Monday, June 23, 2008

Heart Attack Trend Data in England Show that Anti-Smoking Groups' Claims of a Decline Due to Smoking Ban is Junk Science

As I discussed here last week, three anti-smoking and health groups in England endorsed a recent study - reported by the media - showing as much as a 40% decline in heart attacks in some hospital districts following the implementation of a nationwide smoking ban. These groups told the public that this study supports the conclusion that the observed decline in heart attacks was due to the smoking ban. Based on the study, one group even called the ban "the most significant public health initiative this century."

The Rest of the Story

Data on trends in hospital heart attack admissions in England posted by Colin Grainger on his Freedom to Choose web site do not support the anti-smoking groups' claims. These data show that there is wide random variation in heart attack admissions in England, with swings of up to 5.4% from year to year. From 2002-2003 to 2003-2004, heart attack admissions fell by 2.4% in the absence of a smoking ban. Clearly, one cannot take the observed change of about 3% in the first nine months following the smoking ban and attribute it to the smoking ban. It could easily represent random variation in the data. In addition, since heart attack admissions have declined the past two years, it is also possible that the observed decline simply reflects a secular trend of declining heart disease morbidity.

Moreover, since these data represent only nine months of the year, it is possible that after a full year's data are available, the decline in heart attack admissions will not be as great as it now appears. But even if the decline is truly 3%, there is no way to conclude that such a small change is due to the smoking ban, as opposed to random variation and/or secular changes in heart attacks that would have occurred anyway, even without the smoking ban.

Since anti-smoking groups in the U.S. and elsewhere are arguing that smoking bans cause immediate, dramatic declines in heart attacks - on the order of 15% to 40% - it would seem that these data from England actually invalidate the conclusions from these prior studies. It is amazing to me how the anti-smoking groups can twist and distort the data to support a conclusion that this study is consistent with these prior results that show a dramatic decline in heart attacks following smoking bans.

The problem with the anti-smoking groups spinning the data in this way is that it exposes them. If they are going to conclude that smoking bans have a dramatic effect on heart attacks no matter what the data show, then their conclusions are obviously not based on science, but purely on ideology. The agenda is dictating their "scientific" conclusions, rather than their science dictating their public statements. Unfortunately, they appear to have things backwards. The science is supposed to guide the agenda, not the reverse.

No comments: