One of the things that originally motivated me to write about public health was the realisation that the checks and balances that have sustained scientific progress for centuries are being undermined or simply ignored by people whose interests lie not in painstakingly testing hypotheses to reach well-argued empirical conclusions, but in torturing data to suit preconceived outcomes consistent with a narrow philosophy.
The output produced by this unfortunate perversion of science is often newsworthy as it offers public health adherents an opportunity to push their doctrine, excites certain types of politicians and potentially impacts the lives of many through socially destructive bans or restrictions.
Increasingly, the "scientists" producing the headline-grabbing articles are public health activists who simultaneously campaign for whatever policy their "science" is supposed to support. This ought to raise issues over the objectivity of their offerings but doesn't because the publishers, editors and journalists responsible for the seemingly incessant public health bombardment appear to care more about being "on-message" than they do about honesty or objectivity.
A good example as to how advocacy presented as science makes headlines and influences policy is provided by a paper published in the European Heart Journal earlier this year that claimed to link passive smoking in childhood with a cardiovascular effect observable in adulthood.
The theory behind the project has merit in that the researchers tried to associate childhood exposure to parental smoking with something measureable in adults, namely the thickness of the inner layers of the carotid artery. The results were less than compelling but nonetheless resulted in sensational media headlines.
Essentially the study found a tiny difference in carotid artery thickness of no established clinical relevance in those exposed to two parents smoking that was not present when only one parent smoked. It did so by marrying data from two original studies with many methodological caveats and airy assurances that confounding variables were all accounted for. Any self-respecting scientist faced with a less than robust methodology and a meaningless marginal effect dependent on a contrived unquantifiable variable would conclude that the study was insufficiently robust to support any substantial claims.
So how did this mediocre excuse for science result in headline news from the BBC announcing that "Passive smoking damages children's arteries"?
To answer that question, we need to first look at its lead author Dr Seanna Gall. Gall does have a degree in science and a PhD in epidemiology but according to Australian Rotary Health:
So Dr Gall is not exactly objective in the context of this study and even if her method and statistics were more solid, it is not unreasonable to suggest that she might be inclined to seize any serendipitous opportunity to promote her cause by, for example, using a subset of subjects in order to claim statistical significance for the outcome that she very much hoped to find.'Since 2006 she has worked with the Tasmanian Government on smoking policies and has been a board member for Quit Tasmania for many years. In recognition of her work in this area, Quit awarded Dr Gall their 'Friend of Quit Tasmania' award in 2010. Dr Gall has been a member of the executive for the Tasmanian Branch of the Public Health Association of Australia. She is also a member of the Tobacco Special Interest Group, co-ordinating advocacy around tobacco in Tasmania.'
Having herself advised caution with respect to interpreting her data she amazingly concludes that, "results show the pervasive effect of exposure to parental smoking on children's vascular health up to 25 years later".
The data and methodology do not support this conclusion as written. It is dishonest wishful thinking.
She goes on to call for continued efforts to reduce smoking and the BBC quotes her as stating, "Parents, or even those thinking about becoming parents, should quit smoking".
One has to wonder, based on the evidence of history, how long it will be before the likes of Gall are campaigning to ban smoking in private homes or to take the children of refusenik smokers into care, for their own good of course.
Gall's work should have rung alarm bells during peer or editorial review so it is no surprise that this study did not appear in the scientific press but in a medical journal that claims that it "aims to publish the highest quality material, both clinical and scientific, on all aspects of Cardiovascular Medicine"
The flaws in this paper led me to sincerely doubt this so I wrote a strongly worded email to its editor, the Zurich-based cardiologist Professor Thomas Lüscher. His response did nothing to assuage my concerns about the quality and objectivity of his journal. He wrote:
For those without biology backgrounds, catecholamines are hormones involved in the "fight or flight" response. Their levels are increased by hard exercise, stress and smoking. For those who don't know me personally, I don't smoke. No special knowledge is required to understand the arrogance and bigotry."Thanks for your Mail - I am sorry that you wrote it in bad mood under the influence of catecholamines. Nevertheless, I am happy to respond.
Of course, all papers undergo peer review as well as statistical review as in this case.
If you have convincing statistical Arguments, please let me know - I will be happy to forward those to our statistical editor.
A last question: Are you a smoker?"
The professor's assumption that only someone who smokes would question publication of a scientifically weak study with heavy political overtones is worrying. Worse is his attempt to hide behind statistics, which reflects a modern trend that undermines good scientific practice. The observed difference in the results for one and two smoking parents is so fundamental to the study that irrespective of the finer points of the mathematics, it requires explanation before any claims pertaining to the study should be considered ethical or scientific.
Despite its very significant flaws the study was promoted via the usual cascade of hyperbolic uncritical press releases and, as some of the public health faithful are favoured by the BBC, Gall's dubious headline claim was thereby ensured a global audience.
In fairness, BBC health editor Michelle Roberts does point out some problems with the study and at least questions why the effect disappears when only one parent smokes. She hypothesises that this is "possibly because exposure was not so high", which is surprising as earlier in the same article she tells us that "experts say there is no safe level of exposure to second-hand smoke." Perhaps the impact of exposure is a variable that can be "expertly" adjusted depending on the message that activists wish to convey? Gall herself admits that she has no evidence-based explanation but that does not appear to prevent her leaping to conclusions.
In concluding her analysis Roberts sums up the approach of public health to science in general with the sentence, "Regardless, experts say all children should be protected from second-hand smoke."
This is how less than mediocre "science" translates into sensational unjustifiable headlines and a "growing body of evidence"
It seems that we can look forward to further campaigns for increasingly Draconian public health legislation regardless of the quality of the supporting evidence. With editors such as Lüscher and influential broadcasters on board it is demonstrably possible to gain global exposure for illiberal authoritarian campaigns without worrying too much about apparently trivial concepts such as ethics and scientific integrity.
Chris Oakley has written about public health issues for The Free Society before - BBC comment boards show public health has no credibility
Reader Comments
to our Newsletter