Skip navigation

Cancer and the internet: The strange, sad case of Wei Zexi

The death in April of Wei Zexi, a Chinese cancer patient who died after mistaking an ad for an experimental cancer treatment for medically reliable information, should give pause to anyone who thinks that searching online for health information is, at worst, irrelevant but harmless. This is a topic that Elad Yom-Tov has been studying for years—he's a researcher at Microsoft in Israel and the author of Crowdsourced Health: How What You Do on the Internet Will Improve Medicine. We asked him about the Wei case—here are his thoughts.

Wei Zexi, a 21-year-old student Chinese student, was treated for a rare form of cancer with radiation and chemotherapy; when these did not work, he looked online for other options. He found a hospital in Beijing that claimed to have an experimental treatment for his condition, but after travelling to the hospital and taking the expensive treatment, he found it was neither effective nor experimental.

In a widely circulated post before his death, Mr. Wei blamed China’s largest search engine, Baidu, for directing him to the hospital by providing him with paid advertisements that masqueraded as objective information. “We thought: Baidu, a top-ranked hospital ... everything must be legitimate,” CNN reported him writing. “A Chinese student in the United States helped me Google relevant information…Only then did we find out that American hospitals had long stopped using the technology (used in the treatment) due to poor results in clinical trials.”

Baidu is now facing criticism in China from both the government and its citizens for failing to clearly delineate paid links (advertisements) from unpaid ones (“organic results”). While this failure may cause Baidu to lose credibility among its users, it should also alert us to the wider problem of how we perceive information presented to us by search engines, even—or especially—when we focus on the organic results. Do these pages contain truthful, objective information, or simply content that the search engine algorithm considers most likely to satisfy our query?

Consider childhood vaccinations, one of the most successful public-health campaigns in recorded history. When I search Bing or Google for “what are the dangers of vaccines”, most of the links these search engines show provide largely fallacious information on the “dangers” of vaccines. Very few links provide the truth—that “overwhelming medical evidence proves that negative side effects are rare and minor.”  The reason for this is that most people who query for the dangers of vaccines do not begin their search with a clean slate; they already have an opinion, which they wish to reinforce it. They are, therefore, more likely to appreciate those pages that claim that vaccinations are detrimental to one’s health than the pages which provide truthful information. Search engines, which learn from people’s behavior, rank these pages higher.

Indeed, our work has found that selective exposure, the tendency of people to seek and read information congruent with their prior beliefs, leads people to search for information that reinforces their views (the “filter bubble”). In politics, only around five percent of people at the most extreme ends of the political spectrum will bother reading information from the other extreme. This tendency also leads people to interpret the information they are given according to their prior beliefs. Two people might read the exact same page about vaccines, but their takeaway will be wholly different because of their prior opposing views. So the fact that a user has found information online does not mean that it’s truthful, nor does it necessitate that she correctly interprets it.

Search engines, together with advocacy groups and other organizations, have begun addressing the problem of information accuracy. Google, for example, provides verified health information in clearly delineated boxes. The Health on the Net Foundation provides certification for websites that contain useful and reliable medical knowledge. These efforts may not solve the problem of which information people seek or how they interpret it, but they will provide people with accurate, dependable, information.

The information we receive when we search online is the result of a complex interplay between what we seek, what others have sought, search engine algorithms, and the information available on the Internet. It is critically important, especially when we seek information that can affect our health, to realize this and to treat the information wisely.


Or, if you prefer to use an RSS reader, you can subscribe to the Blog RSS feed.


Books, news, and ideas from MIT Press

The MIT PressLog is the official blog of MIT Press. Founded in 2005, the Log chronicles news about MIT Press authors and books. The MIT PressLog also serves as forum for our authors to discuss issues related to their books and scholarship. Views expressed by guest contributors to the blog do not necessarily represent those of MIT Press.