A recent issue of Nature (3rd November 2011) had two unrelated pieces that together merit comment. The first is a viewpoint by James Shelton, who is a science advisor at the US Agency for International Development, discussing a paper published in The Lancet Infectious Diseases recently claiming that injectable hormonal contraception doubles the risk that women will become infected with HIV. The paper was based on observational studies in African populations. The paper also implicated oral contraceptives, although the effect of the latter did not reach statistical significance.
This is a question that has been the subject of many observational studies with diverse conclusions. Shelton evaluates the likelihood of a causality relationship, and in doing so he shows the importance of looking rigorously and critically at a study, particularly when the conclusion may have important implications.
In itself that is nothing distinctive, rather just an example of how we should all look at scientific papers. But the reason for the commentary in Nature arises because of the wider and alarmist media impact this paper had, and the lack of any critical thinking in the media reports. Shelton suggests that responsibility lies with authors, with scientific editorial writers and with scientific journals’ own press offices to ensure balanced and appropriate statement of what are rational conclusions. To that list I would have added the increasing role of institutional press offices, who can also over-hype and over-simplify complex results to enhance institutional profile. The role of the lay media, and in this case particularly the New York Times, in exaggerating the conclusions is also discussed.
In the same issue of Nature there is a series of reviews regarding the conundrum of autism. Amongst the papers is a discussion as to whether there really has been a change in the prevalence of autism or not. The paper suggests that perhaps the popular view that the prevalence is rising is a misconception. The issue of whether there is an environmental factor acting in early life is discussed, but the evidence is far from clear.
There is an obvious connection with the story above. A number of years ago, Wakefield published in The Lancet a paper that claimed a linkage between autism and childhood vaccinations. The paper was controversial at the time, and many experts criticised the data and interpretation and said that it should not have been published in such a high impact journal. Indeed, we now know that the data were suspect; the paper was retracted by most of the co-authors and Wakefield himself was found guilty of misconduct by the UK’s General Medical Council.
This sad story involves the failure of authors, of peer review and of editorial policy, and the consequences of inappropriate media hype. It led to a dramatic fall in immunisation rates, to all sorts of unjustified angst from parents and healthcare professionals, and arguably a big shift in focus in understanding the causes of autism. Redressing the scientific error and the public misconception is hard and has taken a long time, and we still see the consequences today: immunisation rates in some parts of the population are still low. How much is this contributing to the current resurgence of measles?
Indeed the conflation of different interests — scientific egos, competition for funding, institutional competition, journal impact and sales, trends in the public media and the changed nature of access to public information — creates real challenges for how we communicate responsibly. For in many areas of science, not only in health, faulty communication can have long-term consequences. We all have a collective responsibility. It is not easy, but as the issues science deals with are so pervasive we have to find a way to do better.