Abstract
BACKGROUND: People with opioid use disorder (OUD) often exhibit psychiatric symptoms representing primary (PMD) and substance induced mental disorders (SIMD). PMD are associated with increased discontinuation of medications for OUD (MOUD). It is unknown if this relationship holds for SIMD. OBJECTIVE: To examine the association between SIMD and time to MOUD discontinuation. METHODS: In this retrospective cohort study using the Merative™ MarketScan® Commercial and Multi-State Medicaid insurance claims Databases (2015-2019), we identified individuals initiating MOUD (buprenorphine, methadone, and extended-release naltrexone) following a 180-day pre-period of continuous enrollment without MOUD. Episodes were classified based on > 1 SIMD diagnosis in this pre-period. We estimated days to MOUD discontinuation using stepwise multivariable Cox regression, adjusting for age, sex, insurance type (Medicaid or Commercial), Charlson Comorbidity Index, MOUD at initiation, co-occurring substance use disorders (SUD), and PMD. RESULTS: Of 31,134 MOUD initiation episodes, 3.5 % exhibited SIMD. Episodes with SIMD demonstrated a higher prevalence (p < .001 for all) of mood and psychotic disorders (42.8 vs 16.4 %; 4.0 % vs 0.7 % respectively) but lower prevalence of anxiety-related disorders (51.3 % vs 59.9 %) relative to episodes without SIMD. Unadjusted models illustrated a 44 % (95 %CI:1.34-1.55) increased hazard of MOUD discontinuation associated with SIMD. After adjusting for all covariates but PMD, SIMD retained a 13 % increased hazard of MOUD discontinuation (95 %CI:1.05-1.21). Adjusting for PMD eliminated this association. CONCLUSION: SIMD is associated with reduced time to MOUD discontinuation, which may be accounted for by PMD. Further research on SIMD treatment among individuals with OUD is needed to improve OUD treatment outcomes.