(324) Identification of Individuals with Severe Asthma in United States Claims Data

Background: Given the increased likelihood for severe asthmatics (3.7% of all asthmatics, according to the Global Initiative for Asthma [GINA]) to experience comorbidities, disease complications, emergency room visits, and hospitalizations, the ability to stratify asthma populations on severity is often important. Although pharmacoepidemiologic studies sometimes categorize asthma severity, there is no consensus on an approach. Understanding the efficiency and validity of methods to identify severe asthmatics in claims is necessary for future studies.

Objectives: To develop and compare algorithms to classify asthma patients by severity in claims data.

Methods: Individuals age ≥6 years diagnosed with asthma (≥2 ICD-10 J45) between January 2017 and September 2019 were identified in Optum’s de-identified Clinformatics® Data Mart Database. All included patients had ≥365 days prior continuous enrollment, during which they were unexposed to dupilumab and undiagnosed with chronic obstructive pulmonary disease (ICD-10 J41–J44). Severe asthma was inferred, consistent with GINA, from prescription claims for high-dose inhaled corticosteroids (ICS) in combination with long-acting beta-agonists (LABA). Two algorithms were employed. The “code-based method” considered high-dose ICS-LABA to be an inhaler property and defined severe asthma based on presence of US National Drug Codes from our pre-determined list. Alternatively, the “calculation-based method” considered high-dose ICS-LABA to be a regimen property and defined severe asthma based on derived patient-level daily ICS dose. Descriptive cohort characteristics, including the frequency of ICD-10 coding for severe persistent asthma (J45.5), were compared.

Results: A total of 484,621 asthmatics were identified, 2.3% of which were severe based on the code-based method and 3.9% based on the calculation-based method. No meaningful differences were observed in the age and sex distributions of the two cohorts. Calculation-based severe asthmatics not captured by the code-based method (59%) suggest physician-directed inhaler usage to vary from US Food and Drug Administration (FDA) suggestions. Only 4.6% and 3.3% of severe asthmatics from the code- and calculation-based methods, respectively, had a severe persistent asthma diagnosis code. Consistency of calculation-based severe asthma prevalence (3.9%) with the gold-standard cited by GINA (3.7%) suggest this method to have face validity.

Conclusions: Estimation of exposure to high-dose ICS-LABA as a proxy for severe asthma status is subject to limitations without details of physician instructions for inhaler use, which may vary from FDA-recommended maximum daily inhalations. Future work will further validate these asthma severity algorithms using electronic medical records.