(PS12-B42) Psychiatric and Racial-ethnic Predictors of Incident and Long-term Benzodiazepine Use: A National Study of Commercial Health Care Claims

Benzodiazepine use is common, with over 30 million people in the United States receiving prescriptions yearly. Although clinical guidelines only recommend use for short intervals, many individuals use benzodiazepines for longer periods. Despite this, there is a lack of large, methodologically sound studies documenting predictors of incident and long-term benzodiazepine use. This case-control and cohort study examined employer-sponsored insurance (ESI) claims among 2,102,788 individuals (ages 13-65 from 2010 to 2019) from Optum’s de-identified Clinformatics® Data Mart Database who filled a new benzodiazepine prescription (i.e., after at least a year without a prescription) to evaluate psychiatric and demographic predictors of a) filling a benzodiazepine prescription and b) transitioning to long-term use, defined as at least six months of continuous use. We compared cases with matched-pair controls in a series of separate conditional logistic regressions, matched 1:1 on sex, calendar year, and age of first enrollment. We then modeled long-term use among cases with Kaplan-Meier estimates and Cox proportional hazards regression. All psychiatric conditions in the year before initiation were associated with an incident benzodiazepine prescription, and most subsequently predicted greater risk of long-term benzodiazepine use among benzodiazepine initiators. The strongest predictors for incident use were anxiety disorders [e.g., specific phobia (Odds Ratio [OR]=9.63; 95% Confidence Interval [CI], 8.91-10.41), panic disorder/agoraphobia (OR=8.74; 95% CI, 8.50-8.99)], and having a sedative-, hypnotic-, or anxiolytic-related substance use disorder (OR=7.37; 95% CI, 6.76-8.04). Individuals from all racial/ethnic minoritized groups were less likely to have incident benzodiazepine use compared with White individuals. By two years after the initial prescription fill, 10.26% of the sample population met criteria for long-term benzodiazepine use. The strongest predictors of long-term use were severe mental illness [e.g., schizophrenia (Hazard Ratio [HR]=3.60; 95% CI, 3.52-3.69), bipolar disorder (HR=3.13; 95%CI, 3.08-3.17)] and sedative-, hypnotic-, and anxiolytic-related substance use disorder (HR=3.62; 95% CI, 3.42-3.83). These findings highlight that many people with ESI transition to long-term benzodiazepine use in the United States, particularly those with severe mental illness, which should inform future studies of benzodiazepine prescription patterns and their consequences.