Introduction: Though opioid agonist therapies are the mainstay of treatment for opioid use disorder, treatment retention remains suboptimal. Improved prediction of who will remain in treatment could lead to improved treatment outcomes. Whether craving predicts reduced retention in treatment remains debated. We performed analyses to determine whether craving predicted treatment attrition or treatment switching in people with non-heroin opioid use disorder initiating opioid agonist therapy.
Methods: Our data came from the OPTIMA trial - a pan-Canadian, pragmatic, open-label, randomized controlled trial that compared a flexible, early take-home buprenorphine/naloxone model of care (n = 137) to standard treatment with methadone (n = 132) for non-heroin opioid use disorder over a period of 24 weeks. We performed Cox proportional hazards regression to conduct survival analyses of time (days) to treatment attrition, and time to switch to another treatment, with craving as a time-varying covariate, controlling for assigned treatment group, lifetime history of heroin use and province. Craving was measured at baseline, week 2, 6, 10, 14, 18, 22 using the Brief Substance Craving Scale.
Results: We found that craving predicted both treatment drop out and treatment switching. A 1-point increase in craving was associated with a 15.3 % increase of risk of dropping out of the study (HR = 1.153, 95 % CI = 1.065 to 1.248, p < 0.001) and with a 11.5 % increase of risk of switching treatment (HR = 1.115, 95 % CI = 1.016 to 1.225, p = 0.022).
Conclusions: Craving predicted both treatment attrition and treatment switching in people receiving buprenorphine/naloxone or methadone models of care for non-heroin opioid use disorder. These findings highlight the importance of targeting and better addressing craving during treatment with opioid agonist therapies.
Keywords: Buprenorphine; Craving; Methadone; Opioid; Retention; Switching.
Copyright © 2025 The Authors. Published by Elsevier Inc. All rights reserved.