A Novel Nomogram for Individually Predicting 30-Day Pneumonia Mortality Risk in ILD Patients with Long-Term Use of Glucocorticoid

Objective : Long-term glucocorticoid use in patients with interstitial lung disease (ILD) is associated with a significantly increased risk of death within 30-day following pneumonia, indicating poor prognosis. This study aims to identify the risk of mortality after pneumonia onset to optimize treatment strategies and enhance patient management. Methods : This study retrospectively analyzed ILD pneumonia patient data from DRYAD. Patients were randomly split into training and validation sets. LASSO regression selected predictive factors, and a nomogram model was built. ROC curves and AUCs assessed the model's 30-day mortality prediction. Bootstrap resampling (500 times) on the validation set confirmed the model's robustness with a 95% CI for AUC. The model's calibration and discrimination were evaluated in both sets. Results : A total of 324 patients with ILD who developed pneumonia were included in this study, among which 82 patients died within 30-day. LASSO regression identified respiratory failure, vasoactive drug use, ventilator use, and lymphocytopenia as predictors for constructing a nomogram model. The model showed good calibration in both training and validation datasets, with AUCs of 0.897 (95% CI: 0.8642-0.9292) and 0.903 (95% CI: 0.8680-0.9321), respectively. Decision curve analysis suggested clinical benefits when the threshold probability was <77%. Conclusion : The nomogram developed in this study effectively predicts the 30-day mortality risk in patients with ILD following pneumonia, demonstrating strong discrimination and calibration. This provides a valuable tool for optimizing treatment strategies and improving patient outcomes.