Best Practices for Starting MOUD in the Hospital
What landed
On May 7, JAMA Network Open published a Delphi-method consensus from 42 hospital addiction-medicine experts laying out best practices for initiating medications for opioid use disorder (MOUD) — specifically in the patient population where current standard protocols have been struggling: people whose primary opioid is fentanyl or another high-potency synthetic opioid (HPSO).
The panel addresses four areas: rapid methadone initiation, high-dose buprenorphine induction, low-dose (“micro-dose” or “Bernese”) buprenorphine initiation, and integration of long-acting injectable formulations (Sublocade, extended-release naltrexone). It also covers adjunctive medications for withdrawal management when standard inductions stall.
Why this matters
Hospitalization remains one of the highest-yield moments to start MOUD — patients are present, often in withdrawal, often newly aware of the danger of their use. But the standard buprenorphine induction was developed for a heroin-dominant supply. In a fentanyl-dominant supply, the half-life math of the drug already in someone’s system makes precipitated withdrawal much more likely, and patients have correspondingly been less willing to start. The result is a missed window — patients discharged without MOUD, often with the highest post-discharge overdose mortality risk in medicine.
The Delphi consensus is meaningful because it codifies what front-line addiction-medicine teams have been doing in practice — varying induction strategies by what the person was using and how recently — into something a community hospital without a dedicated addiction-medicine service can adopt. It is, in effect, the playbook for the most common point of clinical contact with the highest-risk patients.
For navigation infrastructure like Rize, this is operational: facilities that offer the full induction toolkit (rapid methadone, high-dose bup, low-dose bup, LAI integration) should be ranked higher for users reporting fentanyl as their primary opioid. Our V2 service matcher already incorporates evidence-based service mappings; the JAMA consensus is the next iteration of that taxonomy.
If you or someone you love is in crisis, call or text 988.
Sources Cited
- 01.B
Filed Under
treatmentsciencepolicy