Better search: a cure for cyberchondria?

  • 3 December 2008

Search engines may need to be redesigned to stop people turning into “cyberchondriacs”, a study by Microsoft Research has suggested.

The study found that general search engines can lead people to escalate medical concerns – so that, for example, they start by searching for “headache” and end up looking for “brain tumour.”

It also found that in a small number of cases, this can cause anxiety, distract people into making repeated further searches for information, or even seek medical attention they do not need.

US-based researchers Ryen White and Eric Horvitz argue that the algorithms that search engines use to rank information may need to be adjusted for medical search terms to reduce the risk of this happening.

“We are currently pursuing the creation of classifiers that indicate when a user is employing a search engine as a diagnostic system,” they say. “Search services might [then] provide a list of diseases sorted by likelihood, along with assistance and caveats in interpreting the results.”

White, a specialist in information retrieval, and Horvitz, an artificial intelligence expert, also suggest that new interfaces could be designed to guide users through medical information, such as flow charts and decision trees.

The term “cyberchondria” emerged in 2000 to refer to the tendency to leap to dire conclusions while researching health matters online. The Microsoft study set out to find out why and how frequently this happens, and what impact it has on searchers.

The researchers looked at three search terms – headache, muscle twitches and chest pain – and found that general search engines were more likely to associate these with serious but uncommon diseases than specialist engines and retrieved web pages.

They then surveyed 500 Microsoft employees, and found that many thought search engines ranked results according to the likelihood of their having a disease – rather than by algorithms that judge sites by click rates, dwell times and similar information.

As a result, they argue that search engines can amplify well known biases towards regarding dramatic or recent events as more likely to occur than they really are.

White and Horvitz also looked at how people actually search for medical information, using data mined from the Windows Live Toolbar over 11 months. They found that around 2% of web inquiries were health related and that 250,000 users made at least one medical search during the duration of the study.

Just over 8,700 of these users searched for one of the terms chosen by the researchers and around 5% of them “escalated” the query, either by adding terms such as “chronic” or “fatal” to their search or looking for more serious conditions.

The researchers also found evidence that some people repeated their searches or looked for information about more serious conditions over a number of web sessions, suggesting they were worried and distracted by them.

“Search engine architects have a responsibility to ensure that searchers do not experience unnecessary concern generated by the ranking algorithms their engines use,” they write. “They must be focused on serving medical search results that are reliable, complete and timely as well as topically relevant.”

Link

Cyberchondria: Studies of the Escalation of Medical Concerns in Web Search

 

Subscribe to our newsletter

Subscribe To Our Newsletter

Subscribe To Our Newsletter

Sign up

Related News

WHO launches collaborative network for data and digital health

WHO launches collaborative network for data and digital health

WHO is bringing together its European region member states with partners for a network focused on advancing data and digital solutions in health.
Harnessing AI and cybersecurity to transform healthcare in the UK

Harnessing AI and cybersecurity to transform healthcare in the UK

The UK healthcare sector is in a transformative era, driven by advancements in artificial intelligence (AI). AI has the potential to revolutionise healthcare by improving…
Calderdale and Huddersfield awarded HIMSS stage 6 for analytics capabilities

Calderdale and Huddersfield awarded HIMSS stage 6 for analytics capabilities

Calderdale and Huddersfield NHS Foundation Trust has achieved a stage 6 validation from HIMSS for its use of data and approach to data science.