Outcomes of buprenorphine maintenance in office-based practice

Buprenorphine is an efficacious treatment for opioid dependence recently approved for office-based medical practice. The purpose of the study was to describe the background characteristics, treatment process, outcomes and correlates of outcomes for patients receiving buprenorphine maintenance in "real world" office-based settings in New York City, without employing the many patient exclusion criteria characterizing clinical research studies of buprenorphine, including absence of co-occurring psychiatric and non-opioid substance use disorders. A convenience sample of six physicians completed anonymous chart abstraction forms for all patients who began buprenorphine induction or who transferred to these practices during 2003-2005 (N = 86). The endpoint was the patient's current status or status at discharge from the index practice, presented in an intent-to-treat analysis. The results were: male (74%); median age (38 yrs); White, non-Hispanic (82%); employed full-time, (58%); HCV+ (15%); substance use at intake: prescription opioids (50%), heroin (35%), non-opioids (49%); median length of treatment (8 months); median maintenance dose (15 mg/day); prescribed psychiatric medication (63%). The most frequent psychiatric disorders were: major depression, obsessive-compulsive and other anxiety, bipolar. At the endpoint: retained in the index practice (55%); transferred to other buprenorphine practice (6%); transferred to other treatment (7%); lost to contact or out of any treatment (32%). Outcomes were positive, in that 2/3 of patients remained in the index practice or transferred to other treatment. Patients living in their own home or misusing prescription opioids (rather than heroin) were more likely, and those employed part-time were less likely, to be retained in the index practice. At the endpoint, 24% of patients were misusing drugs or alcohol. Co-occurring psychiatric disorders and polysubstance abuse at intake were common, but received clinical attention, which may explain why their effect on outcomes was minimal.