Thursday, March 18, 2010

Another Published Study Finds No Evidence that Smoking Ban Resulted in Short-Term Decline in Heart Attacks

Negative Finding Doesn't Stop Researchers from Telling the Media that They Found an Effect

Another published study - this one in the Australian and New Zealand Journal of Public Health - fails to provide evidence that smoking bans result in dramatic, immediate reductions in heart attacks, as claimed by many anti-smoking researchers and groups and as claimed by an Institute of Medicine committee that reviewed the scientific evidence (see: Barnett R, Pearce J, Moon G, Elliott J, Barnett P. Assessing the effects of the introduction of the New Zealand Smokefree Environment Act 2003 on acute myocardial infarction hospital admissions in Christchurch, New Zealand. Australian and New Zealand Journal of Public Health 2009; 33:515-520).

The study examined the effects of New Zealand's smoking ban, implemented in December 2004, on first-time acute myocardial infarction (i.e., heart attack) admissions to Christchurch Hospital during the two years before and after the smoking ban. Thus, the baseline period was January 2003 through December 2004 and the intervention period was January 2005 through December 2006.

The researchers found that the overall rate of heart attack admissions declined by 5.1% from 2003/2004 to 2005/2006. They did not include any comparison group nor go further back in time to assess whether this level of decline was similar to pre-existing secular trends or to trends in other countries.

When examining the raw data (see Figure 1 in the paper), it is quite clear that there is no significant effect of the smoking ban on heart attack admissions. In fact, the number of heart attacks during the winter of 2006 was higher than during any other winter during the study period.

The graph also suggests that there has been a slight increase in heart attacks during 2006. The number of heart attacks was higher in 2006 than in 2005 during the last eight months of that year.

The paper accurately concludes that the evidence provided does not allow one to conclude that the smoking ban resulted in a significant decline in heart attacks. As noted in the paper: "it is possible that the observed decline in AMI admissions is a manifestation of long-term secular trends, for example in statin prescription, reduced excess winter mortality or dietary change, that have operated synergistically alongside the national smokefree policy introduced in 2004."

In other words, in the absence of any control group or any comparison to existing secular trends, it is not possible to attribute the small observed decline in heart attacks to the smoking ban. The paper's authors acknowledge this in the manuscript and because of this weakness, conclude that the present evidence does not support a conclusion that the smoking ban led to a significant reduction in heart attacks.

Nevertheless, one of the study authors apparently told the media: "This short-term research indicates a link between a smoking ban in bars and restaurants and a reduction in severe heart attacks." The media certainly got the impression that the researchers were concluding that the smoking ban led to a decline in heart attacks.

Finally, despite their conclusion within the paper that the evidence is not sufficient to conclude that the smoking ban had any effect on heart attacks, the abstract section of the paper states: "At this early stage following the smokefree legislation, there are hints emerging of a positive impact on AMI admissions...".

Also, in the discussion section of the paper, the authors state: "Overall, our results provide only
limited indication that the introduction of the legislation was associated with a reduction in AMI admissions." [emphasis is mine]

The Rest of the Story

This is now the second published study which fails to find a significant effect of a smoking ban on heart attack admissions. Remember that a much larger study, which included an analysis of all heart attack admissions in the entire country of New Zealand, found no evidence for any effect of the smoking ban on heart attacks. In that study, the implementation of the ban was associated with a slight increase in heart attacks (which seems to be the case during the second year of the smoking ban in Christchurch).

The results of this study are consistent with those of the first, larger study, which studied all hospitals in New Zealand and found no effect of the smoking ban on heart attacks during the first year of the ban. This study examined only Christchurch and while there was a slight decline in heart attacks, the rate of heart attacks actually increased during the second year post-ban.

Besides adding evidence that refutes the conclusion disseminated by anti-smoking groups and by the IOM Committee which reviewed this issue, this story is also important because it shows how biased the reporting of science within the tobacco control community has become. Here we have a study where the paper itself asserts that no causal conclusion can be drawn, yet the media were apparently told that the study showed a reduction in severe heart attacks due to the smoking ban.

Why bother doing the research if you are going to tell the media the same thing no matter how the study results come out?

In addition, the study itself uses misleading, inappropriate, and non-scientific language in describing its findings.

First, it states its conclusion as follows: "there are hints emerging of a positive impact on AMI admissions."

To translate into plain English, this means: "We failed to find the evidence of a positive impact on AMI admissions that we were hoping for. However, we don't want to say that there was no effect found. So instead, we'll weasel our way out by stating that there are hints of an emerging positive impact ... in other words, an impact that hasn't yet been shown, but it is due to emerge at some point since we believe in it."

Second, the paper also states its conclusion as: "Overall, our results provide only limited indication that the introduction of the legislation was associated with a reduction in AMI admissions."

To translate that into plain English, it means: "our results actually don't indicate that the legislation reduced heart attacks. But we don't want to state that, so we'll call it a "limited" indication."

Why can't researchers writing about this issue of smoking bans and heart attacks simply tell it like it is? Why do they have to use non-scientific weasel language to avoid what is apparently a pre-determined conclusion? And how does this get past the peer review of the journal?

No comments: