Development and Validation of a Nomogram to Predict Airway Fibrostenosis in Tracheobronchial Tuberculosis

Airway fibrostenosis, a severe complication of tracheobronchial tuberculosis (TBTB), causes respiratory morbidity including atelectasis, pneumonia, and respiratory failure. Early risk prediction remains challenging due to the lack of validated assessment tools.

This retrospective cohort study analyzed TBTB patients undergoing bronchoscopic interventions between January 2021 and June 2024 with 6-month follow-up. A Cox regression model was developed in all 305 patients, internally validated with 1,000 bootstrap resamples. Performance was evaluated via C-index, ROC-AUC, calibration, and decision curve analysis. Kaplan-Meier analysis was used to stratify groups, with log-rank tests assessing differences.

Airway fibrostenosis incidence was 60.33% (184/305). Eight independent predictors were identified: symptom duration, affected lung lobes, diabetes, multiple TBTB types, bronchoscopic intervention frequency, initial sputum acid-fast bacilli smear grade, neutrophil-to-lymphocyte ratio, and CD8+ T-cell count. The nomogram demonstrated strong discrimination (C-index 0.77, 95%CI 0.75-0.81) with increasing predictive accuracy over time: 6-week AUC 0.773 (0.708-0.838), 8-week 0.792 (0.740-0.844), 12-week 0.830 (0.782-0.878), and 16-week 0.883 (0.842-0.923). High-risk patients exhibited a significantly higher probability of developing airway fibrostenosis compared to low-risk patients (P<0.001). Calibration and decision curve analyses confirmed clinical utility.

This validated nomogram effectively predicts airway fibrostenosis risk in TBTB patients, enabling early identification of high-risk individuals for targeted interventions.