Background: The stepwise treatment approach recommended by the Global Initiative for Asthma (GINA steps) can be a useful proxy for asthma severity in studies using health claims data. However, the GINA 2019 report made major changes from previous versions that may pose challenges when applied to health claims data, e.g. low dose ICS-formoterol is a recommended controller across steps 1-3.
Objectives: 1) Develop a US claims-based algorithm to define patients’ GINA step using 2019 report, 2) describe the distribution of GINA steps, including patients who could not be clearly classified, and 3) compare the distributions of 2019 GINA steps and 2016 GINA steps.
Methods: Using the GINA reports, we defined potential treatment regimens for each treatment step. We selected patients aged 12+ between April 2019-March 2020 from Optum’s de-identified Clinformatics® Data Mart Database. GINA step was assessed based on medications filled in the 90 days before the most recent asthma related clinic visit. Single medication fills within 30 days of each other were considered combination therapy. Inhaled corticosteroid (ICS) daily dose was calculated using number of inhalers dispensed, puffs per inhaler, strength per puff, and days supplied, and mapped to GINA dose categories (low, medium, high). For comparison, we replicated this approach in 2017 claims data using 2016 guidelines.
Results: Among 386,851 eligible asthma patients, the algorithm classified patients as follows: 230,082 (59.5%) to steps 1-3, 94,081 (24.3%) to step 4 and 45,529 (11.8%) to step 5. Within steps 1-3, 11,853 (3.1%) received low-dose ICS/formoterol and could not be classified more specifically based on the 2019 report. The remaining 218,229 patients included 81,541 (21.1%) step 1, 83,985 (21.7%) step 2, and 52,703 (13.6%) step 3. 353,078 patients were identified in 2017, according to the 2016 GINA report, 202,978 (57.5%) were classified to step 1-3 (23.0%, 21.1% and 13.4%, respectively), 116,389 (33%) to step 4, and 20,544 (5.8%) to step 5.
Conclusions: Distribution of the 2019 GINA steps was broadly in line with 2016 steps, suggesting that it may be possible to use 2019 GINA based algorithm as a proxy for disease severity in US claims data. Despite ambiguity in GINA guidelines for patients taking low-dose ICS/formoterol, a relatively small proportion of patients were left unclassified. Future research may be able to leverage additional assumptions to classify these patients. In comparison to 2016 guidelines, a higher proportion of patients in 2019 were classified as Step 5, perhaps due to high dose ICS-LABA being moved to step 5. Further validation of the proposed algorithm is needed.
