Published this week in Plos One, Camponetto et al.'s twelve month study found that...
In smokers not intending to quit, the use of e-cigarettes, with or without nicotine, decreased cigarette consumption and elicited enduring tobacco abstinence without causing significant side effects.
The sample group was split between three sub-groups who were initially given a set of e-cigarettes to last them twelve weeks. Group A were given standard 7.2mg nicotine cartridges. Group B was given 7.2mg cartridges as well as a set of 5.4mg cartridges. Group C was given cartridges with no nicotine in them.
The results are shown below. All groups saw quit rates that were well above the norm for unmotivated smokers (8.7 per cent). Those who used the e-cigarettes that contained nicotine were nearly three times more likely to be abstinent after a year than those who used the zero nicotine products.
The remarkable thing about this study—as with the 2011 Polosa et al. study—is that the smokers were not motivated to quit beforehand. It is notoriously difficult for smokers to quit even when they want to. The Cochrane Review found that only 3-5 per cent of unassisted quit attempts are successful after 6 to 12 months (p. 13). Getting smokers who do not intend to quit to do so by merely giving them e-cigarettes to take home is extraordinary. To put it another way, the unmotivated smokers in this study who were given nicotine-containing e-cigarettes had a higher success rate than motivated smokers observed in other studies who are given nicotine patches. If the study had involved motivated quitters (and, perhaps, stronger nicotine cartridges), I am sure the results would have been even more impressive.
So this is good news for anti-smoking campaigners, no?
No. Not for those whose careers depend on following the neo-prohibitionist model. As Michael Siegel reported yesterday, Stanton Glantz has been sniffing around for information that might help him debunk the study and today he has launched his attack.
Glantz offers two criticisms. Firstly, he says that this double-blind, randomised controlled study doesn't really have a control group. In other words, we don't know how many of the unmotivated smokers would have suddenly given up smoking when the experiment began.
It seems to me that Group C is the control group. Not only is it the control group, but it is a control group that has been given the placebo of a zero nicotine e-cigarette. It is true that the study doesn't include a group of people who are given nothing at all but, as the study's authors note, that is hardly necessary when we know that the quit rate amongst unmotivated smokers is close to zero.
We considered it unrealistic to have a control for e-cigarette use per se in a study in which smokers were not interested in quitting. However, to provide an idea of the size of the effect, please consider that the quit rate of up to 8.7% at 1-year follow up in ECLAT compares very favorably with the national average cessation rate of 0.02% on an yearly basis over the 2001–2011 period in the general population
Glantz's second criticism is that the difference between quit rates in the nicotine groups (A & B = 11 per cent) and the quit rate in the zero nicotine group (C = 4 per cent) is not statistically significant. The authors say that the difference is significant at the normal 95 per cent confidence level (p = 0.04) but Glantz claims that if the result is adjusted in the way he thinks is appropriate, it is only significant at the 90 per cent level (p = 0.07). Therefore, he says, "the level of nicotine in the e-cigarette (including zero nicotine) has no detectable effect on quitting smoking".
I will let the statisticians argue about this technical point, but it is richly ironic to see Glantz dismissing a finding on the basis that it can only be shown with 90 per cent, rather than 95 per cent, confidence. I have written before about the way Glantz defended the Environmental Protection Agency when, in 1993, it dropped the 95 per cent confidence interval and replaced it with a 90 per cent confidence interval for its meta-analysis on secondhand smoke. Without doing this (and various other tricks), the EPA could never have found a correlation between secondhand smoke and lung cancer.
When the EPA was criticised for lowering the burden of proof, Glantz said that quibbles about confidence intervals was "hairsplitting that only professors care about". He also said: "I know that scientifically it's widely used, but there is a strong body of thought that people are too slavishly tied to 95 percent."
A few years later, when the World Health Organisation's IARC study failed to find a statistically significant link between secondhand smoke and lung cancer, Glantz suggested that the whole concept of 95 per cent confidence intervals was an invention of the tobacco industry:
"The [tobacco] industry imposes a one-sided interpretation of confidence intervals, focusing the entire discussion on whether the lower bound of the 95% CI [confidence interval] for a relative risk includes 1. By definition, if the lower bound exceeds 1, then the risk is statistically significantly raised (with p=0.05).
Whether or not there is anything magic about 95%, the true risk is equally likely to be anywhere inside the 95% CI, including values above the point estimate. In environmental and health and safety regulation, it is common to take the health-protective approach of basing public policy on the upper 95% confidence limit... The industry has represented the fact that the increase in risk observed did not reach statistical significance as indicating that the study did not find any increased risk."
For the record, I agree that epidemiological findings that fail to reach significance at the conventional 95 per cent level should not be taken too seriously. I therefore have something in common with the Stanton Glantz of 1978, who wrote an article to remind researchers about the importance of sticking to the 95 confidence interval. I also have something in common with the Stanton Glantz of 2013, who has just reaffirmed his rigid belief in 95 per cent confidence intervals. Alas, I have nothing in common with the Stanton Glantz of 1986-2012 who consistently defended weaker epidemiological findings and who claimed that those who insisted on standard epidemiological methods were "hairsplitting".
For more on Stanton Glantz's flexible views on science, see this old blog post.
(PS. If there is a legitimate criticism to be made of this study, it is the ludicrous non-acronym of its name. 'EffiCiency and Safety of an eLectronic cigAreTte' stands for ECLAT, apparently. What's wrong with ESEC?)
(PPS. A quick reminder that the IEA is holding a panel discussion about e-cigarettes on July 15th. Free if you RSVP.)
Michael Siegel and Rory Morrison take a similar view.