In researching that last post I spent some time at the web site of TRDRP, California's Tobacco Related Disease Research Program.
Halfway down the front page there's a link: NCTOH 2002 New! Nicotine Products
Now, I know it's the state's job to keep track of this stuff and I don't mind them printing up fact sheets on it. But do they have to use the promotional images? And doesn't that exclamation mark look make it seem that they're a bit too excited about this?
Colby Cosh weighs in on the secondhand smoke study I wrote about earlier.
Taking Colby's points in order, then:
First, there's this:
the [CBC] story doesn't quote or identify very many critics of Enstrom and Kabat. Just the one, actually.
That's just because the CBC sucks, actually. The CBC quotes one of the authors (Kabat), the BMJ's publisher (Smith), and a cancer prof at the University of Toronto. The crosswalk story quotes three critics of the study, one from the American Cancer Society, one from the British Medical Association, and one from an anti-smoking group; and one smoker's rights group, FOREST. Reuters gets all the anti quotes but no pro.
And all of these news stories ignore the primary initial criticism of the study, written by the BMJ's editors.
Back to Colby, then.
Ferrence is making the extraordinary suggestion that the home is not the natural first place to look for ETS risk.
Hardly extraordinary. In fact this point is made succinctly in one of the rapid responses to this article. The place to look for ETS risk is where the subject spends time: at work or at home. Sadly, many people do not spend most of their waking time in the company of their spouse. So why is spousal smoking accepted as a proxy for ETS exposure?
Well, Enstrom and Kabat accept it because their 1999 questionnaire asked respondents to categorize their exposure to ETS, and the respondents' self-reports were correlated with the smoking status of their spouse. (Tables 4 and 5 of the study.) I am very skeptical of this result -- especially without having seen the survey instrument. First, there is self-selection in respondents. Then we're relying on respondents' self-reports of historical ETS exposure to determine if this correlation exists. And without seeing the survey, I can't tell if the question used to determine "Regular exposure to cigarette smoke from others in work or daily life" -- the linchpin of the argument -- was sufficiently clear in excluding the exposure from the spouse. So I'm not convinced they have a good proxy variable.
And ultimately, the real problem with the paper lies in the conclusion. First there's the part where spousal smoking is not named as a proxy variable: "The results do not support a causal relation between environmental tobacco smoke and tobacco related mortality". Then there's the segmentation of the results. If you divide the survey population into many small groups, then you reduce the power thus increasing the confidence interval and making it more likely for the result to be null. As the BMJ editor observes:
They may overemphasise the negative nature of their findings. With respect to chronic obstructive pulmonary disease—plausibly related to exposure to environmental tobacco smoke— the estimates based on the most accurately classified exposure groups give relative risks of 1.80 in men and 1.57 in women. These are said to be non-significant, but combining them—and there is no good evidence that exposure to environmental tobacco smoke has a different effect for men and women—gives a relative risk of 1.65 (95% confidence interval 1.0 to 2.73)
Colby finishes up with:
Both men, as I understand it, do believe there is a microscopic risk of harm from ETS; it's just way, way too small to be measured--or, for public policy purposes, to be considered at all.I actually agree with this. Previously published findings of ETS risk are typically based on meta-analysis and show a small risk. A good prospective study aimed particularly at ETS exposure needs to be done to conclusively answer the question: does ETS exposure increase mortality?
Colby continues:
If you were absolutely determined to look for measurable ETS risk, you would do it in exactly the way they have done it. Assuming you couldn't initiate a massive prospective cohort study, you would--as they did--go back and see what a previously existing study of that sort can tell us. Or do critics of the Enstrom and Kabat study have a better idea?
If it were up to me, and I had bottomless funds, I would do the followup on the entire CPS I population, not just those living in California. The results would be more representative of the US population, and the study might well have greater statistical power. It would also avoid the accusation that I had gone over the data beforehand to isolate the subset that agreed with my preformed hypothesis. (Why California and not New York, after all? Probably because of the Prop 99 funding in California.) I would not introduce new questions asking respondents to self-assess ETS exposure, because I don't think a 40-year retrospective self-assessment and self-report is useful.
But ultimately, it doesn't matter if there is an actual public-health risk from ETS. If enough people find secondhand smoke unpleasant, public smoking will be banned even if it's not dangerous.