Popular Science - - CONTENTS -


over­shar­ing as a life­sav­ing habit, but maybe it is. For more than a decade, epi­demi­ol­o­gists and data sci­en­tists have scanned our search-en­gine queries and so­cial-me­dia posts with the goal of dis­cern­ing who is in­fected, what they have, and where they live. But de­riv­ing mean­ing from our con­sul­ta­tions with Dr. Google faces an ironic ob­sta­cle: For all our co­pi­ous snaps, self­ies, and sta­tus up­dates, we’re just not shar­ing enough to con­sis­tently fore­cast dis­ease out­breaks—in­clud­ing the flu.

Of course, in­fluenza’s reign of ter­ror started long be­fore the birth of our mod­ern so­cial net­works. A hun­dred years ago, the in­fa­mous “Span­ish flu” spread rapidly around the world, in­fect­ing a third of the pop­u­la­tion and killing at least 50 mil­lion peo­ple. With the rapid evo­lu­tion of the virus, and in­creas­ing in­ter­na­tional travel and ur­ban­iza­tion en­abling the quick spread of ill­nesses, a mod­ern ver­sion of that pan­demic could cause twice as many ca­su­al­ties, along with wide­spread dis­rup­tion to the global sup­ply of food, medicine, and en­ergy. It doesn’t mat­ter where you live or what you do. The flu could in­fect you.

Even in the ab­sence of Flumaged­don, im­prov­ing our abil­ity to fore­cast the ill­ness is vi­tal. In­fluenza viruses kill up to 646,000 peo­ple world­wide ev­ery year, in­clud­ing as many as 56,000 peo­ple in the U.S. Amer­i­cans pay as much as $5.8 bil­lion in med­i­cal care an­nu­ally to fight the pesti­lence. If we know when it’s coming, health agen­cies could push peo­ple to get vac­ci­nated. Hos­pi­tals could plan ahead.

Aug­ment­ing of­fi­cial flu reports from the Cen­ters for Dis­ease Con­trol and Pre­ven­tion (CDC) with data har­vested from the in­ter­net is an­other step in our on­line evo­lu­tion. Ac­cord­ing to a 2012 Pew Re­search Cen­ter study, about 184 mil­lion Amer­i­cans (more than half the na­tion’s res­i­dents) use the Web to find healthre­lated in­for­ma­tion. These searches are like tips to a crime hot­line, en­abling re­searchers to iden­tify sus­pected flu cases. In 2006, Gun­ther Ey­sen­bach, as­so­ciate pro­fes­sor of pub­lic health at the Uni­ver­sity of Toronto, found that searches for the terms “flu” or “flu symp­toms” spiked a week be­fore a jump in doc­tor vis­its. “The in­ter­net has made mea­sur­able what was pre­vi­ously im­mea­sur­able,” he wrote in 2006, chris­ten­ing the new field “in­fodemi­ol­ogy.”

In 2008, Google rolled out Flu Trends, har­ness­ing its own big data to look for world­wide flu surges and hot spots through symp­tom searches in 29 coun­tries. Google scrapped the pro­gram in 2014—be­cause of at least one fac­tor that re­searchers hadn’t counted on.

Your search his­tory, it turns out, can be mis­lead­ing. It’s im­pos­si­ble for data col­lec­tors to know whether you were look­ing up “headache and fever” for your­self, or be­cause you heard your co-worker com­plain­ing about their kid’s symp­toms. In 2007, Amer­i­cans sud­denly started Googling “cholera”—had a new epi­demic taken hold? Nope. Oprah Win­frey had just rec­om­mended Love in the Time of Cholera for her book club. “You should have seen what hap­pened when Brad Pitt had vi­ral menin­gi­tis,” says Lone Si­mon­sen, pro­fes­sor of epi­demi­ol­ogy at Roskilde Uni­ver­sity.

Af­ter culling search data from pub­lic re­sources, re­searchers run them through

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.