Predictors of compliance with naltrexone among alcoholics

Objective: Naltrexone has been found to be an effective adjunct to treatment to reduce the rate of drinking among alcoholics. However, adherence to the medication has been of considerable concern; the high rates of noncompliance with the medication limits the benefits that could potentially be realized from this pharmacotherapy. Knowledge of predictors of noncompliance could result in interventions targeted at these variables. Method: Participants were 128 alcohol-dependent patients who participated in a clinical placebo-controlled trial of naltrexone. Upon discharge from a 1- to 2-week partial hospital program, patients were randomly placed into 12 weeks of naltrexone (50 mg/day) or placebo (n = 64 per condition). Patients met with a physician and a research assistant weekly for 4 weeks then biweekly for 8 weeks. Results: Compliance (number of days taking medication) was not predicted by demographic or pretreatment alcohol use variables. Number and severity of side effects in the first week, particularly nausea and fatigue, predicted early termination. Compliance was not predicted by commitment to abstinence or self-efficacy about abstinence, but was greater among patients who believed more strongly that the medication would help them stay sober. Compliance was not predicted by general level of urge to drink during the first week on medication but compliance was greater among those with a higher urge to drink in response to alcohol stimuli in the laboratory. Conclusions: Implications for approaches to increase compliance include reducing side effects and increasing patients' beliefs in the efficacy of naltrexone.