Denialism: The New Heresy?

(Download this article as a PDF. It makes it so much easier to read.)

In an five-page article published online in The Expert Review of Pharmacoeconomics & Outcomes Research, Martin Dockrell, the Policy and Campaigns Manager for Action on Smoking Health, has launched an extraordinary attack on the journalist and broadcaster Michael Blastland. Calling him a "conspiracy theorist" and a "dissident", Dockrell explicitly compares Blastland to the 'AIDS dissidents' who disputed the link between HIV and AIDS.

The denialist

Blastland's crime was to criticise a study which claimed that incidence of acute coronary syndrome fell by 17% in Scotland after the smoking ban (Pell, 2008) and, therefore - post hoc ergo propter hoc - because of the smoking ban. Blastland covered the story for the BBC in November 2007, two months after the findings were reported by the international media following a presentation at a tobacco control conference.

Since the paper was as-yet unpublished, Dockrell accuses Blastland of "reject[ing] the research before they had had the opportunity to look at it."

It is notable that Dockrell does not criticise any other journalists for unquestioningly reporting a study that they had not yet read. Nor does he criticise the Scottish Government for producing a press release to promote the findings, nor Jill Pell herself for announcing a study that had been neither peer-reviewed nor published. Nor, for that matter, does he criticise his own boss at Action on Smoking and Health - Deborah Arnott - for greeting the news, saying: "We knew from epidemiological statistics there was a risk from secondhand smoke to cardiovascular health but not how much of a risk until now."

For Dockrell, it seems, blind faith is the only acceptable response to an unpublished study. Scepticism is not.

This argument would be slightly stronger if Blastland had had no idea of what the study said and had no evidence with which to counter it. In fact, he had plenty of both. Based on what Pell had told the media, he knew that the study had claimed a 17% drop in heart attacks and, using official data from the Scottish NHS, he could see that no such fall had occurred:

"These [data] show a fall in heart attacks for the year from March 2006 - not of 17%, but less than half as much at about 8%. What's more, taking out the recent trend, this is halved again. Heart attacks have been falling steadily for some years now."

This was the essence of Blastland critique. Using official data, rather than the case group selected by Pell, it was plainly obvious that the fall in 2006-07 was an unexceptional extension of an existing downward trend.

Attacking the official figures

One of the most puzzling things about the Pell study was why the team chose to use a sample group when hospital admission data was freely available. Dockrell comes to her defence, saying:

"Pell makes no secret of the difference between the data from the StopIt study and the routine discharge data...Although Pell and colleagues had access to the AMI [acute myocardial infarction] discharge data, the StopIt study refers to ACS, [acute coronary syndrom] a broader measure for heart attacks, verified by assay."

This is true, but Dockrell fails to mention that Pell had access to the ACS discharge data too. Everybody does - it is freely available on the ISD Scotland website - and it shows that ACS admissions fell by 7.2% in the first year after the smoking ban and rose by 7.8% in the second year. In a bold attempt to ignore the elephant in the room, Dockrell does not mention the rise in admissions in year two, despite the fact that he cannot have avoided reading about this while conducting his research.

Later in the article, Dockrell points out that "raw discharge data" from hospitals is "not peer-reviewed" and suggests that it is not, therefore, reliable. This is not only a ridiculous assertion, it is also a slur on those who compile them. The NHS employs professional statisticians to compile and publish this data and my own experience with them has shown them to be exceptionally cautious and thorough (which is why the data takes so long to be published). There is no requirement at all for these figures to be peer-reviewed by a medical journal. They are official statistics, reviewed meticulously before being published.

It is hard to believe that Dockrell is seriously suggesting that Pell's twenty-month study of a patients in a selection of hospitals trumps ten years of data from the Scottish NHS, but that does indeed seem to be what he is saying.

Warm weather

In his BBC article, Blastland also pointed out that the hospital discharge data showed a large drop in admissions in 1999-2000 (of around 11%), thereby showing that relatively large percentage drops occurred before the smoking ban came into effect. Dockrell suggests that this was due to an unusually warm winter:

"Neither Pell nor Blastland offer any explanation for the drop in 1999-2000, despite that winter being described as "the sunniest on record", with monthly average temperatures in Scotland from January to March 1.5-2C warmer than normal."

This is a valid and interesting comment, since cold weather is associated with high incidence of heart attacks. Blastland, however, had not given any explanation of his own as to why the rate fell in 1999-2000; his point was only that it was possible to see large drops with or without smokefree legislation.

But having raised the issue of warm weather, Dockrell does not follow it to its logical conclusion. Had he done so, he might have found that a similar explanation for the drop of 2006:

"The UK has had its hottest summer since records began 234 years ago in 1772. Sizzling July and August were followed by a hot September, which saw temperatures for that month climb to an average of 15.4C."
(The Daily Mirror, Oct. )

"Unusually large numbers of dolphins and whales have been sighted off the eastern coast of Scotland this summer, including several species that are rare in these waters. Scientists believe the sightings may be the result of the heat wave currently gripping the U.K."
(National Geographic, June 2006)

Unwittingly, then, Dockrell may have made a useful contribution to a debate that he claims is "over".

The new heresy

Dockrell displays a more unpleasant side when he equates criticism of Pell's study with "AIDS dissidents" and other assorted cranks. Ad hominem attacks are, sadly, often used against those who question any aspect of passive smoking epidemiology. The typical modus operandi is to accuse critics of being employees or "allies" of the tobacco industry. Such claims are usually untrue and are clearly false in the case of Michael Blastland.

