Context Alcoholism affects approximately 10% of Americans at some time in their lives. Treatment consists of psychosocial interventions, pharmacological interventions, or both, but which drugs are most effective at enhancing abstinence and preventing relapse has not been systematically reviewed. Objective To evaluate the efficacy of 5 categories of drugs used to treat alcohol dependence-disulfiram; the opioid antagonists naltrexone and nalmefene, acamprosate, various serotonergic agents (including selective serotonergic reuptake inhibitors), and lithium. Data Sources Reports of randomized controlled trials, nonrandomized trials, and other study designs in English, French, and German identified from multiple searches of MEDLINE, EMBASE, and specialized databases; hand searching bibliographies of review articles; searches for unpublished literature; and discussions with investigators in the field. Study Selection We included all studies on alcohol-dependent human subjects aged 18 years or older from all inpatient and outpatient settings between 1966 and December 1997 that met our inclusion criteria. Data Extraction We abstracted the following information: study design and binding, diagnostic instrument and severity assessment, drug interventions and cointerventions, demographic and comorbidity details about patients, compliance, and numerous outcome measures (eg, relapse, return to drinking, drinking or nondrinking days, time to first drink, alcohol consumed per unit of time, craving). We graded quality of the individual articles (scale, 0-100) independently from the strength of evidence for each drug class (A, strong and consistent evidence of efficacy in studies of large size and/or high quality; B, mixed evidence of efficacy; C, evidence of lack of efficacy; and I, insufficient evidence). Data Synthesis Of 375 articles evaluated, we abstracted and analyzed data from 41 studies and 11 follow-up or subgroup studies. Naltrexone (grade A) reduces the risk of relapse to heavy drinking and the frequency of drinking compared with placebo but does not substantially enhance abstinence, ie, avoidance of any alcohol consumption. Acamprosate (grade A, from large-scale studies in Europe) reduces drinking frequency,al though its effects on enhancing abstinence or reducing time to first drink are less clear. Controlled studies of disulfiram (grade B) reveal a mixed outcome pattern--some evidence that drinking frequency is reduced but minimal evidence to support improved continuous abstinence rates. The limited data on serotonergic agents were not very promising (grade I), although most studies were confounded by high rates of comorbid mood disorders. Lithium lacks efficacy (grade C) in the treatment of primary alcohol dependence. Conclusions Recent reports documenting that naltrexone and acamprosate are more effective than placebo in the treatment of alcoholism justify clinical interest in use of these medications for alcohol-dependent patients. Use-of disulfiram is widespread but less clearly supported by the clinical trial evidence; however, targeted studies on supervised administration of disulfiram may be warranted. Useof existing serotonergic agents or lithium for patients with primary alcohol dependence does not appear to be supported by the efficacy data available at this time; these medications may still have a positive effect in patients with coexisting psychiatric disorders.
