Summary
Introduction:
This text discusses a study on the effectiveness of e-cigarettes in smoking cessation among unselected smokers and argues that the conclusion of the study is incorrect. The authors suggest the use of Bayesian methods as an alternative analysis to avoid misleading conclusions.
Key Points:
* The study is a pragmatic trial of e-cigarettes, incentives, and pharmacotherapy for smoking cessation in a population of unselected smokers.
* The authors concluded that the offer of e-cigarettes did not provide a benefit compared with usual care due to a p-value greater than 0.05.
* The odds ratio for this comparison was 8.21 (unadjusted 95% CI: 1.07, 63.25), indicating that e-cigarettes were more effective than usual care.
* The authors argue that the conclusion of no benefit is incorrect and the data provides weak evidence of there being an effect of e-cigarettes.
* The authors used Bayesian methods to disambiguate the results and calculated a Bayes factor of 2.05, indicating support for there being an effect of e-cigarettes.
* The study design was a pragmatic trial, with a usual care group and an e-cigarettes group.
* The primary outcome was the rate of sustained 6-month abstinence from smoking.
* The authors used Bayesian methods as an alternative analysis to avoid misleading conclusions.
Main Message:
The main message of this text is that the use of Bayesian methods in the analysis of randomized controlled trials, particularly in avoiding the error of concluding no effect based on failing to find a statistically significant difference, is important. The authors argue that the conclusion of the study they discuss is incorrect and that the data provides weak evidence of there being an effect of e-cigarettes. They encourage the addiction research community to make greater use of Bayesian methods and avoid misleading conclusions.
Citation
Brown J, Shahab L, West R. Does the offer of e‐cigarettes benefit smoking cessation among unselected smokers? Addiction. 2019;114(1):186-187. doi:10.1111/add.14415
