Development and validation of a prognostic nomogram model for severe osteomyelitis patients.

After severe infection in osteomyelitis patients in the Intensive Care Unit (ICU), there's a higher risk of mortality. However, limited research exists on predicting prognosis. Develop a predictive model for 1-year mortality risk in ICU-admitted osteomyelitis patients to inform clinical diagnosis and treatment. MIMIC IV database was used to retrieve ICU data for osteomyelitis patients. The data were randomly split into training and validation sets (7:3 ratio). Univariate and multiple logistic regression identified independent predictors of one-year mortality and constructed a risk prediction nomogram in the training set. Predictive value of the nomogram was assessed using C-indexes, ROC curves, DCA, CIC and calibration curves. This study included a total of 1153 osteomyelitis patients, with 137 deaths within one year. These patients were randomly divided into training (N = 807) and validation (N = 346) sets. In the training set, multiple features were identified as key predictors of one-year mortality in osteomyelitis patients in the ICU. These factors were incorporated into the nomogram model, demonstrating good identification performance, with AUCs of 0.872 and 0.826 for the training and validation sets, respectively. The calibration curve and ROC curve indicate excellent predictive accuracy. DCA suggests strong clinical utility and robust predictive efficiency. Further analysis through CIC illustrates the clinical effectiveness of this predictive model. We have developed a nomogram model to predict the 1-year mortality rate for osteomyelitis patients admitted to the ICU, providing valuable predictive information for clinical decision-making.