Real‐world comparative effects of slow‐release oral morphine vs. methadone on healthcare use for opioid use disorder: A population‐based target trial emulation

AIMS: To examine the comparative effects of slow-release oral morphine (SROM) vs. methadone for the treatment of opioid use disorder (OUD) on healthcare use. DESIGN: This retrospective cohort study emulated a target trial using electronic medical record data from Vancouver Coastal Health Authority linked with population-based administrative health data from British Columbia (BC), Canada. SETTING: Vancouver Coastal Health region in BC, between 1 July 2017 and 30 June 2024, operating under a universal healthcare system, where the majority of illicit opioids are contaminated with fentanyl. PARTICIPANTS: OUD patients aged between 18 and 65 years. INTERVENTIONS: Intent-to-treat (ITT): receiving a new prescription of SROM or methadone for the treatment of OUD (regardless of treatment initiation or adherence). Per-protocol (PP): initiating and adhering to the prescribed SROM vs methadone (real-world on-treatment exposure). MEASUREMENTS: The daily rate of healthcare use-including all-cause and opioid-specific emergency department (ED) visits, all-cause and opioid-specific hospitalization days, all-cause physician encounters/services, OUD-related physician visits, and non-OUD-related primary care physician visits-over 1 year following treatment prescription or initiation was modelled using weighted modified Poisson regression with generalized estimating equations. Cumulative healthcare use at 1 year and corresponding adjusted rate ratios (aRRs) were estimated. FINDINGS: We identified 3254 unique individuals [median (Q1-Q3) age 37 (30-46) years, 64.2% male] contributing to 4059 person-trials (eligible treatment episodes) for the ITT analysis (32.6% SROM) and 1992 unique individuals contributing to 2276 person-trials for the PP analysis (27.4% SROM). ITT analysis shows that receiving a SROM vs. methadone prescription was not associated with statistically significant differences in the rates of all-cause [aRR = 1.10, 95% confidence interval (CI) = 0.997-1.20] and opioid-specific ED visits (aRR = 1.00, 95% CI = 0.90-1.13), all-cause (aRR = 0.99, 95% CI = 0.81-1.18) and opioid-specific hospitalization days (aRR = 0.97, 95% CI = 0.78-1.20), all-cause physician encounters/services (aRR = 1.00, 95% CI = 0.94-1.06), OUD-related physician visits (aRR = 0.94, 95% CI = 0.87-1.01) and primary care physician visits (aRR = 1.11, 95% CI = 0.98-1.24). PP analysis also did not observe statistically significant differences across healthcare outcomes among those who initiated and remained on SROM vs. methadone treatment, with aRRs of 1.24 (95% CI = 0.91-1.50), 1.01 (95% CI = 0.67-1.36), 1.06 (95% CI = 0.65-1.58), 0.95 (95% CI = 0.53-1.58), 1.03 (95% CI = 0.90-1.17), 0.97 (95% CI = 0.82-1.14) and 1.06 (95% CI = 0.73-1.46), respectively. CONCLUSIONS: This emulated trial on opioid use disorder patients under a universal healthcare setting in Canada did not observe statistically significant differences in risk of healthcare use outcomes between the patients receiving slow-release oral morphine and those receiving methadone following treatment prescription and initiation. This supports slow-release oral morphine as a viable alternative to methadone with similar real-world effectiveness in preventing all-cause and opioid-related acute healthcare needs and the potential to expand treatment options for opioid use disorder without increasing health system burden.