Construction of a nomogram to predict the probability of new vertebral compression fractures after vertebral...

Area of Science:

• Orthopedic Surgery
• Biomedical Engineering
• Radiology

Background:

• Osteoporotic vertebral compression fractures (OVCFs) are a significant cause of morbidity.
• Predicting new vertebral fractures post-surgery is crucial for optimizing treatment and patient outcomes.
• Current predictive models for refracture after OVCF surgery require enhancement.

Purpose of the Study:

• To develop and validate a novel nomogram for predicting new vertebral fractures following surgery for OVCFs.
• To enhance surgical planning and minimize the occurrence of secondary vertebral compression fractures.
• To provide a clinically applicable tool for risk stratification in OVCF patients.

Main Methods:

• A cohort of 420 patients with OVCFs was randomly divided into training (80%) and validation (20%) sets.
• Least Absolute Shrinkage and Selection Operator (LASSO) regression identified key predictive factors.
• Multivariate logistic regression and nomogram construction were employed for model development.

Main Results:

• Significant predictors for refracture included cement leakage, poor cement dispersion, and endplate fractures.
• The nomogram demonstrated strong predictive ability with high Area Under the Curve (AUC) values (0.974 training, 0.965 validation).
• The model exhibited high accuracy (Brier scores 0.043 training, 0.070 validation) and excellent calibration and discriminant ability.

Conclusions:

• Cement leakage, poor cement dispersion, and endplate fractures are critical factors in predicting refractures after OVCF surgery.
• The developed nomogram is a simple, accurate, and feasible tool for clinical decision-making in managing OVCFs.
• This predictive model has the potential to optimize surgical strategies and improve outcomes for patients with osteoporotic vertebral compression fractures.