Helping Smokers Towards Help: A Reply to Hughes

Abstract

Willemsen et al . [1] reported that an innovative intervention to help smokers decide on the use of efficacious smoking cessation treatments increased quit attempts and point prevalence abstinence. They also concluded that, surprisingly, this occurred despite the fact that the aid 'failed to increase the usage of efficacious treatment'. I think this latter conclusion may be incorrect. This conclusion was evidently based on a comparison of the incidence of use of treatments in intervention versus control groups 'among respondents who made a quit attempt' in which the 95% CI for the difference always included 0. The authors acknowledge that this non-significant difference may be due to the smaller sample size of the subset of quit attempters ( n = 264). When I calculated the odds ratio for use of nicotine replacement therapy (NRT) in intervention versus control it was 1.8 and for bupropion it was 1.4. These are of the same magnitude as the effect of the intervention on quit attempts and abstinence (O R s = 1.5). To me, this suggests that changes in medication use may have been real and, thus, the abovequoted conclusion in the abstract (based on the P < 0.05 rule) may be misleading.One other puzzling finding is that p. 446 states: 'Those who attempted to quit and those who did not make a quit attempt did not differ significantly with respect to the usage of treatment aids'. I would conjecture that this result occurred due the very broad definition of treatment (e.g. including unsolicited brief advice from others). It would be interesting to hear from the authors whether this result occurred when medications were examined. This is of interest because, as overall about 18% of quit attempters in the control group used a medication (data from Table 3), then, based on the above statement, I would assume about 18% of smokers who did not attempt to quit also used a stop smoking treatment in the next 6 months. This level of use of medication by smokers not trying to stop would contradict prior surveys [2] and would suggest widespread misuse of medications. Perhaps the authors could clarify this.