A placebo-controlled, double-blind study of fluoxetine in severe alcohol dependence: adjunctive pharmacotherapy during and after inpatient treatment.

Twenty-eight male patients with severe alcohol dependence (mean pretreatment consumption of 18.6 standard drinks per day) completed a placebo-controlled, double-blind clinical trial of fluoxetine (60 mg/day). They were assigned to medication group in the second of 4 weeks on a voluntary inpatient chemical dependency ward and continued medication during a 12-week follow-up phase. Fluoxetine did not reduce clinically significant relapse rates; only 8 of 15 (53%) of fluoxetine subjects remained sober at 12 weeks, compared with 9 of 13 (69%) of the placebo group (Fisher's exact test, p = 0.46). Subjects with comorbid cocaine dependence relapsed more than twice as often (3 of 4, 75%) as those with alcohol dependence alone (8 of 24, 33%), although this trend did not reach statistical significance because of the small number of dually dependent subjects (Mann Whitney U test = 68, p = 0.13). Supportive living arrangements after hospital discharge did reduce relapse rates: 8 of 9 subjects (89%) discharged to a Veterans Affairs domiciliary were sober at 12 weeks, compared with 9 of 19 (47%) subjects discharged back to the community (Mann-Whitney U test = 125, p = 0.02). Fluoxetine-treated subjects who remained sober at 12 weeks reported a significant decrease in mean subjective alcohol craving scores from 2.9 to 0.7 on a 10-point scale (t = 2.828, p = 0.02). In summary, fluoxetine did not reduce clinical relapse rates in this sample of male severe alcoholics without other axis I disorders who completed 4 weeks of inpatient alcoholism treatment.