Development and validation of a claims-based algorithm to identify moderate exacerbations in patients with asthma treated in the US
Respiratory Medicine, 2024
Introduction
Definitions of moderate asthma exacerbation have been inconsistent, making their economic burden difficult to assess. An algorithm to accurately identify moderate exacerbations from claims data is needed.
Methods
A retrospective cohort study of Reliant Medical Group patients aged ≥18 years, with ≥1 prescription claim for inhaled corticosteroid/long-acting β2-agonist, and ≥1 medical claim with a diagnosis code for asthma was conducted. The objective was to refine current algorithms to identify moderate exacerbations in claims data and assess the refined algorithm's performance. Positive and negative predictive values (PPV and NPV) were assessed via chart review of 150 moderate exacerbations events and 50 patients without exacerbations. Sensitivity analyses assessed alternative algorithms and compared healthcare resource utilization (HRU) between algorithm-identified patients (claims group) and those confirmed by chart review (confirmed group) to have experienced a moderate exacerbation.
Results
Algorithm-identified moderate exacerbations were: visit of ≤1 day with an asthma exacerbation diagnosis OR visit of ≤1 day with selected asthma diagnoses AND ≥1 respiratory pharmacy claim, excluding systemic corticosteroids, within 14 days after the first claim. The algorithm's PPV was 42%; the NPV was 78%. HRU was similar for both groups.
Conclusion
This algorithm identified potential moderate exacerbations from claims data; however, the modest PPV underscores its limitations in identifying moderate exacerbations, although performance was partially due to identification of previously unidentified severe exacerbations. Application of this algorithm in future claims-based studies may help quantify the economic burden of moderate and severe exacerbations in asthma when an algorithm identifying severe exacerbations is applied first.
Authors
Roberts MH, Duh MS, Rothnie KJ, Zhang S, Czira A, Slade D, Cheng WY, Thompson-Leduc P, Greatsinger A, Zhang A, Mapel D