Serious non-fatal and fatal opioid overdose events in a cohort of hospital patients seen by an addiction consultation service: Secondary analysis of the NavSTAR trial.

Aims: North America is experiencing a crisis of opioid-associated mortality. The current study examined opioid overdose events in a cohort of hospital patients enrolled in the Navigation Services to Avoid Rehospitalization (NavSTAR) randomized trial, which found that augmenting hospital-based addiction consultation services with post-discharge patient navigation reduced hospital readmissions compared to usual care.

Methods: Participants were adult medical/surgical hospital patients with comorbid opioid, cocaine, and/or alcohol use disorder (N=400; 43% female). All participants were seen by the hospital addiction consultation service, and randomized 1:1 to Patient Navigation or Usual Care. Non-fatal opioid overdoses requiring acute hospital care were tracked for 12 months through regional health information exchange. Fatal opioid overdoses were tracked from state death records, covering 3.3-5.5 years post-enrollment. Non-fatal and fatal overdoses were examined using logistic and Cox regression, with predictors of sex, race, age, study arm, and (in the death analysis) serious non-fatal overdose.

Results: In the first year post-enrollment, there were 122 non-fatal opioid overdose events requiring hospital care (experienced by 63/400 of participants), and 17 fatal opioid overdose. Over the longer-term follow-up (3.3-5.5 years), there were 52 fatal opioid overdoses (13% of the sample). Participants in the Patient Navigation arm were less likely than Usual Care to experience non-fatal opioid overdose requiring acute care (OR=0.50 [0.29, 0.88]; P=0.017], but did not differ on fatal overdose. Likewise, neither fatal nor non-fatal overdose differed by participant race, sex, or age (Ps> 0.10). The association between non-fatal opioid overdose and subsequent fatal overdose was non-significant (HR=1.75 [0.92, 3.31; P= 0.087).

Conclusions: Opioid overdose events were common in this sample of hospital patients with comorbid SUD, highlighting a need for harm reduction and treatment linkage services. The Patient Navigation intervention reduced the likelihood of serious non-fatal opioid overdose, but did not impact fatal opioid overdoses.