What's that symptom? Experts warn of self-diagnosis via the web

Many people prone to leap to worst-case conclusions when investigating own ailments, study finds.

28 july 2012-- Got a weird ache or pain? A rash that's hung around too long? With the wealth of information now available at the click of a mouse, it's common to search the Web to figure out what may be wrong with you.

But a new study suggests that when the symptoms are your own, "self-diagnosis" via the Internet -- or anywhere but a doctor's office -- too often leads to inaccurate, worst-case conclusions.

Research suggests that people tend to overestimate their own risk for serious ailments, in a way that they wouldn't do if they were thinking about someone else's symptoms.

"This is particularly true when the disease is rare," said study co-author Dengfeng Yan, a doctoral student at the Hong Kong University of Science and Technology. "That is, given the same set of symptoms, people will overestimate their own likelihood of getting such rare (often serious) diseases than that of other people."

The study was published online recently in the Journal of Consumer Research and will appear in the journal's February 2013 print issue.

In a series of six experiments, Yan and co-researcher Jaideep Sengupta gave nearly 250 college students information about such diseases as flu, HIV, osteoporosis and breast cancer.

In the experiment on flu, for example, the students were told they were simply being tested on what they knew about the illness. But the researchers then told them to imagine that they had a number of symptoms -- cough, fever, runny nose, headache -- and then asked whether they thought they had "regular" flu or the (at the time) rarer and scarier H1N1 "swine" flu. They were also asked to diagnose a hypothetical "someone else" with these symptoms.

According to the researchers, when referring to their own symptoms people were much quicker to diagnose the rarer H1N1 strain of flu compared to if they thought the symptoms were someone else's. In other words, their accuracy in correctly gauging the likelihood of H1N1 infection dropped if they were thinking about their own symptoms.

Why the difference? "We explain these effects using the concept of psychological distance," Yan said. With distancing -- in other words, thinking about someone else rather than yourself -- people tend to rely more on broader information such as statistics (that is, the likelihood any one person will get disease X) and less on information specific to the individual, such as the symptoms he or she is having.

"Consumers often fear the worst when it comes to their own health, while maintaining a calm objectivity with regard to others," Yan said. If you've got pain in the chest, you think: heart attack. If a friend of a friend has the same symptoms, you say: probably indigestion.

But this type of thinking can have downsides, the authors write, often leading "to mistakenly diagnosing oneself as possessing a serious disease, causing both unnecessary anxiety and wasteful medical expenditure." They add that, "mistaken self-diagnoses of this sort are particularly likely given the ease of information access on the Internet, which frequently leads consumers to engage in 'symptom-matching' exercises."

The findings ring true with Dr. Peter Galier, an internal medicine specialist at Santa Monica-UCLA Medical Center, who reviewed the findings.

"I think it's human nature," he said. People are much more likely to worry about themselves and family members, he said, than about strangers.

Getting information via the Web can also make it difficult to decide what symptoms mean in the absence of a doctor's analysis. "When people are able to access a lot of information that isn't filtered, and they don't have expertise in the field, they don't know how to prioritize the information," Galier explained.

That's true whether you're investigating that strange knocking noise in your car but have no clue about mechanics, he said, or whether you've just spotted a lump in your thigh and have no medical expertise.

For instance, a healthy young man may have just drank a very cold soda and then has crushing chest pain. If he looks that up online, he is likely to see that the symptoms may point to a heart attack, Galier said. But that information doesn't take vital facts such as the man's age and medical history into account, he said. "Chest pain in a 55-year-old guy is looked at much differently than in a 25-year-old guy," he said.

Instead of doing amateur diagnosing on their own, Yan said, "We advise people to see a real doctor."

And as the new study points out, ''The advantage of seeing a real doctor isn't just because he or she is an expert," he said. "It's also that they aren't you."

For that reason, they will take into account less "emotional" information, such as how many people in the population actually get the condition you worry that you might have.

And what about that worrying time before you can get to your doctor's office? "People may simply tell the situation to their colleagues," Yan said, since they're likely to put those symptoms into a less panicked perspective.