Predictors of clinical trial dropout in individuals with co-occurring bipolar disorder and alcohol dependence

Abstract

Background: Individuals with co-occurring bipolar disorder and alcohol dependence have particularly low rates of retention in clinical trials. Past research has identified a variety of factors associated with dropout in this population, but few have been replicated. The present study investigated the ability of several baseline variables to predict clinical trial dropout in a sample of individuals with co-morbid bipolar and alcohol use disorders.

Methods: Demographics, psychiatric diagnoses, recent alcohol use, mood pathology, and risk taking behavior (measured with the Balloon Analogue Risk Task) were evaluated as predictors of dropout from a randomized clinical trial of acamprosate for individuals with co-morbid bipolar and alcohol use disorders (n=30) using stepwise logistic regression.

Results: Risk taking behavior was the only significant predictor of dropout in the present study (OR=1.44, p=0.03); opiate dependence marginally predicted dropout as well (OR=13.46, p=0.08). A model consisting of these predictors, as well as acamprosate group status (p=0.13), provided excellent prediction of dropout (i.e., area under the ROC curve=0.94; R(2)=0.53).

Conclusions: Given the robust relationship between risk taking and dropout in the present study, the Balloon Analogue Risk Task may represent a valuable tool for researchers to predict who will drop out of clinical trials for comorbid bipolar and substance use disorders.