Cross sectional analysis of an addiction consultation service, substance co-use patterns, and receipt of medications for opioid use disorder during hospitalization

Riley D Shearer,Gavin Bart,Timothy J Beebe,Beth A Virnig,Nathan D Shippee,Tyler N A Winkelman
DOI: https://doi.org/10.1016/j.josat.2024.209505
Abstract:Introduction: Despite effective medications for opioid use disorder (MOUD), treatment engagement remains low. As the overdose crisis is increasingly characterized by opioids co-used with other substances, it is important to understand whether existing models effectively support treatment for patients who use multiple substances. Hospital-based addiction consultation services (ACS) have shown promise at increasing MOUD initiation and treatment engagement, but the effectiveness for patients with specific co-use patterns remains unknown. Methods: Using 2016-2023 admissions data from a large safety net hospital, we estimated a random-effects logistic regression model to determine whether specific co-use (methamphetamine, cocaine, alcohol, sedative, and other) moderated the effect of being seen by ACS on the receipt of MOUD. Adjusting for patient sociodemographic, health, and admission characteristics we estimated the proportion of patients who received MOUD across specific co-use groups. Results: Of 7679 total admissions indicating opioid use, of which 5266 (68.6 %) indicated co-use of one or more substances and 2387 (31.1 %) were seen by the ACS. Among admissions not seen by the ACS, a smaller proportion of admissions with any co-use received MOUD (23.5 %; 95 % CI: 21.9-25.1) compared to admissions with opioid use alone (34.0 %; 95 % CI: 31.9-36.1). However, among admissions seen by the ACS a similar proportion of admissions with any co-use received MOUD (57.8 %; 95 % CI: 55.5-60.1) as admissions with opioid use alone (56.2 %; 95 % CI: 52.2-60.2). The increase in proportion of admissions receiving MOUD associated with being seen by the ACS was larger for admissions with methamphetamine (38.6 percentage points; 95 % CI: 34.6-42.6) or cannabis co-use (39.0 percentage points; 95 % CI: 32.9-45.1) compared to admissions without methamphetamine (25.7 percentage points; 95 % CI: 22.2-29.2) or cannabis co-use (29.1 percentage points; 95 % CI: 26.1-32.1). Conclusions: The ACS is an effective hospital-based treatment model for increasing the proportion of admissions which receive MOUD. This study shows that ACSs are also able to support increased receipt of MOUD for patients who use other substances in addition to opioids. Future research is needed to further understand what transition strategies best support treatment linkage for patients who use multiple substances.
What problem does this paper attempt to address?