This study aimed to identify clinical characteristics associated with the prevalence of progressive pulmonary fibrosis (PPF) in interstitial lung disease (ILD) and to develop a prognostic nomogram model for clinical use.
In this single-centered, retrospective study, we enrolled ILD patients with relatively comprehensive clinical data and assessed the incidence of PPF within a year using collected demographics, laboratory data, high-resolution computed tomography (HRCT), and pulmonary function test (PFT) results. We used a training cohort of ILD patients to identify early predictors of PPF and then validated them in an internal validation cohort and subsets of ILD patients using a multivariable logistic regression analysis. A prognostic nomogram was formulated based on these predictors, and the accuracy and efficiency were evaluated using the area under the receiver operating characteristic curve (AUC), calibration plot, and decision curve analysis (DCA).
Among the enrolled patients, 120 (39.09%) cases had connective tissue disease-associated interstitial lung disease (CTD-ILD), 115 (37.46%) had non-idiopathic pulmonary fibrosis idiopathic interstitial pneumonia (non-IPF IIP), and 35 (11.4%) had hypersensitivity pneumonitis (HP). Overall, 118 (38.4%) cases experienced pulmonary fibrosis progression. We found that baseline DLco% pred (OR 0.92; 95% CI, 8.93–0.95) was a protective factor for ILD progression, whereas combined pneumonia (OR 4.57; 95% CI, 1.24–18.43), modified Medical Research Council dyspnea score (mMRC) (OR 4.9; 95% CI, 2.8–9.5), and high-resolution computed tomography (HRCT) score (OR 1.22; 95% CI, 1.07–1.42) were independent risk factors for PPF. The AUC of the proposed nomogram in the development cohort was 0.96 (95% CI, 0.94, 0.98), and the calibration plot showed good agreement between the predicted and observed incidence of PPF (Hosmer–Lemeshow test:
ILD patients with combined pneumonia, low baseline DLco% pred, high mMRC marks, and high HRCT scores were at higher risk of progression. This nomogram demonstrated good discrimination and calibration, indicating its potential utility for clinical practice.