Abstract
BACKGROUND: Few studies have reported on the prevalence and health risks associated with substance use and substance use disorder (SU/SUD) in Veterans who use long-term opioid therapy (LTOT). We leveraged health record data to estimate SU/SUD prevalence and its association with mortality among Veterans on LTOT. METHODS: We conducted a secondary analysis of cohort data for Veterans on LTOT within Veterans Health Administration outpatient settings (2014-2019). SU/SUD was defined as a positive screen for risky alcohol use; a positive urine drug screen for cannabis, benzodiazepines, or stimulants; or a documented SUD diagnosis. We fit Cox models for all-cause mortality, fatal overdose, and traumatic deaths, comparing Veterans on LTOT with SU/SUD vs. LTOT-only. RESULTS: One in four (25.0 %) Veterans on LTOT have risky alcohol use, tested positive for other substances, or had a diagnosed SUD. Alcohol was the most common SU/SUD (9.8 %), followed by sedative (8.1 %), cannabis (6.6 %), and stimulant (0.6 %). Relative to Veterans on LTOT only, mortality rates were higher for Veterans on LTOT with cannabis (HR=1.16, 95 % CI=1.03, 1.30), sedative (HR=1.29, 95 % CI=1.10, 1.52), or stimulant SU/SUD (HR=1.54, 95 % CI=1.17, 2.02). Fatal overdose rates were higher for LTOT with alcohol (HR=1.43, 95 % CI=1.10, 1.86), sedatives (HR=1.40, 95 % CI=1.04, 1.91), or stimulant SU/SUD (HR=3.29, 95 % CI=1.60, 6.77). LTOT with sedative SU/SUD was associated with traumatic death rates (HR=1.30, 95 % CI=1.05, 1.61). CONCLUSION: Substance co-use is common among Veterans on LTOT and is associated with elevated mortality and overdose risks. Comprehensive screening and targeted interventions may be needed.