A new piece of research has been released suggesting that the primary benefit of e-cigarettes, that they help people quit smoking, isn’t true when it comes to cancer patients. This finding has led to headlines such as “E-cigarettes fail to help cancer patients give up smoking,” and given the importance of stopping smoking for cancer patients, seems as though it could be a damning piece of evidence against the technology. However, for anybody that’s been following the research to date, the disturbing trend of headlines like this being completely unsupported by the study that purports to show it may lead you to expect that there’s more to this story than the media would have you believe. And you’d be completely correct in that assumption. In their infinite credulity, reporters have repeated these claims without so much as reading the research to see if it’s a fair analysis. Unless you think those who’ve already tried to quit using e-cigs but still smoke are the people we should be basing sweeping conclusions on, then there is definitely more to this story than meets the eye.
Summary
- Researchers aimed to determine whether or not cancer patients who’ve used e-cigarettes are more or less likely to quit smoking than those who haven’t, and also look at differences in their characteristics.
- All of the participants in the study were still smoking at the time of the baseline survey, and were classed as “e-cig users” if they’d vaped in the month prior to baseline but still continued to smoke. There was no confirmation that they used them in the studied period.
- All participants (regardless of group) were given the standard program of treatment at the cancer center, including counseling and medication as needed.
- The “e-cigarette users” in the study smoked more cigarettes per day, were more highly dependent on nicotine, were more likely to have tried to quit two or more times in the past, and hadn’t maintained abstinence for as long in the past as the non-users. In short, they appear more addicted than the non-users.
- Based on the follow-up data they received (discounting non-responses), there was no significant difference in the quit rate or odds of still smoking between the “e-cig users” and non-users in the study. If non-respondents were classed as returning to smoking, the “e-cig users” were twice as likely to still be smoking.
- The authors use this data to advise cancer doctors against suggesting e-cigs to smokers suffering from cancer, and make several unrelated and evidentially unsupported statements about e-cigs in the paper’s discussion.
- The findings clearly stem from the fact that the authors only included those who’d already failed to quit smoking by vaping in the “e-cig users” group – they’re more addicted to smoking, and will obviously not experience increased quit rates compared to ordinary cancer-suffering smokers, especially when given identical treatment.
What They Did – Looking at Smoking Cessation Outcomes Among Cancer Patients
The study (the full text is paywalled) was published online in the journal Cancer on September 22nd, and the researchers aimed to look at how frequently cancer patients who smoke use e-cigarettes, to determine the characteristics of vapers and to see whether vaping helps these patients quit smoking. Patients who presented to the Memorial Sloan Kettering Cancer Center between the beginning of 2012 to the end of 2013 were screened, and the current smokers among them (who had smoked cigarettes in the past 30 days) were then referred to “multi-component, evidence-based behavioral and pharmacological treatment” to help them quit. This includes up to five sessions of behavioral counseling, either conducted in person or over the phone, and patients are offered cessation medicines according to the group’s practice guidelines. The initial screening also assessed nicotine dependence (using the Fagerstrom test), and took general demographic information.
According to the full paper, the researchers asked the participants whether they’d used e-cigarettes in the last 30 days at the initial assessment, and those answering “yes” were classed as “e-cigarette users.” This immediately shows the severe, debilitating flaws with this study. Firstly, and most importantly, in order to be included in the study at all, the “e-cigarette using” participants had to be still smoking, thus clearly indicating that they’d used e-cigarettes but not successfully quit with them. If they had successfully quit by vaping, they would not be classed as smokers and would not be eligible to participate.
As Dr. Michael Siegel puts it, if 1000 e-cigarette users entered the cancer center, and an amazing 900 had been able to quit smoking entirely, the 100 remaining smokers would be those who struggled to quit even using e-cigarettes (the “tough nuts to crack,” so to speak), but these would be the only ones included in the research. The 90 percent of this hypothetical group who did quit smoking would be ignored, and the authors instead would focus on the unsuccessful ones. Some smokers obviously won’t be helped by vaping (unless e-cigs were 100 percent effective), and just focusing on these is at best disingenuous and at worst purposefully misleading.
But this one small paragraph on their assessment of e-cig use reveals even more issues with the study. The problem is that the initial assessment is the only time e-cigarette use was measured. Between 6 to 12 months after the initial assessment (in the study, this ended up being an average of 10 months afterwards), patients were followed up to see if they were still smoking, with the researchers asking if they’d had even a single puff of a (traditional) cigarette in the past week, and if so how many cigarettes per day they smoked, as well as whether they’d gone a full 24 hours without smoking during the study period. There was no question to determine whether the “e-cigarette users” had continued to vape through the studied quit attempt. In fact, the only thing the researchers can say is that the “e-cigarette users” had tried vaping (but continued to smoke) at some point in the month prior to the research, and they were otherwise treated in the same way as the remainder of the participants (receiving their standard pattern of treatment).
One final important detail from the study’s methodology section is how they dealt with those who weren’t contactable at follow-up. They either counted these individuals as still smoking or removed them entirely from the analysis, but both of these results were reported. They also conducted statistical analyses to determine differences between the “e-cig users” and non-users.
What They Found – No Difference in the Likelihood of Quitting Smoking
At the initial assessment, the researchers had a sample of 1074 participants, but they only successfully obtained cessation outcomes at follow-up for 414 of these (discounting those ineligible, those who died and those who couldn’t be contacted). E-cigarette users were much more likely to drop out (66.3 percent compared to 32.4 percent of non-users), indicating that how those who drop out are treated statistically (either being discounted or being classed as smokers) will notably influence the findings.
