It's repeated so often that it has long been regarded as a cliche, but we live in an increasingly information-intensive world, bombarded by facts and figures from an endless queue of media outlets, websites, television shows and Windsor-based science bloggers. This abundance of information often comes with a cost. If my grandfather wanted to learn something about his health - and of course like many men of his generation he didn't - he would have seen a doctor or read a reputable book. These days, we receive much of our information on the fly in bite-sized chunks from websites and media articles. Fast food culture applied to research.

The result is like trying to build an igloo by catching snowflakes. We snatch little snippets of information here and there, but often they lack any real substance, failing to really contribute to the building of a complete understanding of a subject. Stripped of context or reference, in the end these factlets are as intangible as the ether they travel through.

Just as the quality of information has... if not necessarily declined then certainly altered, the importance of health in the public eye has increased. Perhaps this is coincidental, and perhaps not. Certainly there is a public craving for health information matched only by the modern obsession with celebrity, and media outlets have been keen to fill this need. In this information market quality is not necessarily that important to the producer, and neither is it particularly easy for the consumer to measure.

So that's the situation we find ourselves in today. The question is what effect does this have on the public's understanding of health? That question is rather too large for one study to cover, so Young, Norman and Humphreys address a sub-question in their recent paper, "Medicine in the Popular Press: The Influence of the Media on Perceptions of Disease" [1]. How does media coverage affect the perception of disease risk?

The authors came to their question from a well-known observation, that the public understanding of health risks (and risk in general) bears little relation to actual risks [2], but strongly correlates with media coverage. As the authors note, "The media tend to focus on rare and dramatic events. As a topic receives repeated coverage in the media, public attention is drawn towards that particular topic and away from competing sources of concern."

The media effect is so strong that it can produce significant health consequences in its own right. For example, the rate of Post-Traumatic Stress Disorder (PTSD) and other stress reactions associated with the 9/11 attacks, among people not in the immediately affected areas, was directly related to their level of exposure to media reports [3].

So the specific question the authors set out to address is as follows: "...to what extent is public perception of infectious disease modulated by the high levels of popular media coverage in North America?" In order to address this, they carried out three experiments.

The first experiment involved a survey of around fifty undergraduate students. They were asked about ten different medical conditions, for which they had to rate the seriousness (out of ten), the likelihood that the condition was a real disease (out of four), and how many out of 1,000 of their peers might get it in the next year. They were further split into two groups - one group was simply given the name of the disease, while the other was given the name along with associated information about symptoms, treatment, mortality and so on. The second experiment simply repeated the first, but with a pool of forty-three first year med students.

The third experiment took just twelve graduate students, presented them with two sets of conditions with the disease names removed, and asked them to assess which was worse.

Ten conditions were referenced in these experiments. Five of the disease were sexy, celebrity diseases frequently seen in various states of undress, falling out of nightclubs and plastered all over the the pages of all the hot medical columns: anthrax, SARS, West Nile virus, Lyme disease and avian flu. The other five were the kinds of diseases that need to hire a new agent: tularemia, human babesiosis, yellow fever, Lassa fever and hantavirus. If you've heard of all five of those, I'm impressed.

You're probably thinking - as I did when I first saw it - that the third experiment sounds a bit crap, but actually it's not really part of the study. The aim of experiment three was basically to get a panel of people to independently and objectively confirm which diseases the first two experiments covered were actually worse. The low-media frequency conditions were generally seen as worse than the "sexy" diseases.

With that baseline established we can look at the first two experiments. The purpose of them to to see what impact the name of the disease had on its own, and how this was modified by associated knowledge.

On the "seriousness" scale from 1 to 10, the undergrad students (experiment one) gave an average score of 7.8 to the fashionable diseases, and only 6.66 to the others (with a standard deviation of 0.174 in both cases). On the "disease-like" scale from 1 to 4 they rated the sexy diseases 3.04 (SD 0.088) and the others 2.74 (0.075), so there was a clear bias towards the popular diseases. That trend continued even among the more knowledgeable med students (experiment two), with average "seriousnesses" of 7.98 versus 6.68, and "disease-like" score of 3.14 versus 2.95 for the diseases with high and low media profiles respectively.

