CLINICAL UNIT BASED RESEARCH
CPDD 2008 Annual Meeting. San Juan Puerto Rico
790 METHODOLOGICAL ISSUES IN DESIGNING A CLINICAL TRlAL FOR OPIOID ANALGESIC DEPENDENCE:Â DEFINING THE POPULATION
R.D. Weiss l,2, J.S. Potter l,2, K. Prather l, S. Provost l, J. Selze 3 and W. Ling 4
1 Alcohol and Drug Abuse Treatment Program, McLean Hospital, Belmont, and
2 Psychiatry, Harvard Medical School, Boston, MA, 3 Psychiatry, Hillside
Hospital, Glen Oaks, NY and 4 Psychiatry, University of California-Los Angeles
Aims: With the increased prevalence of OAD, studying treatments for this population is critical. This report presents an approach to defining a distinct population of patients with OAD so that we can communicate clearly the population we have chosen to study. Methods: The NIDA Clinical Trials Network is studying combinations of buprenorphine and counseling for OAD. In identifying the study population, physical pain and heroin use were critical issues to address. Chronic pain was defined as "other than everyday kinds of pain" for> 3 months. We excluded subjects with a major pain event in the past 6 months or pain requiring ongoing opioids. Subjects were excluded for lifetime opioid dependence due to heroin use; ever injecting heroin; or using heroin > 4 of the past 30 days. Thus, some patients with pain and/or heroin use could enter the trial. Subjects met DSM-IV opioid dependence criteria and had to have ≥ 20 days of opioid use in the past month and be physically dependent; physical dependence alone was insufficient. We hypothesized that these criteria would yield a representative, distinct population. Results: To date, 2,016 individuals have been phone-screened. Reasons for ineligibility (45%) include recent major pain event (7%), heroin injection (40%), and current heavy heroin use (15%) Of those potentially eligible, 361 have completed an in-person assessment; 274 (76%) have been randomized. Reasons for exclusion include lack of medical clearance (16%), lifetime heroin dependence (II %), heroin injection (7%), current heavy heroin use (13%), < 20 days of opioid use in the past month (18%), and no diagnosis of opioid dependence (27%) 34% of those randomized have chronic pain, and 22% have ever used heroin. Conclusions: It is feasible to identify individuals with OAD for a multisite study and recruit a sufficient number of participants. Support: Supported by NIDA Grants DA15831, DA022288, DA13035, and DAO13045