Some of the heterogeneity of clinical findings in studies evaluating the efficacy of pharmacotherapies and behavioral therapies in the treatment of alcohol dependence can be attributed to the wide use of standard statistical analytical tools of summary drinking measures that poorly reflect the distributions and complexity of drinking data. Novel statistical analysis tools based on trajectories over time provide a more realistic and complete picture of treatment effects on drinking behavior, and of the interplay of treatment compliance and treatment response. The COMBINE Study is the largest randomized clinical trial to date designed to answer questions about the benefits of combining behavioral and pharmacological interventions. It was successful in confirming some hypotheses but provided discouraging results on other primary efficacy hypotheses. Further understanding of both the positive and the negative results of this study, identifying subgroups of subjects for whom treatment is effective and assessing the effects of compliance on response are vital to moving the field of alcohol research forward towards improved therapies and individual-based treatments. Herein we propose to assess patterns of change over time in drinking and compliance in the COMBINE Study using the trajectory-based approach of Nagin (1999). The study enrolled 1,383 abstinent alcohol dependent patients and tested the efficacy of naltrexone, acamprosate, and Combined Behavioral Intervention (CBI). Analyses of the two primary endpoints, time to the first day of heavy drinking and percent days abstinent, revealed that either naltrexone or CBI improved outcomes compared to MM + placebo but surprisingly, the combination of naltrexone + CBI did not yield better outcomes than either treatment alone, and acamprosate was not efficacious either alone or in combination with naltrexone or CBI. The trajectorybased approach will allow characterization of the heterogeneity in participants'drinking behavior and treatment compliance, will shed light on some of the negative findings regarding acamprosate and the combination of naltrexone and CBI, and will allow assessment of the interplay between compliance, treatment and drinking behavior.
Our specific aims are 1) to identify different trajectory classes of drinking over time and to assess medication and CBI main effects and interactions on the probability to follow particular trajectories, 2) to examine trajectories of drinking prior to randomization as potential moderators of naltrexone, acamprosate and CBI effects, 3) to examine trajectories of treatment participation (e.g., MM, CBI, and medication adherence) and their effects on treatment response, and 4) to explore trajectories of drinking following discontinuation of treatment. The results of these analyses may offer new theoretical and clinical insights into the use of medications and behavioral interventions for the treatment of alcohol addiction.
