Methadone vs Buprenorphine After Opioid Overdose

A matched cohort study published in JAMA Network Open found that people who began methadone after a recent opioid overdose had lower 1-year all-cause mortality than those who began buprenorphine-naloxone. The investigators cautioned that the observational result may reflect unmeasured confounding and requires confirmation in other settings.

The target trial emulation used health administrative data from Ontario, Canada, and included residents who started either medication from 2017 through 2023 after an emergency department visit for opioid overdose in the preceding year. Participants had not received either treatment during the previous 7 days. Propensity-score matching produced 2 equal treatment groups within a cohort of 5,882 people; the mean age was 35.8 years, and 32.1% were female.

Within 1 year, 165 people who initiated methadone, or 5.6%, died compared with 210 who initiated buprenorphine-naloxone, or 7.1%. Methadone initiation was associated with a lower mortality hazard (hazard ratio [HR], 0.78; 95% CI, 0.63-0.95). Median time to treatment discontinuation was 25 days with methadone and 16 days with buprenorphine-naloxone (HR, 0.78; 95% CI, 0.74-0.82).

The initiator analysis found no significant difference in time to first opioid overdose (HR, 1.07; 95% CI, 0.97-1.18). In per-protocol analyses limited to periods before treatment discontinuation, mortality did not differ significantly (HR, 0.84; 95% CI, 0.48-1.47), while methadone was associated with a higher overdose hazard (HR, 1.52; 95% CI, 1.19-1.93).

The authors noted that treatment duration was short in both groups and that the mortality association was not present during active treatment. They concluded that additional epidemiologic studies are needed to determine whether the observed difference generalizes beyond Ontario and to clarify the potential influence of residual confounding.

Source: Kleinman RA, Larney S, Sule NO, et al. Methadone vs buprenorphine-naloxone in opioid overdose survivors. JAMA Netw Open. 2026;9(8):e2629313. doi:10.1001/jamanetworkopen.2026.29313