Walker 2019: how well veterans stuck with their alcohol medicines

What they tested

Researchers reviewed the records of veterans prescribed FDA-approved alcohol use disorder medicines at Veterans Affairs hospitals and community clinics in South Carolina and Georgia from 2010 through 2015. They included 715 people and 807 courses of medicine.

Adherence was measured from pharmacy fills: the share of days over a 6-month period when the person had the pills or injection available, from 0% to 100%.

What they found

Mean adherence was 41.3% for disulfiram, 44.7% for acamprosate, 49.8% for oral naltrexone and 54.6% for extended-release injectable naltrexone. The gap was significant between disulfiram and oral naltrexone (p = .002) and between disulfiram and the shot (p = .004).

Few courses reached 80% adherence: 11.9% with disulfiram, 19.4% with acamprosate, 22.7% with naltrexone and 24.4% with the shot. Both naltrexone forms again beat disulfiram significantly (p = .004 for pills, p = .05 for the shot).

The abstract doesn't report a significant difference between oral naltrexone and the shot.

Limits

• Having medicine on hand isn't the same as taking it. Pharmacy fills show what was picked up, not what was swallowed.
• It was a chart review, not a randomized trial, so people on each medicine may have differed in other ways.
• It covered veterans in two states, who may not reflect other patients.

What it means if you're considering naltrexone

Whatever medicine someone chooses for alcohol, sticking with it is a common problem: on average, these veterans had their medicine on fewer than 55% of days. Naltrexone, as pills or the shot, did better than disulfiram, and the pills and shot were close to each other.