Background: Persons with opioid use disorders (OUD) represent a disproportionately high percentage of hospitalized patients, have greater lengths of stay and readmissions, can incur higher costs, and are at greater risk of drug-related death immediately post-discharge. Further, despite the availability of effective medications for OUD, patients often are not offered treatment during the inpatient encounter. While addiction medicine consult services have been shown to decrease ongoing substance use, decrease readmissions, and increase engagement with treatment, most hospitals do not have such services. The Journal of Hospital Medicine recently published an editorial calling for hospitalists to ensure that hospitalized patients receive treatment for substance use disorders, particularly at hospitals without consult services. However, the ability of a hospitalist-led team to initiate and retain patients with OUD in treatment, and whether this might affect hospital metrics, is unknown.

Methods: Mount Sinai Hospital is a large, urban, tertiary-care hospital without a formal addiction medicine consult service. REACH-IN is a quality-improvement program co-led by a full-time hospitalist and medical trainees that identifies patients with OUD, assists in the management of withdrawal and with the initiation of buprenorphine, and facilitates the linkage to ongoing treatment post-discharge. Patients are identified via 1) a daily electronic report of inpatient encounters that document opioid use, or 2) a direct referral from providers. Eligible patients are evaluated, started on buprenorphine if appropriate, and offered a facilitated referral to either REACH, an affiliated primary care program, or another community-based provider for continued treatment.

Results: Between April 2018 and October 2019, REACH-IN evaluated 104 inpatients, and initiated buprenorphine for 25 (24%) of them. Of those patients started on buprenorphine during their hospitalization, 17 (68%), 13 (53%), and 11 (44%) were engaged in care – as defined by having an active prescription for buprenorphine – at 30, 60, and 90 days post-discharge, respectively. A sub-analysis compared 48 evaluated patients and 13 patients started on buprenorphine while hospitalized with controls matched to similar patients from 2016-2018 based on diagnosis, complexity, age, treatment status, service, and discharge disposition. It found a non-significant trend toward a reduction in readmissions. Compared to 19.1% for matched controls, 30-day readmission rates were 12.2% for all evaluated patients (p = 0.40) and 7.7% for patients started on buprenorphine (p = 0.34). The ratio of observed to expected 30-day readmissions decreased to 0.78 from 1.22 for the matched controls.

Conclusions: Interaction with REACH-IN and inpatient initiation of buprenorphine were associated with high rates of engagement with treatment post-discharge and a trend toward a reduction in 30-day readmissions. Although this trend was non-significant, it may be more adequately powered to detect a difference with a larger sample size. We describe a feasible innovation that successfully identified patients with OUD, initiated them on and retained them in effective treatment, and observed a non-significant trend towards decreased 30-day readmissions. This initiative could be replicated at other institutions grappling with how best to serve growing numbers of hospitalized patients with OUD.