Survey Rode the Media Wave of Flu Mania

Tweet

As swine-flu mania was at its hottest back in April, Stanford University academics Marcel Salathé and James Holland Jones saw a unique opportunity to gauge both the public’s ignorance and its expertise about an emerging health threat in real time.

Notice the term “swine-flu mania.” Only on Thursday — seven weeks after the profs’ epiphany — did the World Health Organization declare swine flu, or H1N1 if you prefer, a pandemic. The number of new infections and deaths attributable to swine flu grows daily, with a muffled drumbeat of stories across the globe as each locality notches a grim first. As Jones noted in his blog, Monkey’s Uncle, on May 18, “As the potential pandemic fades into the obscurity of a couple weeks’ worth of the 24-hour news cycle, cases continue to mount.”

The media mania about swine flu has deflated, something quickly ascertained by checking the term “swine flu” on Google Trends for 2009. Google suggests April 28 was the peak globally, at least based on Anglophone news accounts. (Since the illness in its current incarnation first gained prominence in Mexico, note that gripe porcina coverage peaked a day earlier, April 27.)

But Salathé, an expert on parasites, and Jones, an anthropologist currently focusing on infectious diseases and demographics, knew — or perhaps intuited — something critical. Not about viruses, but about people: Emotional status drives behavioral response.

On an individual level, that may be pretty self-evident. Angry individuals do angry things, sad individuals do sad things. And even among crowds, it’s apparent, hence the fear of mob mentality. But this was a chance to view that behavior over a longer period, with people who didn’t necessarily interact with each other — and over a matter of genuine public concern. Underlying it all was an important public health question: Could that response alter the progression of the disease.

So they created an Internet accessible survey and put the word out that they wanted random people to fill it out. As Salathé wrote in an e-mail he sent as the survey went up on April 29, “One of the major problems is that emotional reactions and behavioral response to an epidemic is generally assessed quite some time after the epidemic has fizzled out. We would like to address this problem by starting a survey about risk assessment and personal responses to a potential epidemic as it unfolds – that is right now.”

The survey — it was spotlighted on Miller-McCune in a piece penned by Salathé, although it’s biggest boost probably came after being featured on the social-networking site Boing-Boing — had received 8,000 responses by the beginning of this week. It’s still taking responses, although, as Salathé noted, “The response rate followed the media coverage rate — i.e. very high initially, and lower as time went by.”

What they found was both encouraging for those in public health, and discouraging for those in the media.

“In a nutshell,” Salathé wrote in response to e-mailed questions, “people became less anxious with time as more information about the virus became available and as media coverage waned. People continue to be well informed.” The results, while not surprising, were gratifying. “We expected to see this, but now we have the data.”

Meanwhile, the actions people took — both silly and serious — were the behavioral response being played out. Anecdotes from around the globe — travel bans to Mexico, Mexicans quarantined in China, trash-eating pigs slaughtered in Cairo, ineffectual face masks worn on public streets — flared and then trailed off.

As Salathé explained in examining his survey data, “People very quickly became more relaxed — this was a bit unexpected to me, but on hindsight it makes sense, too — the emotional status was apparently mainly driven by the media, rather than by the actual epidemic curve. Whether the change in behavior actually altered the progression is impossible to say (and will remain impossible because infectious diseases in humans are not subject to experiment, thank goodness).”

The pair, with backgrounds in studying biological organisms, were treated to a peek at a different kind of ecosystem: the media. Their observations should warm the hearts of struggling “old media.”

“We can’t say what drives emotional status, but media coverage certainly has a great effect,” Salathé said. “Interestingly, we do not see a clear effect of any type of information source. One might speculate, for example, that getting informed by TV would increase anxiety, while getting informed by health officials would decrease anxiety. We don’t see that — instead, they have roughly the same increasing effect.

“In general, ‘immediate and central’ information sources (i.e. TV, radio, Internet, health officials) increase anxiety, while ‘delayed central’ information sources (print media) decrease anxiety. ‘Decentralized’ information sources such as friends and online social networks have no effect.”

The last point takes on added interest since the Salathé/Jones survey was portrayed in some quarters as taking the pulse, so to speak, of the Twittering classes. In his piece for Miller-McCune, Salathé even pointed to the glut of Tweets dealing with swine flu, some heavy with the aroma of panic.

That the survey was online only might raise concerns about its comprehensiveness, although the Internet is the only current common technology that could allow such a quick setup. But the authors acknowledge up front that their survey is biased in favor of Internet users and those with increased education.

“However,” wrote Salathé, “we also do have an exceptionally large number of respondents, and we think this makes up for the bias. One could even argue that a bias towards increased education makes the sample more conservative if one assumes that more educated people are less prone to emotional manipulation by the media. Whether that is true is, however, an open question as far as I know, at least in the context of infectious diseases.”

The pair are about to publish their results in an open-access journal, and both hope to continue to monitor the pandemic to decide what’s next on their research agenda. More surveys are likely, Salathé said.

“Ideally, we would have more time to think about the questions we want to ask, but the important fact about our survey was that it was so immediate. We are thinking about how we could get a less biased sample in the future, but if things happen so fast again this will be difficult.”

And there will certainly be another chance, perhaps even with the same illness. Swine flu itself has not gone away, evidenced by both the WHO’s declaration and the normal infection patterns of seasonal influenzas.

As Jones wrote on his blog, “I can hope is that people have not become inured to warnings of epidemics because of our recent experience with H5N1 bird flu and this new H1N1 swine flu (there is also the last swine flu scare of 1976). Some saner press coverage would help. Of course, it would mean less grist for the mills of John Stewart and Stephen Colbert, but it might mean a public better prepared for a potentially real public health emergency that we still may face.”

Sign up for our free e-newsletter.

Are you on Facebook? Become our fan.

Follow us on Twitter.

Add our news to your site.

Tweet