Written by Rebecca DiFabio
Spoon Feed
In this randomized trial, there was no difference in engagement with opioid use disorder (OUD) treatment at 7 days among groups that received an extended-release (ER) buprenorphine injection in the ED versus a sublingual (SL) buprenorphine prescription.
Dealer’s choice
Although opioid overdose deaths are decreasing, they still remain the primary cause of death in individuals younger than 45 years old. ED-initiated buprenorphine is cost-effective and has been shown to improve OUD treatment engagement, though wholly underutilized.
This multicenter, open-label, randomized trial investigated engagement in formal OUD treatment at day 7 between individuals presenting with a COWS score of at least 4 who received a 24-mg subcutaneous ER buprenorphine injection vs. a prescription for SL buprenorphine. In the SL group, the first dose was administered in the ED if the initial COWS score was 8 or higher. There was no difference in OUD treatment engagement on day 7 between the 854 ER recipients (40.5%, 95%CI 31.3%–50.3%) and the 870 SL recipients (38.5%, 95%CI 29.5%–48.5%).
Importantly, rates of precipitated withdrawal were low, at 0.6% and 0.8% in the ER and SL groups. There was no difference in self-reported nonfatal overdoses within 30 days after enrollment (2.3% in both groups).
Due to all study sites needing to develop partnerships with OUD treatment programs, follow-up appointments may have been easier for participants to access than those typically available in the community, decreasing generalizability.
How does this change my practice?
I have personally been in a group of EM physicians who have been underprescribing buprenorphine for my patients. Working with a population that faces multiple barriers to care, such as housing instability and being un/underinsured, I will start considering administration of ER buprenorphine as a bridge to outpatient OUD treatment.
Source
Emergency Department-Initiated Buprenorphine for Opioid Use Disorder: A Randomized Clinical Trial. JAMA. 2026 Feb 11:e2527019. doi: 10.1001/jama.2025.27019. Epub ahead of print. PMID: 41670966; PMCID: PMC12895321.
View JournalFeed Critical Appraisal
Critical Appraisal
Study Identification
Background
Study Question
Study Design & Conduct
Prospective / Retrospective: Prospective
Multicenter: Yes
Unit of Allocation: Patient
Unit of Analysis: Patient
Randomization Method: Permuted-block procedure with random block sizes and stratification by site and insurance status
Allocation Concealment: Concealed
Blinding: Open-label
Follow-up Duration: 7 days for primary outcome, 30 days for secondary outcomes
Population
- Aged 18 years or older
- DSM-5 criteria for moderate to severe opioid use disorder
- Urine toxicology results demonstrating presence of an opioid
- Clinical Opiate Withdrawal Scale (COWS) score of 4 or higher
- Methadone detected in urine
- Pregnant
- Received medications for opioid use disorder within the past 14 days
- Required ongoing opioid medication for pain
- Active suicidal ideation or a medical or psychiatric condition requiring hospitalization
- Non-English speaking
- Unable to provide locator information
- Unwilling to follow study procedures
- Incarcerated or in police custody
Number Enrolled: 2000
Number Analyzed: 1724
Key Baseline Characteristics
Sex: 68% male
Disease Severity: COWS score at enrollment: 4-7 (30-31.1%), >=8 (68.9-70%); Median (IQR) 9 (7-13)
Care Setting Distribution: Emergency Department
Additional Baseline Characteristics
- 76% tested positive for fentanyl
- 49.9-52.3% reported unstable housing in past 12 months
- 74.6-75.1% had public insurance
- 40% had buprenorphine detected in urine at baseline
Exposures / Interventions
Description: Extended-release injectable buprenorphine
Definition / Dose: 24-mg subcutaneous injection (equivalent to 16 mg/d)
Timing: Initiated in the ED
Classification Method: Randomization
Protocolized / Discretionary: Protocolized
Description: Sublingual buprenorphine
Definition: Initial dose in ED (8 mg if COWS >= 8, or self-administration instructions for up to 12 mg if COWS < 8) + 7-day prescription for 16 mg/d
Outcomes & Results
Primary Outcomes
Definition: Patient report of engagement verified by the treatment source (clinician, clinic, or opioid treatment program), defined per American Society of Addiction Medicine levels of care 1 through 4.
Time Point: 7 days
Measurement Method: Patient report verified by treatment source
Results: Extended-release: Adjusted 40.5% (95% CI, 31.3%-50.3%); Sublingual: Adjusted 38.5% (95% CI, 29.5%-48.5%); Adjusted difference: 1.6% (95% CI, -2.8% to 6.0%); P=0.44
Secondary Outcomes
Definition: Patient report of engagement verified by the treatment source (clinician, clinic, or opioid treatment program), defined per American Society of Addiction Medicine levels of care 1 through 4.
Time Point: 30 days
Measurement Method: Patient report verified by treatment source
Results: Extended-release: Adjusted 43.8% (95% CI, 34.3%-53.7%); Sublingual: Adjusted 44.9% (95% CI, 35.2%-55.0%); Adjusted difference: -1.5% (95% CI, -6.2% to 3.2%); P=0.68
Definition: Increase of 5 or more points in COWS score during ED course, reviewed by independent investigators.
Time Point: Not reported
Measurement Method: COWS score monitoring and independent investigator review
Results: Extended-release: 6 (0.6%); Sublingual: 8 (0.8%)
Definition: Self-reported nonfatal overdose events.
Time Point: 30 days
Measurement Method: Self-report
Results: Extended-release: 18 participants (2.3%); Sublingual: 18 participants (2.3%)
Definition: Visual analog scale (0-100, higher scores indicate more craving).
Time Point: 7 days
Measurement Method: Self-report (visual analog scale)
Results: Extended-release: Mean 26.5 (SD 32.8); Sublingual: Mean 30.2 (SD 34.2); Adjusted mean difference: -3.85 (95% CI, -7.08 to -0.63)
Definition: Self-reported days of illicit opioid use.
Time Point: Past 7 days
Measurement Method: Self-report (timeline follow-back method)
Results: Extended-release: Adjusted ratio of means 0.77 (95% CI, 0.68-0.95)
Definition: Overall experience with medication (1-5 scale, 5 being completely effective).
Time Point: 7 days
Measurement Method: Self-report
Results: Extended-release: Mean 4.12 (SD 1.21); Sublingual: Mean 3.99 (SD 1.27); Adjusted mean difference: 0.13 (95% CI, 0.01-0.25)
Definition: Number of ED visits.
Time Point: 30 days
Measurement Method: Self-report and electronic health record query
Results: Sublingual: 388 ED visits vs Extended-release: 291 ED visits (driven by small number of individuals with multiple visits)
Definition: Number of hospitalizations.
Time Point: 30 days
Measurement Method: Self-report and electronic health record query
Results: Not reported
Definition: Progress in treatment from patients' perspectives across 4 domains (substance use, health, lifestyle, community) on a 1-10 scale.
Time Point: Not reported
Measurement Method: Self-report
Results: Higher in extended-release buprenorphine group in all domains
Risk of Bias
Risk of Bias - RoB 2
- Bias arising from the randomization process (Low): The study used a permuted-block randomization procedure with stratification by site and insurance status, and allocation was concealed.
- Bias due to deviations from intended interventions (Some concerns): The study was open-label, meaning participants and personnel were aware of the assigned intervention. While the primary outcome was verified by a treatment source, some secondary outcomes relied on self-report, which could be influenced by lack of blinding.
- Bias due to missing outcome data (Low): The study used a weighted generalized estimating equation to account for missing data, and sensitivity analyses were performed for missing primary outcome data.
- Bias in measurement of the outcome (Some concerns): While the primary outcome (engagement) was verified by a treatment source, several secondary outcomes (craving, illicit opioid use, patient satisfaction, overdose events) relied on self-report, which could introduce bias.
- Bias in selection of the reported result (Low): Primary and secondary outcomes were pre-specified in the trial registration and protocol.
Transparency
COI Statement Present: TRUE
