Journal of Consulting and Clinical Psychology, Vol 94(5), May 2026, 256-266; doi:10.1037/ccp0001013
Objective: Reduction of craving is a key mechanism of buprenorphine treatment for opioid use disorder but has not been examined as an indicator of early response versus nonresponse. We aimed to (a) identify cut points for craving in the first week of buprenorphine treatment that best predicted later sustained opioid use and (b) examine associations between early nonresponse based on craving and opioid-related, retention, and psychosocial outcomes. Method: Data were from two buprenorphine trials (N = 562; Mage = 34 years; 38% female). Craving was rated on a 0–10 visual analog scale in each of the first 4 weeks. Area under the receiver operating characteristic curve was used to identify optimal craving cut points predicting sustained opioid use during Weeks 5–12. Associations between craving-based early nonresponse and opioid-free weeks, buprenorphine retention, and psychosocial outcomes were examined using regression models. Results: Sustained opioid use was optimally predicted by craving ≥2 in Week 1 (sensitivity = 0.700; specificity = 0.513, positive predictive value = 0.398, negative predictive value = 0.788). Participants above this threshold had 2.53 times greater odds 95% CI = [1.69, 3.81] of sustained opioid use. Craving-based nonresponse was also associated with fewer opioid-free weeks and poorer end-of-treatment depression and quality of life outcomes. Conclusions: Even mild craving early in buprenorphine treatment may signal an increased risk of poor outcomes, thus identifying those who could benefit from monitoring and low-effort interventions to manage craving. However, future work is needed to optimize the prediction of poor buprenorphine outcomes, potentially using models integrating both craving and opioid use in the first weeks of treatment. (PsycInfo Database Record (c) 2026 APA, all rights reserved)