Construction of a nomogram to predict the probability of new vertebral compression fractures after vertebral augmentation of osteoporotic vertebral compression fractures: a retrospective study.

Objective: This study aimed to develop and validate a new nomogram model that can predict new vertebral fractures after surgery for osteoporotic compression fractures to optimize surgical plans and reduce the incidence of new vertebral compression fractures. Methods: 420 patients with osteoporotic vertebral compression fractures were randomly sampled using a computer at a fixed ratio; 80% of the patients were assigned to the training set, while the remaining 20% were assigned to the validation set. The least absolute shrinkage and selection operator (LASSO) regression method was applied to screen the factors influencing refracture and construct a predictive model using multivariate logistic regression analysis. Results: The results of the multivariate logistic regression analysis showed a significant correlation between bone cement leakage, poor cement dispersion, the presence of fractures in the endplate, and refractures. The receiver operating characteristic curve (ROC) results showed that the area under the ROC curve (AUC) of the training set was 0.974 and the AUC of the validation set was 0.965, which proves that this prediction model has a good predictive ability. The brier score for the training set and validation set are 0.043 and 0.070, respectively, indicating that the model has high accuracy. Moreover, the calibration curve showed a good fit with minimal deviation, demonstrating the model's high discriminant ability and excellent fit. The decision curve indicated that the nomogram had positive predictive ability, indicating its potential as a practical clinical tool. Conclusion: Cement leakage, poor cement dispersion, and presence of fractures in the endplate are selected through LASSO and multivariate logistic regressions and included in the model development to establish a nomogram. This simple prediction model can support medical decision-making and maybe feasible for clinical practice.