Patterns in (es)citalopram prescriptions to Medicaid and Medicare patients in the United States: the potential effects of evergreening

Luke Cavanah ¹ Parita Ray ^1* Brian J Piper ^1,2 Jessica Goldhirsh ¹ Leighton Y Huey ^1,3

• ¹ Geisinger Commonwealth School of Medicine, Scranton, United States
• ² Center for Pharmacy Innovation and Outcomes, Forty Fort, United States
• ³ Behavioral Health Initiative, Scranton, United States

Introduction: Citalopram and escitalopram are among the most used medications and are key treatments for many psychiatric disorders. Previous findings suggest citalopram and escitalopram prescription rates are changing because of the patent for citalopram ending as opposed to evidence of a clear therapeutic advantage-so-called "evergreening". This retrospective study focuses on characterizing the chronologic and geographic variation in the use of citalopram and escitalopram from 2015 to 2020 among US Medicaid and Medicare patients. We hypothesized that prescription rates of citalopram will decrease with a concurrent increase in escitalopram, consistent with "evergreening". Methods: Citalopram and escitalopram prescription rates and costs per state were obtained from the Medicaid State Drug Utilization Database and Medicare Provider Utilization and Payment Data. States' annual prescription rates outside a 95% confidence interval were considered significantly different from the average. Results: Overall, a decreasing trend for citalopram and an increasing trend for escitalopram prescription rates were noted in both Medicare and Medicaid patients. The differences between generic and brand were noted for both drugs, with generic forms being less expensive than the brand-name version. Discussion: Despite limited evidence suggesting that citalopram and escitalopram have any meaningful differences in therapeutic or adverse effects, there exists a noticeable decline in the use of citalopram that cooccurred with an increase in escitalopram prescribing, consistent with our hypothesis. Moreover, among these general pharmacoepidemiologic trends exists significant geographic variability. There was disproportionate spending (relative to their use) on the brand versions of these medicines relative to their generic forms.