At enrolment, 26.5 percent of the whole sample had used e-cigarettes in the past 30 days, and 92 percent of these were dual users of both e-cigarettes and cigarettes. Assumedly, given that you had to be a current smoker to be included in this study, the non-dual users must have tried to completely substitute smoking with vaping in the past month but had failed and returned to cigarettes alone. Consistent with the increase in popularity of e-cigarette during the study period, the authors found that 10.6 percent had used e-cigs in the past month in 2012, compared to 38.5 percent in 2013.
The “tough nut to crack” theory is well-supported by the data collected on the differences between the “e-cig users” and non-users in this study. The e-cig group smoked more cigarettes per day, were more likely to have been diagnosed with lung, head or neck cancers, scored higher on the test of nicotine dependence overall, contained more “highly nicotine dependent” individuals (with 51.8 percent of the e-cig group scoring over 5 on the Fagerstrom test, compared to just 32.2 percent of the non-users) and twice as many of them found it difficult to avoid smoking in places they weren’t allowed to do so. Additionally, e-cigarette users were more likely to have tried to quit twice or more previously, with 76.5 percent of the group reporting this compared to 66.7 percent of non-users. According to the data tables in the full paper (although it was reported in the text the opposite way around), the e-cigarette group had only gone an average of 5.8 days without cigarettes in the past, compared to 7.4 days for the non-users. All of these differences were statistically significant and suggest that the e-cig group had more difficulty quitting smoking than the non-e-cig group.
Now the central finding of the research is clearly put into context. If the participants who didn’t complete a follow-up were ignored, then e-cigarette users and non-users had very similar quit-rates, 44.4 percent of the e-cig group compared to 43.1 percent of the non-users (a non-significant difference). When statistical adjustments were made for dependence, past quit attempts and the differences in cancer diagnosis, e-cig users were just as likely to be smoking at follow up as non-users (an odds ratio of 1.0). Between enrolment and follow-up, only around half of e-cig users had stopped smoking for a period of 24 hours, compared to 76 percent of the non-users. Additionally, follow up indicated that the “e-cig users” smoked slightly (but not significantly) more cigarettes per day than the non-users.
When the non-respondents were classed as having gone back to smoking, it had a predictable impact on the findings, changing the abstinence rate for the e-cig group to 14.5 percent in comparison to 30 percent of the non-users. Analyzed in this fashion, e-cigarette users appeared to be twice as likely to be smoking as non-users (an odds ratio of 2.0). The authors themselves admit that the differences in the number of participants lost to follow-up between groups mean that these figures “should be interpreted with caution.”
What Does it Mean? Are E-Cigarettes Ineffective?
According to the authors, “the findings of the current study raise some doubt about the usefulness of E-cigarettes for facilitating smoking cessation among patients with cancer,” and they go on to argue that the smaller numbers of “e-cigarette users” remaining abstinent for at least a day suggest “that the use of E-cigarettes may have averted or delayed quit attempts.”
This generally one-sided analysis continues, claiming that the study provides no evidence to support oncologists recommending that their patients try to quit through e-cigs, and bizarrely (in absence of any evidence on this subject in the study or elsewhere) that people with thoracic or head and neck cancer may be “more influenced” by claims that e-cigs help smokers quit or reduce harm. They say oncologists should inform patients of the “potential risks” and “lack of known benefits” of e-cigs for long-term quitting, and instead suggest FDA-approved methods. They also suggest (stepping well beyond the evidence provided in this study or anywhere else) that hospitals should include e-cigs in their smoke-free air policies, because the nicotine and “other constituents” in the vapor may result in “respiratory irritation and other as-yet unknown health effects.”
It’s abundantly clear that these comments are politically motivated, since most were not supported by the current study, and any fair analysis of the other available evidence would have called many of the claims into question. So what can they say? Well, based on what they actually found, it would be something like: “smokers who’d used e-cigarettes in the previous month but had not been able to quit smoking are no more likely to have quit after another six months than smokers who hadn’t used e-cigarettes in the previous month.” Or, translated more bluntly: “smokers shown to struggle to quit in the past month are no more or less likely to quit than those who hadn’t been shown to struggle to quit in the past month.” In short, it shows pretty much nothing.
It’s also worth emphasizing again that the treatment for both groups was the same. It’s no surprise that they had similar quit-rates, because aside from prior experimentation with e-cigarettes (since we don’t know if they continued using them for the studied quit attempt), they received exactly the same intervention.
And of course, lest we forget, the e-cigarette “users” were selected pretty much on the basis of their failure to quit using e-cigarettes. If they’d been successful, they would be ineligible for the study. The data collected in the study shows in several different ways that these people are more severely addicted to smoking, and have greater difficulty quitting overall. A comment on the findings from Peter Hajek, of the University of London, sums it up nicely:
“Like smokers who fail with any method, these were highly dependent smokers who found quitting difficult. The authors concluded that e-cigarette (use) was not helpful, but that would be true for any treatment however effective if only treatment failures were evaluated.”
Conclusion
The five researchers on this study all hold PhDs, and this leaves us with some important questions. Could it be that these well-educated individuals weren’t able to realize that their methods were so hideously flawed? Did they not think to find out how many of the “e-cig users” actually used e-cigs in the studied quit attempt? Or was the decision to focus on those who already failed to use e-cigarettes to quit purposeful? Did they not ask how many intended to vape in their quit attempts because it may have jeopardized the result they wanted to get? It seems eerily reminiscent of a study that looked at callers to quit-lines who’d previously used e-cigarettes, with the same tricks dressed up in a slightly different format. In fact, it seems any reasonable observer would immediately peg this as a politically motivated smear campaign desperately attempting to retain an appearance of scientific objectivity. Yet again, we’re left to wonder whether they give PhDs to the scientifically inept or whether the peer review process doesn’t mind painfully obviously misleading data as long as it fits in with the dominant ideology.