The Research
Everything on this site is built on published research. This page shows you the studies themselves, what each one actually found, and where to read it yourself. Check us.
Staying in treatment cuts your risk of dying
Sordo et al., BMJ, 2017. A meta-analysis pooling 19 cohort studies.
- People on methadone died at a rate of 11.3 per 1,000 person-years while in treatment, versus 36.1 out of treatment. Roughly three times the risk when not in treatment.
- For overdose deaths specifically: 2.6 per 1,000 in treatment, versus 12.7 out of treatment. Nearly five times the risk.
- Buprenorphine showed the same pattern: 4.3 in treatment versus 9.5 out.
The first month after you stop is the most dangerous
Santo et al., 2021. A review of 36 observational studies and 15 randomized trials.
- All-cause death during treatment was less than half the rate during time out of treatment.
- Drug-related deaths were 59% lower during treatment. Suicide was 52% lower.
- Death rates spiked roughly sixfold in the first four weeks after leaving treatment.
That last line is the most practically important finding on this page. If you are thinking about stopping, the weeks right after are the window where people die. Talk to your clinic about tapering rather than stopping.
Treatment after an overdose changes what happens next
Larochelle et al., Annals of Internal Medicine, 2018. 17,568 Massachusetts adults who survived an opioid overdose, tracked for 12 months.
- Opioid overdose deaths were 59% lower among those treated with methadone.
- 38% lower among those treated with buprenorphine.
Most people who need this never get it
CDC, MMWR, 2024.
- An estimated 9.4 million U.S. adults needed treatment for opioid use disorder in 2022.
- About 1 in 4 received medication for it.
The problem isn't that the medication doesn't work. The problem is access.
The consensus position
National Academies of Sciences, Engineering, and Medicine, 2019 — Medications for Opioid Use Disorder Save Lives. The title is the finding. This is the closest thing there is to an official scientific consensus statement in the United States.
What this research does not say
Being straight about the limits matters as much as the headline numbers.
- It doesn't say methadone is right for everyone. Buprenorphine and naltrexone exist for good reasons, and the right fit depends on your history and your access.
- It doesn't settle how long you should stay on it. The evidence is clear that longer retention is protective. It does not name a finish line.
- Most of this is observational. These are large, consistent, repeatedly replicated studies — but people are not randomly assigned to leave treatment, so these show strong association rather than proof of cause in every case.
- None of it is about you specifically. Population data tells you what tends to happen. Your clinician knows your chart.
How to read a study like this
- "Per 1,000 person-years" means: if you followed 1,000 people for one year, this many would die. It lets you compare groups of different sizes.
- A meta-analysis pools many studies into one result. It's generally stronger evidence than any single study.
- Check the date. Federal methadone rules changed substantially in 2024. Anything written before then may describe a system that no longer exists.
If you think something on this page is wrong, out of date, or misrepresents a study, tell me and I'll correct it.
I'm not a doctor. I have 20 years of experience in this world, and I share what I know alongside sources you can check yourself. This is information, not medical advice. Your dose and your treatment plan are between you and your provider.