Handful of sites top search list for medical info

Health information on the Internet has seemed as untamed and difficult to navigate as the Wild West. But one study shows that patients are settling down in that frontier, relying mostly on the same few sites to learn about a particular condition or drug.

The study, released in February by Envision Solutions, a New York-based health marketing company, found these sites are run by various sources, from associations to drug manufacturers to retailers. But the research also showed plenty of users flocking to so-called user-generated media -- sites, including those maintained by physicians, with no official ties or formal organizational oversight.

Envision's study also looked at which sites appeared on the first three pages of a Google or Yahoo search for specific health terms, with the company determining that sites on those pages are the ones patients will most likely check out -- thus making it fairly easy for physicians to gauge Web activity in their specialty.

Physicians "need to be aware of where people are getting this information and make sure it's accurate," said Fard Johnmar, founder of Envision Solutions.

Others say physicians are better served by recommending sites they trust to patients rather than monitoring the Web. However, some physicians say it is a good idea to talk to patients about their Internet research. Doctors might even find their patients can steer them to some new, reliable sites, said American Psychiatric Assn. Vice President Nada Stotland, MD, MPH, a professor of psychiatry and obstetrics-gynecology at Rush Medical College in Chicago.

Where patients are led

The Envision study was conducted in two steps. First, in November and December 2006, Envision typed eight common, health-related terms into the Google and Yahoo search engines: bipolar disorder, Blue Cross Blue Shield, Chantix, depression, diabetes, herpes, Kaiser Permanente and Lexapro.

Envision was particularly interested in whether user-generated media appeared in the first three pages of the searches. In seven of eight cases, it did. Blue Cross Blue Shield was the only search term in which it did not. On the first page of the search, Wikipedia entries represented 62.5% of the user-generated media, according to Envision. Bulletin boards -- including Web sites that linked to a separate page where users could write public messages -- represented 25%, and blogs represented 12.5% ("other" was 18%).

Lydia Worthington, vice president of client services for Nielsen BuzzMetrics, a New York-based, consumer-generated-media research firm, said some people think user-generated sites are less biased because they are written by people with firsthand experience and no vested interest with a particular issue. Matt Petersen, director of information technology for the American Diabetes Assn., said he has found user-generated sites accurate.

But many still question such sites' trustworthiness and are particularly worried about Wikipedia. The encyclopedia-modeled site allows users to write and edit entries without identifying themselves or presenting credentials. The site relies on "the wisdom of crowds," a theory that even with factual hiccups, the shared knowledge of many creates better information than the expert knowledge of one.

Since the journal Nature in late 2005 published a study saying Wikipedia was about as accurate as the Encyclopaedia Britannica when it came to science, others have done their own research. A peer-reviewed study looking at history topics said, indeed, that field was accurately represented. A professor at Connecticut's Quinnipiac University intentionally put 13 errors on Wikipedia and found them all corrected within three hours.

Similar studies are not yet available regarding medical topics. But Wikipedia itself says anybody going to its site should fact-check information elsewhere. Sandra Ordonez, a spokeswoman for Wikipedia's parent company Wikimedia, says with medical information, the final word should come from a physician.

Where patients go

The second part of Envision's study looked beyond the search engines to find out where patients actually go to look for health information on the Web. Envision used Hitwise, a company that tracks Internet site use, to determine the percentage of users who, on Jan. 13, clicked on each site that popped up on Google and Yahoo.

Generally, Envision found, more than half of Internet users ended up on one of five sites. For example, 54% of the traffic resulting from the search term "diabetes" went to five sites, while 58% of those searching for "bipolar disorder" ended up on the same number of sites. For the search term "Lexapro," 78.9% of the traffic went to five sites.

Most of that traffic ended up at "official" sites, the most heavily trafficked sites that end up high on the Google or Yahoo search list, according to Envision. For diabetes, 31.9% of all traffic went to the American Diabetes Assn. Web site, while 7.9% went to the Centers for Disease Control and Prevention's Web site. Drug company sites also ended up with a lot of traffic. About 50.6% of all Lexapro traffic went to Forest Pharmaceuticals' official Lexapro site. GlaxoSmithKline's diabetes.com and bipolar.com each got between 5% to 6% of all traffic from the search terms "diabetes" and "bipolar disorder," respectively.

But user-generated sites also ranked in the top five. The fourth-ranked site for Lexapro traffic was crazymeds.org, a blog and message board run by a couple from Montana. Jerod Poore, one of the site's founders, says people have come to his site seeking refuge from drug company-sponsored information, "fear-mongering" sites saying all drugs are bad, and "tea and sympathy" sites that let users vent but don't share research and practical information.

"We're more relevant to the consumer because we're more accessible. We translate doctorese into English," Poore wrote in an e-mail to American Medical News. "We've had the misfortune of firsthand experience with the majority of medications written up on the site. That is, until we've found the right cocktail of meds that actually work. Talk about the proverbial long, strange trip."

While Envision's Johnmar recommends doctors do their own Google and Yahoo searches to see what patients might be seeing, physicians themselves say they have plenty to stay on top of without trying to monitor the Internet.

"The best approach to patients who use the Internet is to listen carefully to what their concerns are and what they have read and try to ascertain what sites they use to get their information," said AMA Board of Trustees member Joseph Heyman, MD, an ob-gyn in Amesbury, Mass. "If they use a reliable site, you can just agree with them [that] what they are planning on is a good idea."