A Nomogram Model for Early Mortality Risk Stratification in Elderly Patients with Idiopathic Pulmonary Fibrosis: An...

Background: Idiopathic pulmonary fibrosis (IPF) has a low incidence but high mortality. Simple prognostic tools for elderly IPF patients in primary care are lacking. This study aimed to develop an accessible nomogram for this population. Methods: A retrospective analysis was conducted on elderly IPF patients from the First Affiliated Hospital of Anhui Medical University (January 2016-February 2023). Early mortality was defined as death within 12 months from diagnosis. A nomogram was developed using predictors identified by univariate and multivariate analyses. To minimize overfitting, we limited the number of predictors to four based on the rule of at least 10 events per variable (39 events). Model performance was assessed using the C-index, calibration curves including the Hosmer-Lemeshow goodness-of-fit test, ROC analysis, and decision curve analysis (DCA). Internal validation was performed using bootstrap resampling (1000 iterations). For missing data, variables with >30% missingness were excluded; for variables with ≤30% missingness, multiple imputation was applied. Risk stratification was performed based on nomogram scores, and survival between groups was compared via Kaplan-Meier analysis. Results: Overall, 83 patients were included. Multivariate analysis identified age, monocyte count, globulin, and DLCO%pre as independent predictors of early mortality. The nomogram incorporating these factors achieved a C-index of 0.846 (optimism-corrected C-index: 0.812 after bootstrap). The AUCs for predicting 1-, 2-, and 3-year overall survival were 0.879, 0.896, and 0.854, respectively. The Hosmer-Lemeshow test showed good calibration (p = 0.42, 0.38, and 0.51 for 1, 2, and 3 years). Kaplan-Meier analysis confirmed significantly worse survival in the high-risk group (p < 0.0001). Conclusions: We developed an accurate and practical nomogram to predict prognosis in elderly IPF patients, offering a useful risk-assessment tool for primary care settings. However, external validation in independent multicenter cohorts is required before clinical implementation. At its current stage, the model should be regarded as hypothesis-generating.