In the evaluation, there were 130 visits with 92 unique veterans meeting inclusion criteria. The final sample included 70 visits with 47 unique veterans (Table 1). Of note, 24 (53%) patients self-identified as homeless or were engaged with VA housing services. Twelve veterans had multiple ED visits (7 patients with 2 visits; 5 patients with ≥ 3 visits). In 30 (43%) visits the veteran’s primary reason for seeking ED care was to obtain treatment for opioid withdrawal or receive MOUD. Type of opiate used was specified in 58% of visits; of these, 69% indicated heroin use and 17% prescription medications. Buprenorphine was initiated in the ED in 18 (26%) visits for 10 veterans. Appendix 4 outlines the clinical course and follow-up after these visits. Some veterans returned to the ED for buprenorphine redosing per the 72-hour rule. SUD clinic appointments were provided in 11 visits, and direct transfer to an inpatient rehabilitation center was arranged in 4 visits. In 42 (60%) visits, across 32 unique veterans, buprenorphine was not given in the ED, but patients were referred for SUD treatment (Table 2). In 10 (14%) visits, patients were not referred for SUD treatment or given buprenorphine, primarily because the presenting reason was not definitively related to OUDs.
A majority of veterans who received buprenorphine and a referral for an SUD appointment went to their initial SUD follow-up appointment and had ongoing engagement in addiction care 30 days after their index ED visit. Among veterans who did not receive buprenorphine but were referred for SUD treatment, about half went to their SUD appointments and about 1 in 5 had ongoing engagement in addiction care at 30 days after the index ED visit. Of note, 2 patients who received referrals died within 1 year of their index ED visit. The cause of death for one patient was an overdose; the other was unspecified.
DISCUSSION
We developed the ED MOUD program as a bridge to SUD specialty care. Our 8 implementation steps can serve as a model for implementing programs at other VA EDs. We demonstrated feasibility, high follow-up rates, and high retention in treatment.
Patients who received ED buprenorphine initiation were more likely to follow up and had higher rates of ongoing engagement at 30 days than did those who received only a clinic referral. In a similar Canadian study, buprenorphine was initiated in the ED, and patients followed up as a walk-in for addiction services; however, only 54% of patients presented to this initial follow-up.11 Our higher initial follow-up rate may be due to our ability to directly schedule clinic appointments. Our 70% 30-day follow-up rate is comparable, but slightly lower than the 2015 D’Onofrio and colleagues study in which 78% of patients remained engaged at 30 days.7 A possible reason is that in the D’Onofrio and colleagues study, all study physicians obtained X-waiver training and were able to prescribe buprenorphine after ED initiation or for self-initiation at home. X-waiver training was not required of our clinicians, and none of our patients were offered a prescription for self-initiation.
Our program demonstrates that it is feasible to develop a protocol without X-waiver licensing. This program provides a supportive framework for the use of MOUD and allows nonspecialists to gain experience and confidence in using buprenorphine. Any clinician could administer buprenorphine in the ED, and patients could be bridged at later ED visits until follow-up with a specialist. Of note, only a small percentage of the total visits for buprenorphine initiation required multiple daily visits for buprenorphine. Appointments with the specialist were assured to fall within a 72-hour window.
Our program has some limitations. First, the number of patients who were candidates for our ED MOUD program was small. In our 7-month review, only 47 patients were identified as potential candidates for MOUD treatment across 70 visits, and only 10 were initiated in the ED. Second, all patients were not actively screened for OUD. There was potential for missing eligible veterans as inclusion criteria relied on clinicians both recognizing OUD and manually entering a correct diagnostic code. We attempted to mitigate this by also reviewing all ED referrals to the SUD clinic and all patients who received buprenorphine in the ED. In addition, we do not have data on preimplementation rates of follow-up for comparison.