Blastland writes and broadcasts for the BBC - hardly a pro-tobacco organisation - and is the author of The Tiger That Isn't, a book which looks at the misuse of statistics. He rarely discusses smoking at all except when, as here, it involves statistical sleight of hand.

Deprived of an opportunity to accuse critics such as Blastland of being in the pay of industry, Dockrell resorts to the slur of saying that:

"their position echoes the so-called 'AIDS Dissidents' who continued to contest that HIV was a causal agent in AIDS long after the scientific debate was over."

This is a very feeble argument, particularly coming from a member of ASH. Anti-smoking activists like Dockrell are the first to complain when smokers' rights groups accuse them of being 'Nazis' on the basis that the modern anti-smoking campaign has "echoes" of Hitler's own efforts to stamp out tobacco.

Comparing critics of these heart studies to AIDS deniers is spurious and offensive. AIDS deniers, as Dockrell states, protested "long after the scientific debate was over." Debate about studies like that of Pell's has barely begun. The first of them - from Helena, co-authored by Stanton Glantz - only appeared in 2003 and the majority of the other studies have been published in the last two years. Pell's paper appeared barely six months ago. To compare this slender and new body of epidemiological evidence to the vast body of scientific evidence about HIV is absurd.

By mentioning the AIDS "dissidents" Dockrell is clearly attempting to persuade the reader that both bodies of evidence are of equal merit, and both sets of detractors equally deranged. Worrying, this fallacious argument comes hot-on-the-heels of an article puiblished in the European Journal of Public Health, titled 'Denialism: What it is and how should scientists respond?" (by Martin McKee, who happens to edit the journal) which groups together creationists, AIDS deniers and critics of the Pell study as being peas in the same pod.

In it, McKee writes:

"It took many decades for the conclusions of authoritative reports by the US Surgeon General and the British Royal College of Physicians on the harmful effects of smoking to be accepted, while even now, despite clear evidence of rapid reductions in myocardial infarctions where bans have been implemented, there are some who deny that secondhand smoke is dangerous."

The implication is clear: those who dispute studies such as Pell's also deny that primary smoking is dangerous. Somewhat inevitably, McKee then introduces the most famous and most repellent type of "denial": Holocaust denial.

"This phenomenon has led some to draw a historical parallel with the holocaust, another area where the evidence is overwhelming but where a few commentators have continued to sow doubt."

This goes beyond the pale. Equating critics of passive smoking studies with holocaust deniers is unacceptable, but it serves as a reminder of why the word "denier" is becoming so popular amongst those who would prefer to close down debate. It is quite deliberately chosen to bring to mind images of cranks, fraudsters and neo-Nazis. It is, perhaps, the ultimate insult and it is, of courtse, utterly fallacious.

The evidence for the Holocaust is documented in hours of film footage and remembered by hundreds of thousands of eye-witnesses. Similarly, the link between HIV and AIDS has been proven by solid biological evidence. What McKee laughably calls the "clear evidence of rapid reductions in myocardial infarctions where bans have been implemented" consists of nothing more than a handful of epidemiological studies of dubious value, created and funded by highly partisan bodies and individuals. Comparing this pitiful selection of flawed studies to the thousands of studies conducted into primary smoking, let alone sixty years of evidence for the Holocaust, is offensive and obscene.

This kind of abuse "echoes" (as Dockrell might say) the accusation - common in Stalinist Russia - that those who failed to accept the 'science' of dialectical Marxism (or the junk science of Lysenko) - were either mentally ill or were "dissidents" (a word that Dockrell uses seven times in his article*). This prevailing view effectively ended scientific and political debate in the Soviet Union in the 1930s just as Dockrell and McKee would like to end the debate over the 'heart miracles' today. Both their articles push for a revival of the notion of scientific heresy, rebranded as denialism, something so devilish that it must not be allowed to be heard. It is hard to think of a concept that could be further removed from the scientific method.

Dockrell refers to the famous criteria of causation set out by the noted epidemiologist Austin Bradford Hill, and claims that the Pell study would pass Hill's test of strength and consistency. Aside from the fact that the rise in heart attacks in year two of the smoking ban shows this to be sheer nonsense, both he and McKee would do well to remember that neither Hill nor Richard Doll ever resorted to character assassination when their early studies were contested by other scientists. They might learn from their example, and from the example of Albert Einstein, who invited Arthur Eddington to put his ideas the most rigorous of tests. Or do they believe the findings of Dr Jill Pell to be more robust than the theory of relativity?

Honest scientists do not engage in ad hominem attacks. Honest scientists are happy to have their theories scrutinised, even when they believe their opponents to be utterly misguided. Honest scientists do not announce their findings to the press and then refuse to answer questions. Honest scientists do not refuse to release their raw data. Honest scientists do not claim that a scientific debate is over before it has been allowed to begin. Above all, honest scientists do not slander their critics with barely concealed accusations of madness, corruption or worse.

Honest scientists must speak out about the erosion of scientific discourse and reject the ugly concept of "denialism", censorship and all the fallacious and twisted reasoning that Dockrell and McKee embody.

*Dissidents or dissidence

Christopher J. Snowdon is the author of Velvet Glove, Iron Fist: A History of Anti-Smoking

If you have any comments or corrections please send them to

Robert Wagner gives an interesting insight into how Pell was able to conjure up the much-quoted 17% drop in heart attacks here.