The efficacy of non-nicotine smoking cessation drugs: a mixed treatment comparison

Article type
Authors
Chaiyakunapruk N, Hora W, Jumpahom J, Jampachaisri K, Sakunrak I
Abstract
Background: Because previous meta-analytic studies determining the efficacy of non-nicotine smoking cessation drugs were based on a classical frequentist approach, they did not provide a comparison of the overall efficacies of the individual products. Pooling relevant studies using Bayesian meta-analytic approach can provide the overall efficacy estimates. Objectives: To evaluate the efficacy of non-nicotine smoking cessation drugs, including varenicline, bupropion, clonidine and nortriptyline, on the long-term abstinence rate. Methods: Multiple computerized databases (1966 to 2008) including the Cochrane Central Register of Controlled Trials, MEDLINE, CINAHL, IPA, EMBASE, and Clinicaltrial.gov, were searched to identify relevant studies. To be included, studies had to be randomized, controlled trials evaluating the 52-week continuous abstinence rate (CAR) of patients receiving one of the following drugs: varenicline, bupropion, clonidine, nortriptyline. Using a standardized abstraction form, two reviewers extracted data on study design, patient population, intervention, and outcomes including a 52-week biochemically-verified CAR, which is the primary outcome measure of this study. Mixed treatment comparison was undertaken using Bayesian metaanalytyic approach. The overall risk ratio (RR) and 95% credible interval were calculated. Results: A total of 13 studies were included in this metaanalysis, comprising 4510 patients. The RR (95% credible interval) of varenicline, bupropion and nortriptyline compared to placebo were 2.53 (2.01, 3.11), 1.84 (1.54, 2.18), 1.70 (1.05, 2.59), respectively. The RR of varenicline versus bupropion calculated from indirect comparison was 1.38 (1.08, 1.71), varenicline versus nortriptyline was 1.57 (0.94, 2.49), and bupropion versus nortriptyline was 1.14 (0.69, 1.78). The probability of varenicline being the best drug is very high (94%), while the probability of bupropion and nortriptyline being the best is 1%, and 5%, respectively. Conclusions: Our findings indicate that varenicline, bupropion, and nortriptyline are effective smoking cessation drugs, compared to placebo. When smokers express interest in quitting, clinicians should consider using these medications as smoking cessation aids in patients without contraindication of these medications. Further research is needed to determine whether these medications are cost-effective interventions.