Google 911

The use of online search tools to find health information has spurred a debate about how to tap their power without sacrificing the precision required when lives are at stake

An estimated eight out of 10 people with Internet access look for health information online at one time or another, so it’s no surprise that companies seeking to deliver Web-based health services have begun to focus on search tools.

Google, whose name has become synonymous with search, made its first formal foray into that market in May 2006 when it launched Google Co-op, a platform that allows Web developers to create customized search engines. It named the health sector as one of the first vertical markets it would focus on.

Many observers had expected more, possibly a separate Google health business. Instead, the company will help trusted medical sources find and tag reliable information to which Google will guide users.

Perhaps officials at Google realized how tough it is to build a business around medical search technology. As Adam Bosworth, the company’s vice president of engineering, wrote in a recent post on Google’s blog site, it’s difficult to know what is trustworthy and what isn’t when it comes to online medical information.

“Search is great at finding us places with relevant information, but it is hard to know which links are reliable and which are less so,” he wrote. “Honestly, this is a hard problem.”

Google officials declined to be interviewed for this story.

Search technology is nothing new to online health sites. Specialized information providers such as WebMD and the National Library of Medicine’s MedlinePlus have included search as a central component of the services they offer for a number of years.

However, such organizations use a relatively narrow set of clinical and academic sources for the information they offer through their search engines. Google, on the other hand, provides access to all the information on the Web that its technology can collect and tag.

Granularity vs. convenience
For better or worse, most people see Google as the trendsetter for health search technology, especially for general consumers.

“Usability studies show us that people want this kind of search to be as easy as possible, and they prefer to use Google,” said Joyce Backus, deputy chief of the National Library of Medicine’s Public Services Division. “It’s a matter of convenience for them.”

The library was one of the first organizations to partner with Google Co-op to help identify Web sites and pages that have reliable information.

But the health care domain is not as straightforward as others, said Tom Eng, founder, chairman and chief executive officer of Healia, which offers a personalized online search engine for health information. The Veterans Health Administration is one of the company’s high-profile customers.

“Health information is highly complex, and unlike other kinds of information, the consequences of not getting high-quality information are very severe,” he said. “General search engines just can’t do it with enough granularity.”

Healia’s engine uses semantic search technology, which allows computers to understand what search terms mean and relate those terms to specific medical concepts. It also evaluates search results according to a quality index that rates information based on certain characteristics of its content.

As a result, Eng said, users receive search results that provide more reliable — and relevant — information.

“Much of the information about such things as diagnoses and test results differ biologically from person to person,” he said. “Unless you can provide a person with that level of control, then the information is not necessarily all that useful to them.”

Health information entropy
Google’s keyword-based approach to searching produces results that don’t reflect the relevance and richness of medical data, said George Krucik, senior product manager at Healthline Networks.

Healthline’s search engine, on the other hand, takes a taxonomic approach to medical searching, ranking the relevance of information according to factors such as how it links to other information in answering queries related to specific illnesses.

Taxonomic searching also ensures more consistent results over time, he said.

“The information on the Web is always changing, and its entropy is always increasing,” Krucik said. “You get different results from the same search on Google from one day to another.”

Developers often overlook the importance of creating search tools that focus on returning relevant results, said Dr. Edward Fotsch, CEO of Medem. The company has developed a suite of Web-based communication services that help create interactive connections between physicians and patients.

It’s less about the sophistication of the search than it is about relating the results to the person conducting it.

“For example, the results for a 69-year-old husband of a person with coronary disease should be different from those for a 29-year-old mother with a kid who has sports asthma,” Fotsch said. “The first one may be looking for a treatment while the second is searching for medication, and it would be nice to know that.”

But the process isn’t that intelligent yet, he said, adding that right now “99.9 percent of search is dumb search.”

Contextualizing health searches
The problem can’t be solved in isolation, Fotsch said, because search technology is already a fixture in the world of online health information. It’s the No. 1 tool for consumers to find information, and he said he believes that’s likely also the case for physicians.

“It’s simply the quickest way to get from here to there,” Fotsch said. “What’s shifting is how people see search [on the Internet] and in what context, and how can we customize search in order to get people exactly the health information they want.”

The market for medical search technology is strong, agreed West Shell, Healthline’s chairman and CEO, adding that online health information is in the process of making an exponential leap forward.

“A health information war is breaking out,” he said. “We’ve had focused health sites that people had to go to before to get information, but that’s rapidly changing. There’s a lot more consumer health information at a lot more places.”

Every consumer-oriented Web site now has health information as one of its top three services, along with financial and entertainment news, he said. It’s a matter of figuring out how to take this increasingly complex subject and deliver the best, most reliable information to consumers and health professionals.

“It’s very hard for either [audience] to filter the junk from the high-quality information and, even when they do get to high-quality information, to sort that out into what is relevant for a particular situation,” Eng said. “You are really trying to produce the equivalent of what a knowledgeable and trusted health librarian can provide.”

It will take some time for search technology developers to resolve all the issues, and even though specialty tools will provide some of the answers, most experts expect Google to lead the way.

If you take Bosworth at his word, however, Google might not be doing that anytime soon. He wrote in his blog that although questions about finding information might seem simple, when it comes to life-or-death issues, they are also profound.

“I’d like to say that we have all the answers,” he said. “But we don’t. Mostly, at the moment, what we have [are] questions.”

Defining the health search universeGoogle Co-op’s health search tool relies on the keyword-based approach the company popularized. Users type in words related to illnesses, symptoms or health topics and refine the terms if necessary based on the search results they get.

However, Google recognizes that not all sites are trustworthy sources of medical information and has recruited the likes of the Mayo Clinic, the National Library of Medicine, the Centers for Disease Control and Prevention, and the Medical Library Association to help it evaluate and tag Web sites as trusted resources.

In time, so the theory goes, users will consistently go to those trusted, tagged and annotated sites in their search for reliable health information.

But that approach assumes that users will be able to comprehend the information when they find it. And that poses a problem. Several years ago, the Institute of Medicine found that one in three Americans could not understand even basic medical instructions.

Therefore, health literacy has become a necessary consideration in any effort to improve search tools.

The World Health Organization defines health literacy as the cognitive and social skills that “determine the motivation and ability of individuals to gain access to, understand and use information in ways [that] promote and maintain good health.”

The National Library of Medicine, a major source of health information in the United States, offers an online course that walks users through what to look for when they visit sites and how to recognize trustworthy information.

So far, such attempts at online search education are limited. But most experts agree that this kind of user knowledge will be vital, even as health search technology improves.

— Brian Robinson

Featured

  • FCW PERSPECTIVES
    sensor network (agsandrew/Shutterstock.com)

    Are agencies really ready for EIS?

    The telecom contract has the potential to reinvent IT infrastructure, but finding the bandwidth to take full advantage could prove difficult.

  • People
    Dave Powner, GAO

    Dave Powner audits the state of federal IT

    The GAO director of information technology issues is leaving government after 16 years. On his way out the door, Dave Powner details how far govtech has come in the past two decades and flags the most critical issues he sees facing federal IT leaders.

  • FCW Illustration.  Original Images: Shutterstock, Airbnb

    Should federal contracting be more like Airbnb?

    Steve Kelman believes a lighter touch and a bit more trust could transform today's compliance culture.

Stay Connected

FCW Update

Sign up for our newsletter.

I agree to this site's Privacy Policy.