Some participants had more information than others, and in the undergraduate group this had a fairly sizeable impact, but less so among the med students. The authors suggest that this is presumably because they already had more objective knowledge to begin with, which seems reasonable.

So on the basis of the above it seems that media reporting has a strong impact on how serious people believe a disease to be, but what about the perceived prevalence? Interestingly, the results for this were all over the place, and there was no significant increase in perceived prevalence for the well-known diseases. It's a result that the authors gloss over somewhat, but I can't help but feel that the style of the question may have played a part in this, and that their study was seriously flawed in this regard.

Asking a person to rate something's relative seriousness out of ten or four is asking for a subjective judgement with a very finite range of possible answers, but asking somebody to estimate a specific absolute value for something is a whole different type of question. On a scale of 1 to 10, people know what the values mean - 1 is low, 5 is average, 8 is high. But asking for an incidence per thousand is meaningless to a lot of people - most people don't have a clue what an average, high or low figure would be in this context. I predict that if this study were repeated, and the question were altered to ask for a "Risk" from 1 to 10, the results would be substantially altered.

So to me, the results for prevalence are meaningless, but the central result is still very solid. Not only did the authors demonstrate that people rate diseases with high media coverage as being more serious, but there was a significant correlation between media frequency and perceived seriousness too.

So what does this all mean? Well it's another piece of research to add to the growing body of evidence that shows that newspapers can have a direct effect on public health attitudes. Indeed, it's hard to to relate this back to the MMR Hoax in the UK, and research conducted into the media's role there.

When the facts are presented dispassionately, people will make rational judgements about relative risk, but of course dispassionate reporting is not what newspapers and media outlets are about. As the authors note: The media function as a critical interface between the scientific community, government, and the public [9,34] with a responsibility to strike a careful balance between raising awareness of issues of public concern and irrationally alarming the public at large.

That responsibility isn't taken seriously by the industry, neither has any attempt been made by outside agencies to push the issue, probably at least in part because politicians depend on friendly media organization to get them elected.

The problem is that the idea of a good media story is the complete antithesis of a good presentation of facts. Media stories thrive on the novel and unusual rather than the mundane, even at the expense of higher risk issues. The authors note that research shows viewers remember less than a quarter of the information content from an average story, and that these stories increasingly focus on personal threat to the viewer, often backed up with exaggerated claims. Again, the MMR controversy stirred up by the British press is a classic example of newspapers exaggerating the claims of one maverick doctor in order to manufacture a scandal. In the end, the only scandal was that parents left their children unvaccinated as a result of media scares.

The paper closes with the following remark: "Given the results presented in this paper, it is imperative that we fully understand the effects of the media on public perceptions of disasters and disease epidemics. In this age of television and internet media it is important to consider the impact of media reporting on public perception of risk, and public health in general."

Imperative it may be, but it is also irrelevent unless some effort can be made to change the behaviour of the media, to make them understand the massive reponsibility they hold for public health.

(PLoS ONE @ Two. Entered for the PLoS 2nd Anniversary Synchroblogging Competition)

[1] Meredith E. Young, Geoffrey R. Norman, Karin R. Humphreys (2008). Medicine in the Popular Press: The Influence of the Media on Perceptions of Disease PLoS ONE, 3 (10) DOI: 10.1371/journal.pone.0003552

[2] Kasperson RE, Renn O, Slovic P, Brown HS, Emel J, et al. (2002). The social amplification of risk: A conceptual framework. Risk Analysis, 8, 177-187

[3] Ahern J, Galea S, Resnick H, Kilpatrick D, Bucuvalas M, et al. (2002). Television images and psychological symptoms after the September 11 terrorist attacks. Psychiatry (65), 289-300