Development and validation of a nomogram for predicting mortality in patients with vertebral osteomyelitis

Retrospective study.

From January 2019 to January 2025, patients diagnosed with vertebral osteomyelitis at Shandong Public Health Clinical Center were enrolled. Clinical data were analyzed using Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariable logistic regression to identify risk factors. A prognostic nomogram was developed and evaluated for discrimination using the area under the receiver operating characteristic curve (AUC), calibration using the Hosmer-Lemeshow test and calibration curves, clinical utility through decision curve analysis, and robustness with 500 bootstrap resamples for internal validation. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated at the optimal cutoff.

Of 248 patients, 55 (22.17%) died within 1 year. Eight independent risk factors were identified: renal insufficiency, higher Charlson Comorbidity Index, elevated C-reactive protein and white blood cell levels, multifocal vertebral osteomyelitis, pneumonia, poor cardiac function, and poor wound healing. The nomogram demonstrated excellent discrimination (AUC: 0.957; 95% CI 0.929-0.985), confirmed by bootstrap validation (AUC: 0.946; 95% CI 0.915-0.977). At a cutoff of 0.135, it achieved 82.5% sensitivity, 93.3% specificity, 88.7% positive predictive value, 81.9% negative predictive value, and 92.7% accuracy. Calibration was adequate (Hosmer-Lemeshow P = 0.750), and decision curve analysis confirmed clinical benefit.

This pioneering nomogram accurately predicts 1-year mortality in vertebral osteomyelitis, facilitating personalized clinical management.