(SLB-027) Association between traumatic brain injury and post-hospitalization opioid dose among prevalent opioid users

Background: The risk of substance use disorder is elevated among individuals with a traumatic brain injury (TBI). However, it remains unclear whether this increased risk is the result of an increased likelihood of developing chronic pain or potential mechanistic changes to neural networks linked to increased risk of addiction and opioid use disorder (OUD). Additionally, whether TBI uniquely impacts subsequent opioid use among prevalent opioid users with traumatic injuries is not well understood.

Objectives: To test the hypothesis that among prevalent opioid users, the change in total opioid dose after hospitalization compared to pre-hospitalization will be different among those hospitalized with TBI compared to those hospitalized with long-bone fracture (LBF) only.

Methods: We used insurance claims data from Merative MarketScan (2016-2022) to identify a retrospective cohort of adult patients without OUD but with evidence of a filled opioid prescription in the 60 days before hospitalization for TBI or LBF. The outcome of interest was the difference in total milligram morphine equivalent (MME) dose of all opioid prescriptions filled after the initial post-discharge period (day 43-103) compared to the 60-day period prior to hospitalization. We excluded prescriptions filled in the 42-day post-discharge window after hospitalization as this represents the standard duration for post-discharge pain management after traumatic injury. We used linear regression to compare the difference in post-hospitalization MME dose adjusting for baseline covariates identified in the 180 days before hospitalization through day 42 after hospitalization, including initial post-discharge opioid dose, concomitant use of non-opioid analgesics and mental health medications, indicators of pain, depression, length of stay, and evidence of surgery during hospitalization.

Results: We identified 7,062 prevalent opioid users without OUD prior to hospitalization for traumatic injury (TBI n=1,020; LBF n=6,042). Individuals hospitalized with TBI had similar pre-hospitalization opioid use (median MME=445) compared to those with LBF (median MME=400), though the unadjusted difference in MME after hospitalization was similar among those with TBI (median total MME decrease: -150) compared to LBF (median total MME decrease: -150). In the adjusted linear regression, we observed a non-significant, smaller reduction in total post-hospitalization MME among those with TBI compared to those with LBF (coefficient: -82.3; 95% CI: -211.6-47.0).

Conclusions: Post-hospitalization opioid use was lower among prevalent opioid users hospitalized for TBI and LBF, though the decrease was smaller among those with TBI compared to LBF.