Phenotyping to Identify Mental Health Trends of Transgender Individuals Using Private Commercial Insurance Data in the United States.

Abstract

Purpose: We evaluated the performance of computational phenotypes (CPs) in commercial insurance claims for identifying transgender (TG) individuals and assessed trends in population size and depression and anxiety prevalence of the TG population. Methods: We compared two previously defined CPs by measuring their concordance. We combined CPs to establish a cohort of TG individuals from Merative^TM MarketScan® commercial insurance claims (2007-2021) to measure population and mental health trends using joinpoint regression. Results: Due to high levels of overlap between CPs, we combined CPs to reach our sample size of 67,809 unique individuals. TG-related International Classification of Diseases (ICD) diagnoses codes increased from 59% of TG claims in 2007 to 97% in 2021. We observed a sharp increase in the prevalence of TG-related claims in 2012 by 42.3% (95% confidence interval [CI] = 35.8-56.8) per year then by 17.0% per year (95% CI = 6.1-23.7) from 2017 to 2021. Among TG individuals there was a gradual increase in mental health-related claims from 2007 to 2015, which remained stable until there was a 10% decrease in 2021. Conclusion: The combined CP identified the largest TG population in commercial insurance claims to date. Most TG individuals were identified through TG-related ICD codes for both CPs. Increases over calendar time may represent an increased access to insurance-covered gender-affirming services. Persistently high depression and anxiety-related claims suggest an ongoing need to reduce the burden of psychiatric-related claims in this population.