Fuller 1986: the VA disulfiram trial in 605 men

What they tested

Men were randomly assigned to 250 mg of disulfiram (202 men), 1 mg of disulfiram (204 men) or no disulfiram (199 men). The 1 mg dose was too small to cause a reaction with alcohol, so it tested whether the threat of a reaction alone made a difference. Everyone got counselling.

The men were assessed regularly for a year. Interviews with relatives or friends and blood and urine alcohol tests were used to check what they reported.

What they found

There were no significant differences among the groups in total abstinence, time to first drink, employment or social stability.

Among patients who drank and had a complete set of assessment interviews, the 250 mg group reported significantly fewer drinking days (49.0) than the 1 mg group (75.4) or the no-disulfiram group (86.5).

In every group, taking the assigned pills was linked to complete abstinence. The authors concluded disulfiram may help reduce how often people drink after a relapse, but does not improve on counselling for staying abstinent or delaying the return to drinking.

Limits

• Everyone was a man in a Veterans Administration program, so it may not apply to women or other settings.
• The fewer-drinking-days result comes only from men who drank and completed every interview, a subgroup rather than everyone randomized.
• Pill-taking was linked to abstinence in every group, so that link can't be credited to disulfiram itself.

What it means if you're considering naltrexone

For someone comparing disulfiram with naltrexone, this large trial found that disulfiram's main feature, the threat of feeling sick after drinking, did not lead to more men staying sober. Its benefit here was narrower: fewer drinking days among men who had started drinking again.