As I correctly predicted yesterday based on his one-line email inquiry, electronic cigarette opponent (and my mentor, hero, and colleague) Stan Glantz was nonplussed with the clinical trial showing that electronic cigarettes achieved an almost miraculous 13% one-year quit rate among Italian smokers who had no interest in quitting. As I also correctly predicted, he quickly scrambled to try to discredit the study in order to support what I believe is a predetermined conclusion in his mind that electronic cigarettes are useless.
Today, he posted on his blog a critique of the study.
In his critique, Stan makes two main arguments:
1. "There is not a control group of people who were not using e-cigarettes that would allow assessment of spontaneous quit rates.
By not having a true control group that would account for spontaneous
quitting without using e-cigarettes one cannot say anything about
whether e-cigarettes affected quitting."
2. "The [second] problem is that the authors failed to include the required Yates
correction in their calculation of the chi-square test statistic and
associated p value. Recalculating the test properly yields p = 0.07,
which is no longer statistically significant. Thus, the correct
conclusion is that there is no statistically significant difference
between the nicotine and non-nicotine e-cigarettes."
The Rest of the Story
Let's take each of these arguments in turn.
First, it is true that the study did not include any control group. However, that doesn't necessarily mean that we have no idea what the quit rate would have been for smokers with no interest in quitting in the absence of electronic cigarettes. We can simply examine the ample scientific evidence reporting one-year quit rates among smokers in the absence of intervention.
However, as I noted yesterday, what we really want is the one-year quit rate among smokers with no interest in quitting. To be conservative, we can use the one-year quit rate among smokers who do want to quit (and try to quit), which is about 3%.
Clearly, the quit rate observed in the high-nicotine electronic cigarette group (13%) is considerably larger than the 3% observed for the population of smokers who want to quit and make quit attempts. The natural quit rate among smokers with no desire to quit would be substantially lower than 3%.
In fact, the paper provides population-based data showing that in Italy, the spontaneous quit rate during the study period was a dismal 0.02%.
So this first argument does not invalidate the study's conclusion that electronic cigarettes did help achieve smoking cessation for some smokers. There is no scenario under which you would observe 13% of non-motivated smokers quitting within one year without any intervention.
It is also important to point out that the comparison group Stan requests is not possible. One could not ethically conduct a clinical trial of smoking cessation in which one group is assigned to receive no intervention.
What is really needed is a head-to-head comparison of electronic cigarettes compared to nicotine replacement therapy (NRT).
Stan's second argument is that the study conclusions are invalid because of failure to apply the Yates correction. This argument is invalid because it is a sort of sleight-of-hand. It is kind of like magic, where the magician can succeed in a trick by diverting the audience's attention away from the central event taking place.
Why is this argument a diversion? It is a diversion because there is no placebo in this trial. Had this study compared electronic cigarettes with a placebo, then the appropriate analysis would indeed be to statistically compare the cessation rates in the e-cigarette vs. placebo groups. But there was no placebo. The group receiving electronic cigarettes without nicotine was in fact an intervention arm. Even without nicotine, electronic cigarettes have been shown to reduce the craving to smoke. And in fact, in this trial, the most remarkable finding was that in the zero nicotine e-cigarette group, 14% either quit or cut down by more than half. This is remarkable for a group of smokers who had no intention to quit and who received no non-tobacco cigarette nicotine for an entire year.
The paper actually acknowledges the lack of statistical significance in quit rates between the three study arms at one year. But this doesn't mean the treatment wasn't effective. It reflects the small sample size of the study along with the fact that there was some effect among the 0 nicotine e-cigarette group.
Stan is absolutely right that the correct comparison to make to evaluate the results of the study is how the observed quit rate (about 9% for all three groups combined) compares to the spontaneous one-year quit rate among the population of smokers, except that one would have to restrict that to the population of smokers who have no interest in quitting.
If anyone can show me data demonstrating that more than 2% of smokers with no interest in quitting achieve smoking cessation over a one-year period, then I will retract my conclusion about this study. But short of that, it cannot be denied that the results achieved in this study are better than what one would have obtained with no intervention.