Models of Screening, Brief Intervention With a Facilitated Referral to Treatment (SBIRT) for Opioid Patients in the Emergency Department

Patients with heroin and prescription opioid dependence are at increased risk for adverse health consequences and often utilize the Emergency Department (ED) as their source of medical care. Screening, brief intervention and referral to treatment has been effective in decreasing high risk behaviors such as alcohol and tobacco use, and unsafe sexual practices. The data on the effectiveness of brief interventions with opioid dependence is limited. This prospective, randomized controlled trial of opioid dependent subjects (N=360) will compare two models of brief intervention with a control condition. ED patients with opioid dependence will be randomized to either: (1)Screening, Brief Intervention with a Facilitated Referral to Treatment (SBIRT); (2) Screening, Brief Intervention with ED initiated Buprenorphine Treatment (SBI+Bup); or (3) standard care (SC) which includes a handout detailing substance abuse treatment centers in the area. The primary outcome will be self-reported engagement in formal substance abuse treatment at 30 days, verified by contact with the treatment program. Other outcomes measured at 30 days, 2, 6 and 12 months include changes in opioid use (self-report and urine toxicology analysis), HIV risk behaviors, and health care service utilization. The three interventions will also be compared on their cost-effectiveness. We will test the hypotheses that SBI+Bup will be superior to SBIRT and SC, and SBIRT will be superior to SC in (1)increasing the proportion of patients engaged in formal substance abuse treatment at 30 days; (2) reducing illicit opioid use; (3) reducing HIV risk behaviors; and (4) reducing health care service utilization. In addition, we hypothesize that the societal costs of SBI+Bup, per number of days of opioid abstinence, will be cost effective relative to SBIRT or SC; and that SBIRT will be cost effective relative to SC. Data analyses will be conducted on the intention to treat sample of randomized patients. This study, conducted