Letting methadone patients take doses home earlier tracked with lower death and dropout
A British Columbia study emulated a trial across 41,000 patients. Those started on take-home methadone within weeks to months of induction stayed in treatment longer and died less than those on daily witnessed dosing.
Making patients swallow every methadone dose under a pharmacist's eye is meant to prevent diversion, but it may also drive people out of treatment that keeps them alive. A large British Columbia study in JAMA Internal Medicine found that patients allowed to take methadone home within the first weeks to months of treatment were less likely to die and less likely to drop out than those kept on daily witnessed dosing [s1]. The finding is observational, but the analysis was built to imitate a randomised trial, and the pattern held across several checks [s1].
Methadone is one of the two main medicines for opioid use disorder, and staying on it is the point: retention is what lowers overdose deaths. In British Columbia, daily witnessed ingestion at a community pharmacy has been standard practice — a patient must attend in person to take each dose [s1]. Whether that requirement, compared with take-home doses, becomes a barrier that pushes people out of treatment had not been established [s1].
How the study was built
Because randomly assigning patients to "no take-home doses" would be hard to justify, the researchers used target trial emulation — a method that arranges observational records to answer the question a trial would, using explicit rules for who is compared with whom and when the clock starts [s1]. They drew on British Columbia data from 2010 to 2022 and analysed both incident users (no prior opioid agonist treatment) and prevalent new users (none in the past month), restricted to adults who had completed induction and were not pregnant, incarcerated, or in cancer or palliative care [s1].
Patients were grouped by when take-home dosing began: not at all, or within 0 to 4 weeks, 5 to 12 weeks, 13 to 24 weeks, or 25 to 52 weeks after completing induction [s1]. The two outcomes were time to all-cause death and time to methadone discontinuation, defined as a gap of five or more days in prescribed doses [s1]. The analysis estimated the adjusted risk difference at 78 weeks — the change in the percentage of patients who died, or discontinued, versus witnessed dosing alone [s1]. It included 9,788 incident users (median age 34.1; 32.8% female) and 31,658 prevalent new users (median age 36.9; 33.3% female) [s1].
What it found
Among prevalent new users, earlier take-home dosing was associated with lower 78-week mortality than no take-home dosing: an adjusted risk difference of −0.87 percentage points (95% CI, −1.51 to −0.23) for initiation within 0 to 4 weeks, −0.98 (−1.62 to −0.33) for 5 to 12 weeks, and −0.82 (−1.59 to −0.06) for 13 to 24 weeks [s1]. For the latest window, 25 to 52 weeks, the estimate was −0.31 and its interval crossed zero (−1.05 to 0.44) [s1].
The effect on staying in treatment was larger. Discontinuation fell by −2.04 percentage points (95% CI, −3.99 to −0.10) for take-home dosing started within 0 to 4 weeks, −5.29 (−6.82 to −3.76) for 5 to 12 weeks, −5.97 (−7.16 to −4.78) for 13 to 24 weeks, and −4.39 (−5.37 to −3.42) for 25 to 52 weeks [s1]. The same patterns appeared among incident users, and multiple sensitivity analyses supported the main results [s1].
What to take from it, and what not to
The direction is consistent: across windows, allowing take-home methadone tracked with better retention and, in most windows, lower mortality, with the strongest and most stable effects on staying in treatment [s1]. That is the outcome that ultimately protects against overdose. The absolute differences are modest in percentage-point terms, but at the scale of a treatment population they represent real numbers of people retained or lost.
The limits are those of the design. Target trial emulation reduces confounding but cannot remove it: clinicians grant take-home privileges to patients they judge more stable, and even careful adjustment may not fully separate the effect of the policy from the characteristics of the patients who receive it [s1]. This is one health system's data over one period, and diversion risk — the reason witnessed dosing exists — is a real countervailing concern the mortality and retention numbers do not price in. What the study adds is evidence that the default of universal daily witnessed dosing carries its own cost, in patients who leave treatment. It sits alongside other work reweighing how opioid use disorder is managed, including trials of repurposed GLP-1 drugs as possible adjuncts.
Sources
- [s1] Timing of Initiation of Methadone Take-Home Dosing, JAMA Internal Medicine, 8 September 2026.
Sources
- Timing of Initiation of Methadone Take-Home Dosing — JAMA Internal Medicine , September 8, 2026
From rodent cages to Medicare claims: GLP-1 drugs and opioid use disorder
A review lays out the evidence chain: animal studies showing reduced drug-seeking, one small human trial, and record-based analyses linking GLP-1 prescriptions to 40% fewer opioid overdoses.
A drug already used for achondroplasia sped up growth in a related skeletal condition
In an 81-child trial, vosoritide raised annualised growth velocity by 2.33 cm a year against placebo in hypochondroplasia, which has no targeted treatment. The endpoint is growth speed, not final height.
Muse and other brain-sensing headbands: real EEG, thin evidence they help
A meta-analysis of consumer neurofeedback found no benefit for cognition, mindfulness or physical health and only a small effect on distress — even though the headbands do record real brain activity.
As ADHD and autism diagnoses rose, newly diagnosed children came to resemble their peers
A Danish study of 71,000 cases found the risk factors tied to a diagnosis weakened over a decade, pointing to changing diagnostic practice rather than falling thresholds of impairment.