SHS, smoking bans and suspect studies

Another “study” showing a decline in heart attacks due to smoking bans. Another headline, this time from the Boston Globe: “Smoking ban tied to a gain in lives.” Uh-huh. Smoking bans no longer just save lives, they “gain” lives. Of course, since these people were alive before they didn’t die, it can’t honestly be said that their not dying was a gain in lives.

The Globe article refers to a study conducted by the state Department of Public Health and the Harvard School of Public Health. It claims there have been 577 fewer deaths from heart attacks every year since a statewide smoking ban was implemented in mid-2004.

Massachusetts banned smoking in most restaurants, bars, and other workplaces four years ago. Anti-smoker crusaders are determined to show that smoking bans save lives to justify the adverse social and economic consequences which usually follow.

Dr. George Philippides, a cardiologist from Boston Medical Center, had this to say: “Secondhand smoke is an insidious killer. As a cardiologist, I’ve witnessed secondhand smoke’s harmful effects on my patients. This study shows how a public policy to eliminate secondhand smoke exposure can save lives on a large scale.”

Over the past few years, many studies have attempted to link smoking bans with declines in heart attacks. Helena (Montana), Pueblo (Colorado), Bowling Green (Kentucky), Scotland, England, France, etc. All previous studies were shown to be unreliable and were panned by critics and smoking ban supporters alike.

Dr. Michael Siegel, a Boston University School of Public Health expert in tobacco control has been highly critical of these studies. But, Siegel is quoted in the Globe article as saying: "This is the strongest study yet done of the effect of smoking bans on heart attacks. You can no longer argue that these declines would have occurred simply due to medical treatment."

Is Siegel right? Is this study legitimate?

The anti-smoker crowd is claiming the people who never died were people who never smoked, protected from secondhand smoke because of the smoking ban. Ergo, smoking bans save the lives of non-smokers by reducing their exposure to secondhand smoke.

But, anti-smoking activist Siegel says: “Moreover, the observed decline in heart attack mortality is likely due to the sharp reduction in the number of smokers. It likely has little to do with reduced secondhand smoke exposure.” Ergo, smoking bans save the lives of smokers by forcing them to quit.

But how many of those 577 were non-smokers saved from the indignity of exposure to SHS? How many were smokers who saved themselves by quitting? Is there any way to tell?

Actually, there may be a way to estimate the number of deaths among smokers and non-smokers. We can use the same technique used by the anti-smoker brigade, a Smoking Attributable Fraction (SAF).

For example, StatCan notes there were 40,607 deaths due to Ischemic Heart Disease, which includes Acute Myocardial Infarction (AMI or heart attacks), in 2002. Health Canada applies an SAF to the total number of deaths to estimate the number of IHD deaths attributable to secondhand smoke exposure annually.

Although the data from Canada is unlikely to be directly applicable to the US, it should serve to provide a rough estimate as to the number of deaths which could be expected due to heart attacks as a result of exposure to SHS.

Health Canada uses an SAF of 1.6% for males (1.2% for females) multiplied by the total number of deaths from IHD to estimate the number of deaths related to SHS exposure. If the same SAF (we’ll use the higher fraction of 1.6%) were applied to the heart attacks that never happened, you get roughly 10 deaths out of 577 that could be expected due to SHS exposure.

The authors of this study are suggesting that at least a majority of these 577 heart attacks that didn’t happen were non-smokers who might have been exposed to SHS had it not been for the smoking ban? But, the numbers just aren’t there. At least, not without a little fiddling.

But, what about Dr. Siegel’s contention that the precipitous drop in heart attack deaths was the result of “a sharp decline in the number of smokers”? While Siegel’s assertion is the more plausible of the two, it raises a few questions of its own.

The SAF used by Health Canada, if applied to the total number of deaths (577) which never happened, can estimate (supposedly) the percentage of smokers who would otherwise be expected to die of heart attacks. Since details are unavailable as to age or sex, we’ll use the SAF for the group with the largest number of smoking related IHD deaths.

In Canada, this occurs in the population aged between 45 and 59 years of age. The SAF for males is 42.2%; for females, it’s 37.3%. We’ll use the highest percentage across the board since gender has not been established. So, again using Canadian SAFs, we can estimate the number of smokers expected to die at 42.2% of 577 or 244 deaths among smokers.

And, the SAF declines dramatically for those 60 to 69, and again between 70 and 79 until it bottoms out at 8.9% for males and 5.1% for females in those 80 years of age and older. In Canada, over 52% of the deaths due to IHD occur in the population over the age of 80 (mean age 87). This means the estimated deaths attributed to smoking will likely be overstated for this exercise.

Without the ban in place, the 577 deaths that might have occurred would have broken down like this: smokers = 244; nonsmokers exposed to secondhand smoke = 10; and non-smokers never exposed to SHS = 323.

As stated previously, the Canadian SAFs likely cannot be directly applied to the US study. But they do provide a rough estimate as to how many deaths might be apportioned in each category: smoker, non-smoker exposed to SHS and never exposed non-smoker.

But, even these rough estimates raise serious concerns about the conclusion of the study authors that the smoking bans protect the public from death due to SHS exposure. These rough estimates suggest the majority of lives “saved” following the smoking ban would have been never-smokers never exposed to secondhand smoke.

Siegel’s contention, that the reduction was due to smokers who had been coerced into quitting by the smoking ban, is more credible. But, the portion of the 577 who didn’t die from smoking would have been roughly 244, who may, or may not, have escaped death by quitting.

The fact that heart attack deaths (along with smoking prevalence) had been declining steadily for many years prior to the ban, suggests that something other than the ban has contributed significantly to the decline in heart attack mortality.

I suspect this study will prove as inconclusive as the rest in demonstrating a direct correlation been smoking bans and a reduction in heart attacks or heart attack deaths. But, of course, there'll be another day, another study and another opportunity for the anti-smoker crusade to prove smokers are murderous scum.


Count on it.