Addiction Consult Service Associated With Improved Rates of 30-day Treatment Engagement After Emergency Department and Hospital Discharge: A Propensity-weighted Analysis.

Anderson, Erik S; Sirey, Lauren; Jakubowski, Andrea; Heeney, Megan; Patregnani, Michelle; Ullal, Monish; Herring, Andrew A · J Addict Med · 2026

retrospective_cohort · Level III

Where this comes from

Abstract

Addiction consult services (ACS) are an effective intervention for hospitalized patients with substance use disorders (SUD); however, their effectiveness for patients discharged from the emergency department (ED) is not well established. We aim to examine the association of ACS with 30-day engagement in outpatient treatment among ED and hospitalized patients at a multi-hospital safety-net health system. We conducted a retrospective study of patients discharged after ED and inpatient visits with SUD-related diagnoses. The exposure of interest was receipt of ACS consultation during an ED or inpatient encounter. We used propensity score matching with frequency weights for the analyses, and the primary outcome was engagement in outpatient SUD treatment within 30 days of ED or hospital discharge. Propensity-weighted logistic regression with and without a moderation analysis examined factors associated with engagement in treatment after discharge. From April 1, 2024, through February 28, 2025, 4779 unique patients were discharged with a qualifying SUD diagnosis; 2783 (58.2%) after ED discharge and 1996 (41.8%) after hospital discharge, and 1279 (26.8%) received a Bridge ACS consultation. The 30-day engagement in treatment after discharge rate was 32.6% with ACS versus 9.4% without (P<0.001). ACS was independently associated with higher odds of engagement [adjusted odds ratio (aOR): 4.7, 95% CI: 3.5-6.4]. The interaction term between ACS and ED versus hospital discharge was not statistically significant (aOR: 0.9, 95% CI: 0.5-1.5). ACS consultation was associated with increased engagement in SUD treatment within 30 days of ED or hospital discharge.