Smoking ban miracle in Germany

Readers of this story recognise the holes in it. Researchers found a city where heart attack admissions have dropped following the introduction of a smoking ban, and quickly decided that it would help in building the case for a smoking ban, regardless of any other factors, such as the comprehensiveness of the ban and whether or not it did actually result in a reduction of so-called passive smoke exposure (rather than simply ensuring that people were exposed at home rather than in recreational venues).

The story says that the drop in heart attacks in Bremen was a 25 per cent drop and compares it to the drop of 2.4 per cent that the Department of Health found in England. A ten-fold disparity (as well as the huge disparities in population) in the figures must surely tell them something! but the figures of various heart attack drops are rattled off as if they proved, rather than cast a heavy shadow over the theory that smoking bans stop heart attack admissions.

A more accurate measure would have involved a larger population, across the whole of Germany. Why pick Bremen anyway – simply because the figures could be made to show a large drop? Small populations are more likely to produce startling variances for a wide variety of reasons.

An example of a study seeking to cover a wider population base is here. No statistically significant drops in heart attacks were found when studies were conducted over wider population bases using publicly available data on admissions, and employing appropriate controls.

Fact checking charity cautiously debunk's Penny's 26 per cent claims

Among the more outrageous things that was claimed in this week's fanfare commemorating five years since the smoking ban was implemented in England was Laurie Penny's claim in the Independent that heart attack admissions had dropped 'as much as' 26 per cent since the ban started.  Even Linda Bauld could only come up with a figure of 2.3 per cent for heart reductions in England (a figure too small to show that it has a single cause): 26 per cent was a figure that demands explanation.

The report by independent fact checking organisation Full Fact even provides an illustration:

Source

This should show without any doubt that there is no appreciable difference in heart attack rates in relation to the smoking ban. The reporter declares herself at a loss to understand where Penny's figures have come from (a request from Full Fact had no reply at the time of writing). She then discusses various other heart attack reports, all of them small ('five US cites', etc.) giving results with smoking ban declines of up to 50 per cent, and mentions Linda Bauld's study linked above from the University of Bath. (No evidence is given that larger-scale studies convincingly undermine the smaller ones, by including cities where no drop heart attacks was recorded following the introduction of a smoking ban.)

The conclusion is stated with moderation:

It is very hard to assess the accuracy of Ms Penny’s claim without first knowing her source. The research we have found from the NHS Information Centre shows at most a 22 per cent decline, although this only refers to fatalities. Among all incidents of heart attacks the reduction was a more modest 13 per cent. [emphasis added]

Ms Penny's claim is so far from the available evidence that it is hard to escape the conclusion that she has not researched the material properly. This article does properly conclude: 'however it is noticeable that trends don't seem to have changed much since 2002', but then we read:  

This suggests that there may be wider social and environmental factors that have also contributed to the fall, and singling out the impact of the smoking ban is a difficult process. 

Suggests? how about 'shows categorically', Full Fact?

Still, look on the bright side. Yet another body has acknowledged that the 'heart attack miracle' is an unsustainable fiction. 

Liquor license group challenges Health Director Wymyslow, Ohio

The Buckeye Liquor Permit Holders' Association issued this press release yesterday. They state that a new Executive Summary Report published by the Ohio Department of Health misrepresents the main study, 'Analysis of the Impact of Ohio's Smokefree Workplace Act', by failing to refer to vital data presented in the report. It was not reported that fewer than 1 per cent of smokers increased their patronage of bars after the smoking ban came in. The fact that venues in the lowest tax bracket were excluded from the study was also left out of the summary report. The group also points to the claim in Ohio Health Department's report that heart attacks dropped drastically (by 26 per cent) following the smoking ban, referring to Michael Siegel, smoking ban advocate of 25 years' standing, who shows that there were more heart attacks after the smoking ban than before it.

The group also questions the appointment of employees of the Public Health and Health Promotion departments of the administration to write an economic study, when their career interests are at stake in the outcome.  The omission of information in the Executive Summary concerning bars specifically is achieved by collating information for the entire hospitality sector, which conceals damage to bars. Leaving out establishments in the low tax bracket enhances the focus on bars and restaurants in areas less affected by the ban (in districts with fewer smokers). (Both these issues are covered in an enlightening 2008 paper by Michael Marlow entitled 'Honestly, who else would fund such research?')

Opponents of Ohio Bans have today gone even further and called for the resignation of the Health Director Dr Theodore Wymyslow, on the grounds that the Executive Summary Report gives misleading and essentially fraudulent information to the press – both broadcast and print media. This report has been released one month before a debate on the constitutionality of the smoking ban is heard in the Ohio Supreme Court and the Ohio group holds that this is a deliberate attempt to influence the outcome of this debate.

The methodology of the Ohio report rings bells. Professor Linda Bauld, a professor in Socio-management recently appointed to Stirling University, published a significant report on the economics of the smoking ban in March: The Impact of Smokefree Legislation in England: Evidence Review. It is curious that the public accepts as impartial and authoritative economic reports about smoking bans that are written by researchers in public health departments, who in general lack insight into the economics of the hospitality industry and are swayed by the requirements of their paymasters.

We should take a leaf from Michael Marlow's paper and remember the 'rhetoric of assent' referred to in Michael Marlow's paper: 'that is an initial readiness to suppose that the other guy or gal is also fairminded and trustworthy and deserves to be heard out'. Unfortunately the influence of the Framework Convention on Tobacco Control makes this very difficult, as the FCTC works to resist 'tobacco industry interference' (interpreted as widely as they can) in public policy (Article 5.3). It is hard not to dismiss the insights of the other camp when they are so determined to dismiss any take on the issue that doesn't emanate from public health.

I look forward to an update on Ohio.

Selecting studies to scare smokers

Here's another one. Your risk of AF (atrial fibrillation) increases when you smoke.

Regardless of the fact that most people, probably correctly, perceive smoking as a health threat, new stories about how bad tobacco is come out daily.

But we know that. Stories about smoking roll off a production line, because there is a ready market for them. We also know that the mass market exists for stories that at least appear to give a result that makes tobacco appear more dangerous. We know that studies are censored at birth.

The same goes for secondary smoke. The diagram below comes from this story on heart attacks – the heart attack miracle that occurred in the year following the implementation of smoking bans everywhere from Scotland to Montana. Except (1) there is no evidence of any follow-up studies to show a sustained effect, (2) the figures frequently disagreed with routine statistics and (3) only the localities that did show a drop in heart attacks featured in published studies. The figure shows the range of results from all localities with comprehensive smoking bans, and the results clearly fail to show that smoking bans always result in a drop in heart attacks.

Chris's post shows more detail. We can be sure that for every study showing us the dangers of smoking and secondary smoking, there are studies or potential studies with results that anti-smokers would prefer us not to know.

This does not mean that there is no risk attached to smoking, of course – just that we are not told of any benefits (although nicotine has been appropriated by the pharmaceutical industry and so is regarded as having benefits), and we may be misled as to the scale of the risk. The study of AF linked above, for example, shows risk doubling, i.e. a risk factor of 2 – still a low risk.