Best Practices for Hospital-Based Initiation of Medications for Opioid Use Disorder

Importance: Due to the impact of high-potency synthetic opioids on medication for opioid use disorder (MOUD) initiation, hospital-based addiction clinicians adapted their practice to improve care, with decisions often guided by limited evidence. Objective: To derive expert consensus on best practices for hospital-initiated MOUD among patients with OUD using high-potency synthetic opioids. Design, Setting, and Participants: This survey study used a 2-round Delphi process conducted from January to April 2025. National experts were hospital-based physicians or advanced practice clinicians who treated at least 100 hospitalized patients with OUD in the last 2 years and self-identified as an expert in methadone and buprenorphine initiation. Experts were purposively sampled based on practice location and addiction board certification. Data were analyzed from April 2025 to March 2026. Main Outcomes and Measures: The main outcome was consensus on the appropriateness of specific practices. Appropriateness was evaluated using a 9-point Likert scale from 1, indicating very inappropriate to 9, very appropriate. The presence of consensus was determined using the RAND/UCLA appropriateness method; if consensus was obtained, it was stratified into inappropriate (median 1-3), uncertain (median 4-6), or appropriate (median 7-9). Results: Of 48 experts, 42 (87.5%; median IQR age, 41 36.0-45.8 years; 25 59.5% female) completed round 1 and 41 of these (97.6%) completed round 2. Most were physicians (36 participants 85.7%); of those, 30 (83.3%) were board certified in addiction medicine, 3 (14.3%) in addiction psychiatry, 2 (5.6%) in both, and 1 (2.4%) in neither. Experts treated a median (IQR) of 200 (100-315) hospitalized patients with OUD in the past year. There was consensus that buprenorphine and methadone initiation were appropriate, while the appropriateness of naltrexone initiation was uncertain. Consensus was reached that rapid methadone initiation was appropriate. Regarding buprenorphine initiation practices, experts reached consensus that high- and low-dose initiation were appropriate, traditional initiation was of uncertain appropriateness, and rescue was inappropriate. There was consensus that provision of non-MOUD full agonist opioids was appropriate to treat opioid withdrawal during methadone initiation, as a bridge during buprenorphine initiation, and to treat withdrawal among patients declining MOUD. Conclusions and Relevance: In this survey study of hospital-based addiction specialty clinicians, consensus was reached on hospital-initiated MOUD and treatment of opioid withdrawal. These best practices can inform current clinical approaches and reflect the need for more robust research to evaluate effectiveness and safety.