How to Read the Relapse Estimate in a Five-Trial Bupropion–CBT Review
A 2023 meta-analysis estimated a relapse relative risk of 0.91 for bupropion plus cognitive behavioral therapy compared with placebo plus cognitive behavioral therapy. Its 95% confidence interval, 0.78–1.05, included the no-difference value of 1.
Which comparison was tested?
Apuy, Barreto and Merino included five randomized trials with an aggregate sample of 539 participants. Both groups received cognitive behavioral therapy. The comparison therefore concerns adding bupropion within that treatment context; it does not compare medication with no care.
The review defined relapse using urine-metabolite detection in the final weeks of follow-up and required at least ten weeks of follow-up. That operational definition is narrower than every possible meaning of recovery or recurrence.
A null result is not proof of equivalence
The point estimate favored the bupropion group, but the confidence interval included both a lower relapse risk and a small increase. The review did not establish a statistically clear reduction. It also did not prove the interventions were exactly equal.
Relative risk should not be converted into an absolute percentage-point benefit without the corresponding baseline risk. Nor does a group estimate determine whether a particular individual will respond.
A review reanalyzes earlier evidence
The authors identified heterogeneous populations and publication bias as limitations. The review is a synthesis of included trials, not a new independent participant cohort. Reading it alongside an underlying trial adds interpretive context without adding another independent experiment.
These findings describe the historical evidence included in the 2023 review. They do not establish the status of every later treatment study or supply a current prescribing recommendation.
Scientific source
Apuy LFM, Barreto MAB, Merino LAH. Efficacy of bupropion and cognitive behavioral therapy in the treatment of methamphetamine use disorder: a systematic review and meta-analysis. Brazilian Journal of Psychiatry. 2023;45:280–285. doi:10.47626/1516-4446-2022-2979. Based on the supplied review.



