Re: Evidence about electronic cigarettes: a foundation built on rock or sand?
This authored analysis recapitulates the claims of an editorial in the Lancet two weeks ago that our MCDA analysis of the comparative harms of nicotine products was unreliable – a claim we instantly refuted http://drugscience.org.uk/blog/2015/08/27/blinded-smoke-why-do-e-cigaret... but is repeated now in the BMJ.
The justification both commentaries use for this claim is not the data we reported but the fact that a few of the experts in our group of twelve had declared potential conflicts of interests [CoIs] in relation to tobacco harm reduction products. Also it was implied that the funder of the MCDA project, EuroSwiss Health, was paid by the tobacco industry. These claims were publicly refuted by those named straight after the Lancet editorial [see link above], so why did the BMJ article continue in the same vein?
We suggest that in repeating these allegations the BMJ authors were following a sadly all-to-common path for those who disagree with research conclusions to use CoI claims to undermine confidence in the findings rather than to critically analyse the data. A charitable view could be that they in fact can find no fault with the data [other than they dislike the conclusions], so have no alternative but to attack the scientists who generated it.
We see such CoI attacks being used to try to discredit work in topics as divergent as climate change and the use of statins in preventing cardiovascular events [recently in the BMJ]. This approach has a long history; the seminal study of the transmission of cholera through infected water by John Snow was in 1855 attacked by the Lancet on the grounds that it supported firms who had previously been blamed for causing this infection e.g. tanners [and it took them over 100 years to admit they were wrong - http://www.thelancet.com/journals/lancet/article/PIIS0140-6736%2813%2960....
What is sad about this current episode is that there is a massive public health need to determine the best approach to alternative forms of recreational nicotine products, which electronic cigarettes and snus provide. The MCDA approach is easily the best way of approaching such complex and difficult topics. It was used to determine the best approach to Nuclear Waste disposal and to compare the harms of legal and illegal drugs [a paper of ours famously published in the Lancet in 2010 http://www.thelancet.com/journals/lancet/article/PIIS0140-6736%2810%2961....
The MCDA process is such that individuals with biases cannot influence the outcome because alternative points of view are represented in the expert group and the numerical inputs judged by participants are subject to on-the-spot peer review, thus rendering the CoI criticism even less relevant than in other areas of research output. Moreover it has proved exceptionally robust when replicated with independent sets of experts http://jop.sagepub.com/content/early/2015/04/23/0269881115581980
Finally there is the issue of giving the accused a chance to explain their position before being publically criticised. I have been attacked by a number of low quality newspapers over the years but each have had the courtesy to inform me of the accusations prior to the article appearing in print, and allowing me a comment in the article. Perhaps the so-called “scientific” press should sign up to the same standards of journalism?
Competing interests: No competing interests