Combining Medications and Behavioral Interventions
(Project COMBINE) Miller, PI
(Site-specific abstract). The University of New Mexico's Center on Alcoholism, Substance Abuse, and Addictions (CASAA) offers to participate as a Clinical Research Unit (CRU) in the multisite trial, Combined Behavioral/Pharmacological Treatment of Alcoholism. Although the trial protocol will ultimately be determined by the trial's Steering Committee, this proposal reflects how CASAA might approach the project. Our factorial design for the main (Phase M) trial crosses four levels of medication (placebo, acamprosate, naltrexone, and combined medications) with two levels of behavioral intervention (Cognitive-Behavioral Therapy and Motivational Enhancement Therapy), both modified from manuals and experience in Project MATCH. Following the initial treatment period (months 1-3), clients are assigned at random to receive or not receive augmented treatment during months 4-6. This embedded trial will provide reliable information about main effects of longer (augmented) versus shorter treatment, and about client characteristics associated with differential benefit from augmented treatment. Augmentation will be retained within each behavioral treatment modality, so as not to confound the interpretableness of follow-up data. The prospective validity of various 'nonresponder' classification systems can be tested via retrospective analyses within this design. In addition to feasibility testing of procedures, the preliminary (Phase II) study will evaluate the main effect of extensive pretreatment and follow-up assessment on treatment outcomes, a crucial issue because such reactivity could wash out main or interaction effects of trial treatments - a critique unaddressed in Project MATCH. Clients in the Phase II study will all be given the both acamprosate and naltrexone, and further randomized within a 2x2 factorial design to receive either extensive or minimal assessment, and, either CBT or MET. The Phase II study design also allows evaluation of whether a client's knowledge that drinking reports will be verified via collateral interviews affects the accuracy of self-reports. Main trial clients will be randomly assigned to therapists within conditions, so that (unlike in Project MATCH), true tests of client/therapist matches can be conducted in this large sample. Prior to randomization, all clients will be challenged with an initial medication dose (which exerts immediate aversive effects in a minority of individuals), to diminish potentially problematic differential attrition from medication and placebo groups. Naturalistic involvement in Alcoholics Anonymous and other mutual help organizations will be studied to determine whether it interacts with these treatments to enhance outcomes.
