Diagnostic performance of a machine-learning algorithm (Asthma/COPD Differentiation Classification; AC/DC) tool versus primary care physicians and pulmonologists in asthma, COPD and ACO

Background: Differential diagnosis of asthma and chronic obstructive pulmonary disease (COPD) poses a challenge in clinical practice, and their misdiagnosis results in inappropriate treatment, increased exacerbations and potentially even death.
Objective: To investigate the diagnostic accuracy of the Asthma/COPD Differentiation Classification (AC/DC) tool compared with primary care physicians and pulmonologists in asthma, COPD, and asthma-COPD overlap (ACO).
Methods: The AC/DC machine learning-based diagnostic tool was developed using 12 parameters from electronic health records of >400,000 patients aged ≥35 years. An expert 98 panel of 3 pulmonologists and 4 general practitioners from 5 countries evaluated 119 patient cases from a prospective observational study and provided a confirmed diagnosis (n=116) of asthma (n=53), COPD (n=43), ACO (n=7) or other (n=13). The cases were then reviewed by 180 primary care physicians and 180 pulmonologists from 9 countries and by AC/DC tool, and diagnostic accuracies were compared with reference to the expert panel diagnoses.

Results: Average diagnostic accuracy of the AC/DC tool was superior to primary care physicians (median difference, 24%; 95% posterior credible interval [CrI]: 17–29%; PPdiagnostic accuracies were 73%, 50% and 61% by AC/DC tool, primary care physicians, and pulmonologists versus expert panel diagnosis, respectively.
Conclusion: The AC/DC tool demonstrated superior diagnostic accuracy compared with primary care physicians and pulmonologists in diagnosis of asthma and COPD in patients aged ≥35 years and has the potential to support physicians in the diagnosis of these conditions in clinical practice.