Wakeman 2020: which opioid treatments were linked to fewer overdoses

What they tested

Researchers looked back at insurance claims for people aged 16 or older with opioid use disorder who had commercial or Medicare Advantage coverage. Each person was placed in one of six treatment pathways: no treatment, inpatient detox or residential services, intensive behavioral health care, buprenorphine or methadone, naltrexone, or nonintensive behavioral health care.

The outcomes were an opioid overdose or serious opioid-related acute care in the 3 and 12 months after treatment began, used as stand-ins for a return to opioid use. Nobody was randomly assigned, so the results show links rather than cause and effect.

What they found

Most people did not get medicine. 24,258 (59.3%) received nonintensive behavioral health care, 6,455 (15.8%) went to inpatient detox or residential services, 5,123 (12.5%) got buprenorphine or methadone, 1,970 (4.8%) got intensive behavioral health care and 963 (2.4%) got naltrexone.

Within 3 months, 707 people (1.7%) had an overdose and 773 (1.9%) had serious opioid-related acute care.

Only buprenorphine or methadone was linked to fewer overdoses, with adjusted hazard ratios of 0.24 at 3 months and 0.41 at 12 months. A hazard ratio below 1 means a lower risk. Buprenorphine or methadone was also linked to less serious opioid-related acute care (0.68 at 3 months and 0.74 at 12 months).

Naltrexone and the other pathways were not linked to a lower risk of overdose.

Limits

• It used past claims and no one was randomly assigned, so the people in each pathway may have differed in ways that shaped the results.
• Only 963 people (2.4%) were in the naltrexone pathway, a small group next to the others, and the abstract doesn't say whether they took tablets or the injection.
• It covered only people with commercial or Medicare Advantage insurance.
• Overdose and acute care were counted from claims as stand-ins for a return to opioid use, not measured directly.

What it means if you're considering naltrexone

In this large real-world sample of people with opioid use disorder, buprenorphine and methadone were the only treatments linked to fewer overdoses and less serious acute care. Naltrexone was not, though the naltrexone group was small and an observational study can't explain why.

The study is about opioid use disorder only. It says nothing about naltrexone for alcohol.