LAUSR

Health disparities adversely affect people based on characteristics that are historically linked to discrimination or exclusion. There is limited research on understanding consequences of health disparities in the diagnosis and treatment of obstructive sleep apnea (OSA). This retrospective study sought to investigate health disparities affecting OSA patients using real-world data. A national sample of administrative claims data from OSA patients who had a claim for a sleep test (home sleep test or polysomnography) was used. Gender, insurance payer, race/ethnicity, and 3-digit zip code were characterized from the claims data at the time of the first sleep test to describe differences across patients. The sample included about 2 million patients with 44.9% female, an average age 51.0 years, and an insurance payer breakdown of 74% commercial, 18% Medicaid, and 8% Medicare Advantage. There were a number of challenges that made it difficult to draw conclusions about the impact of health disparities in this data. The source of gender data was unknown; therefore, we do not know if it represents identified gender or biological sex. Race/ethnicity data was listed as “unknown” for 89% of commercial patients, 18% of Medicaid patients, and 36% of Medicare Advantage patients. Data reporting did not allow for details on multiple races or separation of race and ethnicity. Using 3-digit zip codes proved to be unreliable for drawing conclusions about a patient’s residence and access to care as 3-digit zip code areas are heterogenous and could cover a small section of a major city, a large rural portion of a state, or both. Lastly, ICD-10 Z codes, which describe social determinants of health (SDOH), appear to be underutilized. Currently, using administrative claims as a real-world data source to assess health disparities is difficult due to data incompleteness, lack of adequate data definitions, and data aggregations made to protect patient privacy. Strategies to increase usability of claims data for investigating health disparities may include improvements in payer data reporting, increased utilization of SDOH Z codes, and linking to other more patient-centric data sources. Analyzing large samples of real-world data may help identify disparities and improve care for all patients.