Tuesday, 13 April 2010

Junk from Canada and a surprising admission

Public smoking bans seem to have paid off in fewer hospital admissions for heart and lung problems, a Canadian study suggests.

The study looked for any effects of Toronto's 2001 ban on smoking in restaurants, aimed at reducing exposure to second-hand smoke, which the researchers say is a major factor in preventable poor health and premature death.

In Monday's issue of the Canadian Medical Association Journal, the researchers reported the following changes since the ban took effect:
  • 17 per cent decrease in the heart attack hospitalization rate

Ah, the magical 17%! As reported by the infamous Jill Pell, and as reported by the still-more infamous Stanton Glantz. With such remarkable consistency, perhaps we should see it as evidence of a genuine effect? Alas, no, since Pell's fictitious 17% was for 10 months, Glantz's was for all sorts of time periods and this new study's was for 8 years. Quite the opposite of consistent, really.

Michael Siegel explains why there are "huge problems with this study." His critique is recommended reading for anyone who needs convincing that these studies—every single one of them—rely on appalling methodology and sleight of hand. In particular, Siegel points to some huge drops in heart attack admissions in the places without smoking bans, and shows how the researchers buried this evidence.

An editorial about the study appears in the journal, written by British public health demagogue Alan Maryon-Davis. Maryon-Davis is, I'm afraid, a classic busybody, who describes himself as a libertarian despite calling for extraordinarily illiberal policies in the name of health. His views aren't worth paying the $20 the Canadian Medical Journal wants for them, but according to CBC:

In a commentary that accompanies the study, Prof. Alan Maryon-Davis of Kings College London, United Kingdom, argued for comprehensive cost-benefit analyses to weigh the potential health benefits of anti-smoking legislation against infringements on personal liberty and effects on jobs and livelihoods.

What would be the point? We already know that tobacco control flatly denies that smoking bans cost jobs, and we know that they value public health over personal liberty. And if the 'benefits' of smoking bans include these ludicrous ecological studies, to which no serious epidemiologist would put their name, we can be sure it wouldn't be worth the paper it's written on. Why pay these people to conduct another rubber stamping exercise when, as with the so-called review of the UK smoking ban, the result is preordained?

A better idea would be to get some serious, unbiased statisticians to look at heart attack data from around the world and see if there is any real correlation with smoking bans. After they've shown that there is no correlation (as they surely will, if data from the UK, America, Australia, New Zealand and several other countries are any indication), let's launch an investigation into how a small cabal of quackademics have engaged in what amounts to scientific fraud. How about that, Alan?

Speaking of junk science, the CMAJ carries another smoking-related story which has been buried under the media coverage dedicated to the latest heart attack fairytale. Last month I attempted, not very successfully, to convince BBC Scotland that Prof John Britton was—for want of a better word—lying when he said that secondhand smoke in a car was 23 times more dense than in a smoky bar. In the CMAJ, Ross MacKenzie and Becky Freeman have now exposed that lie:

There is no evidence to support the claim that smoking in cars is 23 times more toxic than in other indoor environments, according to a report in the Canadian Medical Association Journal.

Ross MacKenzie and Becky Freeman, from the University of Sydney, have criticised the 'unsubstantiated' figure and plotted its path through both the mainstream press and scientific publications before become widely accepted as 'fact'.

Kim Barnhardt, of the Canadian Medical Association Journal, said: 'There is no evidence to support the fact that smoking in cars is 23 times more toxic than in other indoor environments.'

As I have mentioned before, this "fact" has no basis in science and originated from Chinese whispers and ASH press releases. If it has a source at all, it seems to be a news report in that esteemed medical journal Rocky Mountain News.

'Successful advocacy campaigns often require the translation of complex research findings into short and memorable media quotes,' the report states.

That's being very generous. The '23 times higher' is, quite simply, a lie and always has been so. It would be perfect material for More or Less or Ben Goldacre, but don't hold your breath. Goldacre, in particular, seems able to turn a blind eye to bad science if it's in the name of a 'good cause'.

Finally, a plea. I've read the new heart miracle study, but if anyone has Maryon-Davis' article or the car study, I'd be interested to read them in full: chrissnowdon@yahoo.co.uk. Thanks.


Man with Many Chins said...

Dunno about you Chris, but I am started to get really pissed off at the breathless way the MSM reports this junk science without even the most basic investigation being carried out on the claims made.

Paul said...

I've noticed the same tendency with Goldacre. Its too bad.

Anonymous said...

someone - with some money! - needs to launch libel proceedings against one of these pseudo science by press release stories..


Fredrik Eich said...

The 23 times number that has made it into this RSPH manifesto as discussed along with a transfat ban in the bmj.