Construction and validation of an acute exacerbation risk score for fibrotic idiopathic interstitial pneumonia.

BACKGROUND: Acute exacerbation (AE) of fibrotic idiopathic interstitial pneumonia (f-IIP) is a severe complication associated with high short-term mortality, yet standardized AE prediction tools remain limited. This study aimed to develop and internally validate a clinically accessible risk scoring system for predicting AE in patients with f-IIP. METHODS: This single-centre retrospective cohort study enrolled 638 patients with f-IIP, including 358 patients with idiopathic pulmonary fibrosis (IPF) and 280 with other f-IIPs. A nomogram and simplified risk score were developed based on independent predictors identified by multivariate Cox regression. Internal validation using 1,000 bootstrap resamples was performed to evaluate discrimination, calibration, and clinical utility. Subgroup analyses were conducted for IPF and other f-IIP patients. RESULTS: Multivariate Cox analysis identified six independent predictors of AE: age, diffusing capacity of the lung for carbon monoxide percent predicted (DLCO %pred), forced vital capacity percent predicted (FVC %pred), minimum oxygen saturation during the 6-minute walk test (6MWT Min SpO2), data-driven texture analysis (DTA) fibrosis score, and neutrophil-to-lymphocyte ratio (NLR). Patients were stratified into low-risk (0 points, n=68), intermediate-risk (1 point, n=11), and high-risk (≥2 points, n=559) groups. The risk score showed good discrimination for AE prediction at 24 months, with an area under the receiver operating characteristic curve (AUC) of 0.843. Kaplan-Meier analysis showed significant differences in AE-free survival among the three risk groups (P<0.001), and the score was also associated with overall survival (P=0.001). Subgroup analyses showed stronger discrimination in other f-IIP patients (AUC =0.849) and acceptable discrimination in IPF patients (AUC =0.729). No significant interaction was observed between risk score and disease subtype (P=0.32). CONCLUSIONS: This multidimensional risk scoring system showed good predictive performance for AE in f-IIP and may support individualized risk stratification and clinical monitoring.