Abstract Background and Rationale Pulmonary fibrosis (PF) occurs in children across diverse developmental, genetic, lung injury, and immune-mediated conditions. While standardized radiologic criteria have facilitated accurate diagnosis, prognosis, and clinical trial design for adults with PF, the burden and spectrum of PF remains uncertain in children. Our objectives were to assess the prevalence of features of fibrosis in chest computerized tomography (CT) reports and evaluate whether PF is explicitly recognized in clinical documentation when radiologic features are present. Methods We conducted a retrospective cohort study of all patients evaluated by our pediatric pulmonary division (2010-2025) with at least one chest CT before 21 years of age. CT interpretations were analyzed using a rule-based natural language processing algorithm to identify fibrotic descriptors (e.g., fibrosis, reticulation, traction bronchiectasis, interlobular or septal thickening, cystic lucencies). Scans mentioning fibrosis were reviewed for contextual accuracy, and two study physicians independently reviewed a random, non-overlapping sample of ten patients each to confirm PF. Structured diagnostic codes were used to identify comorbid conditions. Results A total of 8,209 patients met inclusion criteria. Of these, 228 (2.7%) had radiologic PF, confirmed by chart review in all sampled cases. Only 27 of 228 (11.8%) patients with PF on CT scan had fibrosis documented in pulmonary visit diagnosis or problem lists. Children with PF were born at lower gestational age (34.9 vs 36.7 weeks; p 0.001) and more often carried diagnostic codes for interstitial lung disease, bronchiectasis, bronchopulmonary dysplasia, pulmonary hypertension, transplant-related conditions, and neuromuscular disorders. Among 18,620 CT reports, 356 (1.9%) explicitly described fibrosis. The proportion of fibrosis-positive scans increased significantly over time: 1.5% (2010-2015), 1.9% (2016-2020), and 2.4% (2021-2025; p=0.0047), with post-hoc testing showing a significant increase between 2010-2015 and 2021-2025 (bonferroni corrected p=0.0049). Scan reports mentioning fibrosis more frequently contained reticulation (7.9% vs 1.6%), septal thickening (28.1% vs 6.6%), traction bronchiectasis (8.1% vs 0.4%), cystic lucencies (3.7% vs 0.7%), and architectural distortion (11.5% vs 2.1%) (all p0.001). Among scans without explicit fibrosis mention, 174 (0.95%) met ≥ 2 of these radiologic features. Conclusion While the prevalence of radiologist reported PF has increased in recent years, these data suggest that PF may be more frequent in children than explicitly reported by radiologists and documented by clinicians. Standardized pediatric PF imaging criteria and further validation studies are needed to improve recognition of PF and accurately phenotype patients with progressive fibrosis to advance care and enable inclusion in therapeutic trials. This abstract is funded by: None
Dunn et al. (Fri,) studied this question.