a Bootstrap resampling calibration curve, and clinical utility was examined through decision curve analysis.

Univariate analysis indicated significant differences between the two groups in bone mineral density, operative time, surgical type, fracture vertebral segment, presence of multiple fractures, cortical disruption of the vertebra, degree of vertebral compression, bone cement injection volume, cement viscosity during injection, postoperative vertebral height restoration rate, and hypertension (all P < 0.05). Multivariate Logistic regression identified bone mineral density, multiple fractures, vertebral cortical disruption, degree of vertebral compression, and bone cement injection volume as independent predictors of cement leakage. Validation of the nomogram demonstrated good discrimination and excellent calibration.

The nomogram model exhibits robust predictive performance and clinical applicability. It may offer a visual tool for individualized preoperative risk assessment in vertebroplasty.