Tuesday, June 24, 2008

New Study Concludes that Smoker's Breath is Harmful to Children; Kids at Risk Even if Parents Smoke Outside of the Home

A new study published in this month's issue of Indoor Air concludes that even when parents only smoke outside of the home, toxic smoke constituents in their breath pose a major health risk to their children indoors. This conclusion was based on the finding that: "There were significant
differences in the median levels of air nicotine and PM10 between households in which smoking was reported as only occurring outside, and the smoke-free households" (see: Rumchev K, Jamrozik K, Stick S, Spickett S. How free of tobacco smoke are 'smoke-free' homes? Indoor Air 2008; 18: 202-208).

The study compared indoor air levels of respirable suspended particulates (RSP) and nicotine under three conditions: (1) smokers smoke inside the home; (2) smokers only smoke outside the home; and (3) no smoking at all. With smokers smoking inside the home, the average nicotine level inside the home was 1.4 ug/m3. With no smoking at all, the average nicotine level was below 0.2 ug/m3. With smokers smoking outside the home only, the average nicotine level was 0.55 ug/m3.

In response to the study, Action on Smoking and Health (ASH) is promoting policies by which employers fire all existing smokers or refuse to hire smokers, even if they smoke only outside of work, in order to protect nonsmoking employees from the toxic breath of the smoking employees. ASH is also arguing that smokers should not be allowed to adopt children, even if they agree to smoke only outside of the home, because their breath will expose children to unsafe levels of tobacco smoke. Moreover, ASH argues that before children whose parents are separated visit a parent who smokes, that parent should be required to change clothes and rinse with mouthwash.

ASH states: "This study suggests that society must go beyond merely protecting children from being in the presence of parents and others who smoke in their homes, and think about more effective measures to protect children from parents who smoke anywhere." ...

"It also provides a strong scientific basis for agencies which already refuse to permit smokers to adopt children, even if the potential adoptees claim that they only smoke outdoors, and never in the presence of the child." ...

"Judges in almost three-fourths of the states have issued orders prohibiting smoking in homes to protect children involved in custody disputes. But this study would permit the parent who obtained the order to go back and strengthen it, perhaps requiring the smoking parent to change clothing and use a mouthwash before the child visits, predicts attorney Banzhaf, who has helped nonsmokers obtain some of these court orders." ...

"We've always known that a smoker's breath stinks. Now we know that it also creates indoor air pollution which can harm children and perhaps some adults."

The Rest of the Story

The rest of the story is that there is a fatal flaw in the study which invalidates the conclusion: it is highly likely that some smokers who claimed only to smoke outside the home actually do smoke in the home, at least on occasion. This would have completely explained the study result: that levels of smoke in homes with smokers who claim to only smoke outdoors are intermediate between levels in a smoke-free home and levels in a home with smokers who admit smoking inside the home.

It is likely that some smokers are not being honest about smoking inside the home because very clearly there is a social pressure not to smoke in the home. With all of the publicity - much of it due to ASH itself - about how smoking around kids is child abuse and how smokers are child abusers, it is not surprising that smokers would be hesitant to admit that they do smoke indoors with children present. Thus, the hypothesis that smoke smokers are not being honest about smoking inside the home is quite plausible.

Importantly, this hypothesis would completely explain the observed findings. Not all smokers are being dishonest, just a proportion of them. Thus, the smoke levels among those who claim to smoke only outdoors would be intermediate between levels in smoke-free homes and levels in homes where the smoker admits to smoking indoors. This alternative hypothesis is entirely consistent with the observed findings of this study. Thus, unless the authors can disprove this hypothesis or show why it is implausible or unreasonable, the study conclusion is not valid.

The authors, however, do not provide any evidence as to why this hypothesis is not plausible. In fact, they reject this hypothesis without a word of explanation. They point out that the results could be due to a true effect of outside smoking on smoke levels in a home or to smokers not telling the truth and then they simply conclude that the former is the correct explanation. But they provide no evidence or even reasoning to suggest why the latter explanation is unlikely to be correct.

I find this to be a striking bias in the article (not atypical of the bias I am seeing these days in many articles in the tobacco control literature). You have two possible explanations for the study findings. Rather than objectively evaluating the evidence to decide which hypothesis is likely to be correct, you instead simply reject the least favorable explanation. That is essentially what is going on here.

Of note, only 4 of the 39 households in which there was a smoker reported that smoking occurred inside the home. On its face, this seems too good to be true and it should have alerted the investigators to the likely possibility that they were not obtaining accurate information about the smoking locations in that household.

This study is plagued by a fatal limitation: it is unable to objectively determine whether smoking occurred in the home or not. Normally, one might not make too much of such a limitation. But because the primary purpose of the study was to determine whether outdoor smoking can result in indoor exposure, it is inexcusable not to develop a means to validate the actual locations where smoking occurred in these households. Or, alternatively, to admit in the study discussion that dishonesty about smoking locations could be an alternative explanation for the study results.

There are four important implications to this story.

First, it makes it clear that this study cannot be used as evidence that outdoor smoking results in substantial secondhand smoke exposure for children inside the home. Action on Smoking and Health has either not taken the care to analyze the validity of the study or else it is so excited about the opportunity to further lambast smokers that it ignored the lack of validity of the study findings.

Second, it reveals the increasing investigator bias in tobacco control articles these days. This is a disturbing trend to me. It suggests that the peer review system is not working particularly well for tobacco control. I think perhaps part of the problem is that the articles are being send to the same cadre of reviewers, all of whom are sharing the same bias towards favorable results.

Third, it has unfortunate implications for children. If tobacco control groups follow ASH's lead and tell the public that smoking outside is of little help to protect kids, then many smokers may well decide that it is not worth taking the trouble to smoke outside and the exposure of children may increase, rather than decrease.

Fourth, and most disturbingly, the misinterpretation of this study's results may lead to efforts to bar smokers from the workplace and to prevent smokers from adopting children, both of which would be tragic mistakes.

No comments: