September 2019, ScientificAmerican.com 97 eynep Tufe ci is an associate professor at the University
of Nor th C arolina School of Information and Librar y Science
and a regular contributor to the New York Times. Her book,
Twitter and Tear Gas: The Power and Fragility of Networked Protest,
was published by Yale University Press in 2017.THE INTERSECTION
WHERE SCIENCE AND SOCIETY MEETIllustration by Enan LiangBeware of
Medical Web
Searches
? ̧ø`D³³l ̧ ̧l³
̧ßDî ̧³j
UøîîxßxÜäD§ ̧î ̧ßxUDl
By Zeynep TufekciWhen a sharp chest pain woke me up around 3 A.M., I had the
obvious question: Was this truly panic-worthy? This had never
happened to me before, and I’m in fairly good health—but I had
lost one parent to a sudden, early death that may have been a
cardiac event (we never learned for sure).
I may have the word “doctor” in my title, but I’m not that kind
of doctor. I thought about dialing 911, but then I noticed that ...
well, the pain was on the right side of my chest rather than the
left. My breathing wasn’t labored. My heart wasn’t pounding. So
like millions before me, I Googled my symptoms. On top of the
screen was an article entitled “16 Causes of Right Side Chest
Pain.” Bingo, I thought—except it was an ad. I moved on to the
actual search results, which were headlined “17 Causes of Pain in
the Right Side of the Chest” and “26 Causes of Chest Pain &
Tightness.” When I got to “3 Types of Chest Pain That Won’t Kill
You,” I started wondering: What were all these bizarre articles?
Suddenly, my medically useless doctorate seemed to be more
and more relevant. The titles were about gaming Google’s search
algorithm to grab people’s attention at their vulnerable moments.
This is called search-engine optimization, or SEO: the art and
science of engineering for higher placement in Google search
results and getting people to click on the links.
Unlike a lot of “medical information” online, SEO itself is
steeped in the scientific method, which shows that this number-
heavy format plays to a human cognitive bias called stand-out.
We notice things that stand out—like oddly precise numbers.
When I was a child in Turkey, my grandmother loved various
over-the-counter remedies for maladies from the most minor to
the serious, and many of them contained menthol. Menthol may
not always have been the active ingredient, but it gave a sooth-
ing, medicinal aura to the lotions and rubs that filled her cup-
board. Similarly, these odd, specific numbers add a scientific
aura to the headlines.
I still needed information, though, so I clicked on a result
from WebMD. I had seen that site before, and it didn’t seem like
it was quackery. I was barely a paragraph or two in when I
noticed links about lung cancer symptoms. Huh? Lung cancer
often causes no pain until its later stages. Then I realized that
the information about lung cancer was an ad, but the “ad” nota-
tion was barely noticeable. During an emergency is obviously
not the best time to scare people into clicking links for unlikely
diseases. Indeed, if you clicked on these “lung cancer symptoms,”
the browser took you eventually to a site advertising a lung can-
cer medication from Merck. Ugh.
Using inappropriate screenings and symptoms to advertise
for drugs is not new at all. As early as 2010, WebMD had gotten
in hot water for a depression screener consisting of 10 questions
that, no matter how you answered, spat out the same an swer:
“You may be at risk for major depression.” You can almost smell
the legalese: we all may be at risk for major depression anyway.
No need for a quiz sponsored by Eli Lilly, a company that just
happens to manufacture the antidepressant Cymbalta.
Having medical information online be financed by advertis-
ers or pharmaceutical companies—or supplement manufactur-
ers—is certainly not healthy. They often have an incentive to
scare us. It’s a lose-lose situation. Sometimes we do need to be
worried and seek medical care. Other times it’s just fearmonger-
ing for clicks. But who can tell when one is in crisis?
Finally, I looked up whether my own university had a health
information site. Sure enough, there was a site with precise
answers for exactly my question, the first statement being that
“pinpointed chest pain” like mine was unlikely to be heart-related.
Just then I remembered that I’d had a glass of carbonated water
right before going to bed—a common cause of random but pin-
pointed chest and abdominal pain. My own “emergency” faded
away, but the state of health information online remains dire.JOIN THE CONVERSATION ONLINE
<äî2_w²í_Ĉ¬wÞ_C² ̧³D`xU ̧ ̧¦D³l5ÿîîxß
̧ßäx³lD§xîîxßî ̧îxxlî ̧ßi[email protected]