Performance of a diagnostic algorithm for fibrotic hypersensitivity pneumonitis. A case-control study

Area of Science:

• Pulmonary Medicine
• Diagnostic Accuracy
• Idiopathic Interstitial Pneumonias

Background:

• Differentiating fibrotic hypersensitivity pneumonitis (HP) from idiopathic pulmonary fibrosis (IPF) is clinically significant yet challenging.
• Current HP diagnostic guidelines recommend multidisciplinary discussion (MDD), but its availability is limited.
• A novel diagnostic algorithm for HP, excluding MDD, was previously proposed.

Purpose of the Study:

• To evaluate the diagnostic accuracy of a proposed HP diagnostic algorithm that omits MDD.
• To assess the prognostic validity of the algorithm-derived diagnoses compared to MDD.
• To determine the optimal confidence threshold for the algorithm's clinical utility.

Main Methods:

• A case-control study was conducted using patients with confirmed MDD diagnoses of fibrotic HP or IPF.
• The proposed HP diagnostic algorithm's performance was tested at diagnostic confidence levels of ≥90% and ≥70%.
• Diagnostic test characteristics (sensitivity, specificity) and prognostic validity using Cox proportional hazards models were analyzed.

Main Results:

• The study included 31 fibrotic HP and 50 IPF patients.
• At ≥90% confidence, the algorithm showed high specificity (0.94) but low sensitivity (0.35) for HP diagnosis.
• At ≥70% confidence, the algorithm demonstrated improved performance with 0.90 specificity and 0.74 sensitivity (J-index 0.64).
• MDD-based HP diagnosis was linked to a significantly lower mortality risk (aHR 0.10), unlike algorithm-derived diagnoses.

Conclusions:

• The algorithm-derived diagnosis at ≥70% confidence offers satisfactory test performance for identifying MDD-diagnosed HP.
• The algorithm's sensitivity at ≥90% confidence was insufficient for reliable HP diagnosis.
• The superior prognostic value of MDD-based HP diagnosis suggests that the algorithm may not fully replace MDD in clinical practice.