Wednesday, June 18, 2008

Media Reports Decline of Up to 40% in Heart Attack Admissions in England Due to Smoking Ban; Three Anti-Smoking and Health Groups Support the Claim

The media is widely reporting that there has been a drop in heart attack admissions of up to 40% in England due to the implementation of the national smoking ban last July (report 1; report 2; report 3).

According to these three articles, the smoking ban resulted in a decline of up to 40% in hospital heart attack admissions during the first nine months that the ban was in place. The articles report that overall, the number of heart attack patients being admitted has fallen in more than half of England's hospital trusts. Specifically, there was a decline in heart attack admissions in 66 of the 114 trusts examined.

The total number of heart attack admissions during the nine-month period following implementation of the smoking ban was 1,384 less than during the same nine-month period a year earlier. One hospital trust saw a 41% decline in heart attack admissions.

Three anti-smoking or health groups - the British Heart Foundation, the British Cardiovascular Society, and Action on Smoking and Health - were quoted in these articles as concluding that the observed decline in heart attacks was attributable to the smoking ban.

According to the articles:

"The British Heart Foundation said that it showed the ban was the 'most significant public health initiative this century'".

"Amanda Sandford, of the pressure group Action on Smoking and Health, added: 'This is excellent news. It seems likely that the drop in hospital admissions for heart attacks is linked to the implementation of the smoking ban. It shows just how quickly the benefits can be felt. Even if the overall percentage reduction appears small, the fact that this amounts to over a thousand people whose lives have been saved is extremely important.'"

"Dr. Nicholas Boon, president of the British Cardiovascular Society, said: 'This is great news. It is exactly what we hoped and expected to see. When you place these figures with the research in Scotland, Ireland, France and Rome, it is consistent with the observation that the ban has been followed by significant improvements in heart attack rates. It is early days, but the benefits may be greater in the long run.'"

The Rest of the Story

This is pure junk science, and it is a shame that the anti-smoking and health groups are willing to stoop down to the level of junk science to promote their agendas.

The fact that heart attacks have declined in more than half of the hospital trusts is of no consequence at all with respect to the hypothesis that the smoking ban led to a reduction in heart attack admissions. Even if one assumes that there is no secular change occurring in heart attacks, under the null hypothesis - that the smoking ban had no effect on heart attacks - one would expect that heart attacks would decline in one-half of the hospital trusts. That is, if smoking bans have no effect on heart attacks, then one would expect that heart attacks would decline in 57 of the 114 hospital trusts. The fact that heart attacks declined in just 66 of the hospital trusts is actually pretty strong evidence that the smoking ban did not cause a decline in heart attacks.

In fact, we can quantify the probability that if smoking bans have no effect on heart attacks, one would observe a decline in heart attacks in 57 of 114 hospital trusts. It turns out that this probability is greater than 5%, the level generally considered statistically significant. Thus, the reported finding is not significantly different than one would expect by chance alone.

The actual 95% confidence interval on the reported proportion of hospital trusts that saw a decline in heart attacks is 48.8% to 67.0%. Since this confidence interval includes 50%, one cannot conclude that the proportion of hospitals that saw a decline in heart attacks is different from what would have been observed by chance alone (if there were no effect of smoking bans on heart attack admissions). Thus, these data provide no evidence of any effect of the smoking ban on heart attack admissions.

Moreover, the overall reported decline in heart attacks is only 3%. This is such a small effect that there is no way to attribute the decline to the smoking ban.

Furthermore, it is well-documented that heart attack rates in Europe have been declining over time in recent years. It is unlikely that the 3% observed decline in heart attacks is significantly higher than the decline that would have occurred anyway due to secular trends. Without any analysis of the trends in heart attack admissions over time, it is impossible to draw the conclusion that the 3% decline was due to the smoking ban, rather than merely reflective of a secular trend.

Even worse, there is no control group in the study. Even if the 3% decline in heart attacks in England were significant, without a control or comparison groups, it is impossible to know that the decline is greater than what would have been expected in the absence of the smoking ban.

It appears that the sloppy science being promulgated by anti-smoking groups and researchers in the area of the evaluation of the effects of smoking bans on heart attacks has become so pervasive that it has infiltrated into the media itself. Now the media are conducting their own shoddy analyses and representing them as having some meaning.

Let me assure my readers that this particular analysis is meaningless. If anything, the data best support the conclusion that the smoking ban had no substantial effect on heart attack admissions. A 40% decline in admissions in Helena was attributed to the smoking ban by anti-smoking researchers. as was a 27% decline in Pueblo. Even if the observed decline in smoking was due to the smoking ban rather than a secular change or random variation, the finding of just a 3% decline would wipe out the conclusion that smoking bans have a dramatic effect on heart attack admissions.

What is disturbing to me is not that the media would put forth such a shoddy scientific analysis and unsupportable conclusion. What disturbs me is that anti-smoking groups find this shoddy science to be convincing. They are so biased in their views that they apparently care only about the direction of the results, not the scientific validity of the findings.

This is a sad state of affairs for the tobacco control movement, because it means that we are continuing to lose our scientific integrity. The political cause has become more important than the science.

Unfortunately, this makes us no better than the tobacco companies that we have consistently criticized for their own shoddy science. And maybe even worse, because the tobacco companies have at least made some changes in their representation of science to the public. They are at least moving somewhat in the right direction. But we are apparently moving in the wrong direction.

We need to take the high road and to protect the scientific integrity of our movement. In the long run, it does no good to stoop to the level of junk science to support our agenda. It puts us on the road to the eventual loss of credibility and of the public's trust.

No comments: