Predictors of Engagement and Impact on Alcohol-Related Hospital Utilization with Primary Care-Based Alcohol Use...

Primary care-based alcohol use disorder (AUD) treatment models may improve engagement with AUD treatment, yet predictors of engagement and impact of these approaches on hospital utilization are unclear.

To characterize patients referred to a primary care-based AUD clinic (AUDC), identify predictors of engagement, and assess associations between engagement and alcohol-related hospital utilization.

Retrospective cohort study of adults referred to the AUDC at a large public hospital using electronic health record data (2022-2025).

Adults (≥ 18 years) referred to the AUDC, classified as attendees (attended ≥ 1 AUDC visit) or non-attendees.

Differences between attendees and non-attendees in characteristics using t-test/chi square and alcohol-related ED visits or hospitalizations 6, 12, and 18 months after referral using multivariate regression.

Of 463 patients, 253 were attendees (54.6%). Attendees were more often referred by primary care providers (PCPs) than non-PCPs (57% vs. 35%, p < 0.001), had public insurance than private or no insurance (30.4% vs. 18.1%, p = 0.003), had previously seen a primary care provider (PCP) (85.8% vs 62.4%, p < 0.001), and less often had any alcohol-related ED visits or hospitalizations before referral (43.9% vs 60.5%, p < 0.001). AUDC attendance was associated with significantly reduced odds of alcohol-related ED visits or hospitalizations at 6 months (OR 0.59; 95% CI: 0.36-0.94; p = 0.03; N = 461; (1-β) = 0.81) and 12 months (OR 0.63; 95% CI: 0.40-0.98; p = 0.04; N = 425; (1-β) = 0.84) but not 18 months (OR 0.65; 95% CI: 0.37-1.12; p = 0.12; N = 290; (1-β) = 0.72), controlling for proportion of patients with an ED visits or hospitalizations in 6 months prior to referral, referral source, and funding source.

Primary care-based AUDC engagement was associated with reduced alcohol-related hospital utilization short- and intermediate-term but not long-term. Existing relationship with primary care was a notable predictor of AUDC engagement, providing insights into who benefits from this model.