Impact of diagnostic guidelines on the diagnosis of hypersensitivity pneumonitis

Abstract

IntroductionHypersensitivity pneumonitis (HP) is an immune-mediated interstitial lung disease from exposure to environmental antigens. Diagnosing HP could be challenging. The American College of Chest Physicians (CHEST) and American Thoracic Society/Japanese Respiratory Society/and Asociación Latinoamericana del Tórax (ATS/JRS/ALAT) have published diagnostic guidelines in 2021 and 2020 respectively. The CHEST guideline uses four grades of confidence: confident (>90%), provisional high (70–89%), provisional low (51–69%), and unlikely (≤50%). The ATS/JRS/ALAT guideline uses five grades of confidence: definite (>90%), high (80–89%), moderate (70–79%), low (51–69%) and not excluded (≤50%). In this study, we determined how these two guidelines could have affected the diagnosis of HP made before the guidelines.MethodsTwo hundred and fifty-nine adult patients from a previous cohort with HP (ICD-9:495) made between Jan. 1, 2008, and Dec. 31, 2013, at Duke University Medical Center were included. We simplified the diagnostic confidence into three categories so we could compare the guidelines: high (≥90%), intermediate (51–89%), and low (≤50%).ResultsThere were 156 female and 103 male. Mean age was 58 (range: 20–90). 68.8% of the patients had restrictive defects (FVC < 80% pred) and 48.6% had lung biopsy. The CHEST guideline classified 33.6% of the patients into high, 59.5% into intermediate and 6.9% into low confidence categories. The ATS/JRS/ALAT guideline classified 29.7% of the patients into high, 21.2% into intermediate and 49.0% into low confidence categories (p < 0.0001 vs. CHEST). Cohen's kappa was 0.331. In patients with identifiable inciting agents (IAs) (N = 168), the CHEST guideline classified 32.1% of the patients into high, 64.3% into intermediate and 3.6% into low confidence categories. The ATS/JRS/ALAT guideline classified 29.2% of the patients into high, 20.8% into intermediate, and 50.0% into low confidence categories. Cohen's kappa was 0.314.DiscussionIn our HP cohort with two-thirds of the patients with restrictive defects, we found the two guidelines had fair agreement in diagnosing HP with or without identifiable IAs. They agreed more when the diagnostic confidence was high. When the diagnostic confidence was lower, however, the ATS/JRS/ALAT guideline was more stringent. Clinicians should be aware of the differences between the two guidelines when evaluating patients suspicious of HP.