When doctors proved unable to cure her Lyme disease, computer scientist Jennifer Mankoff did what many people do when they're chronically ill: She went online to see what medical organizations had to say about it.

There the Carnegie Mellon University associate professor in the Human-Computer Interaction Institute found conflicting information and other users as befuddled by that as she was.

The majority of the sites said the tick-borne bacterial disease is easily treated and cured with antibiotics, especially if caught early. But, some of those sites said, if the disease does not respond to the drugs after a limited period of time -- the Centers for Disease Control and Prevention recommend no more than two, two- to four-week courses of antibiotics -- no cure is possible. "You manage the symptoms," said Ms. Mankoff, 37, of Shadyside.

Other sites called long-term Lyme disease "chronic" and suggested it might be cured by much longer and varied therapy.

Her computer searches showed doctors on the extreme ends of the controversy.

Eventually, Ms. Mankoff found a doctor who agreed to do long-term treatment and after 18 months of antibiotics she felt well enough to stop the drugs. That was in October 2009.

During that time, she managed to get tenured, write a proposal for a research study, and start the paper with four colleagues. The topic: how longtime Lyme disease sufferers handled and were affected by the competing information online.

The study, on which she is lead author, has been accepted for presentation at the 2011 Computer Human Interaction Conference in Vancouver, Canada, next month.

Her inspiration, she said, "was the controversy, what a huge influence it had on my experience with the illness."

Online, she said, Lyme patients "encounter all kinds of different things: People who want to sell a cure; people who want to sell a story. ...

"We have to address that experience, all that variety," she added. That's what is next for her and her research colleagues: online tools that would permit people already skilled in their computer searches to do even more sophisticated ones.

The study comprised 150 participants with long-term disease who were surveyed; 140 of them described their online activities, and 128 provided demographic information. All of them had been sick for months or years, and 71 of the total sample were so sick they were unable to work. The median time since the participants were first symptomatic was six to 10 years, and they had been diagnosed a median of one to two years earlier. A total of 103 of them had seen an average of 7.8 doctors before diagnosis.

The participants were divided into two groups: those symptomatic for five or fewer years and those who were sick longer. All of them had visited Lyme-specific sites more often than general medical sites, but those who had been sick longer visited the Lyme-specific sites more than those sicker for less time.

The Lyme-specific sites were divided into three models, or categories: dominant (the sites which call some long-term cases post Lyme-disease syndrome and incurable); minority (which say antibiotic treatment may need to be extended over months); and alternative (like those based on non-Western medicine).

Twenty of the original participants then were interviewed by telephone for more information. Ten of them told the researchers that at first they had accepted the information associated with the dominant sites and resisted contradictory viewpoints. Those 10 had been diagnosed in half the time -- an average of five years -- of the other 10, who had no clear model of belief at diagnosis time.

Eventually, however, all came around to believe in either the minority or the alternative categories.

Because so many of the participants shifted in their beliefs, the researchers focused on the individual experiences that led to those changes. Those experiences included interaction online with other Lyme sufferers.

What the researchers found, in addition to the conversion in beliefs, were changes in the Lyme patients' self-identities and in how they explained the illness to themselves. The paper said, for example, the failure of dominant doctors to help her son eventually drove one woman to take him to a minority doctor and believe in that category because that treatment was successful.

A second finding was "a level of sophistication in using computer information that went beyond what we expected," Ms. Mankoff said, and a third finding was a high level of activism on behalf of the Lyme community. "People were committed to providing social support," she added.

The researchers' main conclusion: "We believe it's important to expose multiple understandings of the disease when they exist -- if we can -- so they can lead to more informed and then better decisions," Ms. Mankoff said.

Next, she said, the researchers will try to build the online tools that can be used in conjunction with a Google search to provide additional, specific information about a search result's content.