Epidemiology and the media

By Guest Work 11/07/2013


By Andrew Whitehouse, University of Western Australia

Science and the media have long been uncomfortable bedfellows.

Traditionally, much of the prickliness has come from the side of scientists, who (often unfairly) view ‘the media’ as having a tendency to swap scientific truth for headline sexiness.

But in the 21st Century, the landscape has changed significantly.

No longer is it acceptable for scientists to communicate the fruits of their labour solely through the narrow mechanisms of peer-reviewed journals and conference presentations. There is now a clear message from the majority of research stakeholders that the findings of scientific studies are to be transmitted to a broader audience.

This dictum is particularly strong in medical research, where applicants to funding bodies must detail clear plans to communicate their findings beyond the confines of the scientific community.

The media, of course, represent(s) a highly effective means through which science can be communicated, and the nose-wrinkling that many scientists once directed towards TV, radio and newspapers has slowly started to diminish.

Undoubtedly this is a good thing. At times, science can suffer from a hint of elitism, and is too often accessible to only a small slice of the community. Providing everyone with ready access to the wonders of science can have untold benefits to broader society.

The new group-hug between scientists and the media has been embraced by many, but there has inevitably been some growing pains. In my opinion, this is no more evident than in the area of epidemiology.

Epidemiology

Epidemiology is the study of how often diseases occur in different groups of people and why. Epidemiology is an enormously powerful research tool that can be used to plan and evaluate strategies to prevent and manage illness.

Epidemiology is a particularly media-friendly area of science, primarily because it allows researchers to identify and quantify ‘risks’ for a given disease. Barely a day goes by without a news story proclaiming that pesticides/red wine/caffeine/you-name-it increases risk for cancer/autism/heart disease/you-name-it.

I hesitate to write that these news stories are unconstructive – after all, the latest scientific findings are being communicated to the widest possible audience. Nevertheless, these reports do cause me concern.

Discovering causes of disease

The main reason for my concern is that, often, the findings that are broadcast are only preliminary. By using the term ‘preliminary’, I’m not casting judgment on the quality of the science, which is often first rate. I’m more referring to epidemiology being only the very first stage of the scientific pathway that identifies definitive causes of disease.

Let me explain.

Epidemiological studies often form the first stage of the discovery pathway. These studies survey a large population and find a pattern of results that indicate a certain factor may be coinciding with a certain disease more often than we would expect by chance. To use a well-established example: the finding that lung cancer is more common among smokers than non-smokers.

The next stage in the discovery pathway is replication – finding the same result in an independent sample of people. Replication is a cornerstone of science. Quite simply, scientists cannot make sweeping conclusions from one study alone. It is entirely possible that a research team conducts a rigorous study that includes all of the necessary checks and controls, and the study still yields a statistically significant finding by pure chance. Replication of study findings is just as important, if not more important, than the initial result itself. To continue our example, this would mean examining lung cancer rates in a different sample of smokers and non-smokers.

The stage beyond replication is validation. One can make a solid case that epidemiological studies cannot prove causation. Epidemiology is an extremely powerful research tool, but we must be clear that discovering correlates of disease is not identifying the cause of anything. To do this, we need to conduct validation studies: detailed investigations of the exact mechanisms that may be causing the association. Taking up our example, a validation study would be to expose animals to tobacco smoke/extracts under highly controlled laboratory conditions, and then examine whether tobacco exposure led to damage of the lung tissue.

The most important point for us to understand about the discovery pathway is that very few ‘risk factors’ make it through all three phases intact.

Discovery, replication and validation of any scientific finding, is the exception, not the rule.

Serving the public

And this brings us back to the news story reporting the latest epidemiological finding.

The question I ask is this:

Is the public really being served well by being informed of ‘discovery’ findings from epidemiology that have not yet been replicated or validated?

I would argue that when it comes to health and medicine, they are not.

Picture a parent who has just had their child diagnosed with autism (or cancer, heart problems etc.) listening to a news report on an epidemiological study that has linked exposure to traffic pollution during pregnancy with autism. The mother casts her mind back to her pregnancy and remembers living next to a busy road. For reasons of human nature, that mother will feel some guilt, however small, that she is to blame for her child’s autism.

The problem is that, even with the caveats at the end of the news report that ‘the authors caution that more research in this area is needed’, that guilt will survive. Indeed, even if replication and validation studies prove the initial ‘discovery’ to be inaccurate – a very real possibility – that guilt will still be difficult to shake.

This may seem like a contrived situation, but I can assure you that it is not. It happens all the time.

Is this a situation that we want? Again, I would argue that it is not.

What can scientists do?

Of course, there are many factors beyond the control of scientists that contribute to what research findings are taken up by the media: university competition, grant funding, the political climate and the quick media cycle, being just a few.

The media are not the enemy of scientists – far from it. My experience is that most media organisations have wonderful Science Editors, who are extremely keen to convey scientific findings accurately.

But the simple truth is that earth shattering findings do not happen every week.

I truly hope that the relationship between scientists and the media continues to blossom.

What I also hope is that individual scientists – including me – think harder about how, and most importantly when, we choose to publicise our findings. And that we always keep in mind those families in the clinic.

Click here if you would like to be on the mailing list for this column.

Andrew Whitehouse does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.

The Conversation

This article was originally published at The Conversation.
Read the original article.