It’s estimated that over a third of the US population regularly surf the internet in an attempt at self-diagnosis. Although it’s often been claimed that such surfing leads to confusion and unwarranted increased anxiety, “cyberchondria”, it has been difficult to properly study, until now.
In doing an internet search, symptoms are typed into search engines such as Google or Yahoo, which leads to websites with conditions matching the symptoms. Recently, online symptom checkers have been introduced that rely on specific algorithms to establish a diagnosis and the urgency of treatment. Algorithms are often based on triage telephone protocols employed in physician offices.
A study published in the July issue of the British Medical Journal, is the first to evaluate 24 popular symptom checkers. Researchers from the Harvard Medical School concluded that online symptom checkers provide the correct diagnosis 34% of the time, while listing it among 20 others, 58% of the time. Online symptom checkers varied in their diagnostic accuracy, with AskMD, DocResponse, Isabel and Family Doctor showing somewhat greater accuracy.
Regarding advice on further treatment, online symptom checkers were correct in about half of the 45 standard symptom scenarios. Although with emergency conditions, online symptom checkers were correct in 80% of scenarios, while in symptom scenarios requiring no care, online symptom checkers suggested medical treatment two-thirds of the time.
Study findings suggest that current online symptom checkers contribute to confusion regarding diagnosis, “cyberchondria”, over-treatment and increased health care costs. Nevertheless, online symptom checkers are better than simply surfing the internet and when a medical emergency is suspected, the wiser course is to over respond rather than wait. With rapid technological advances, there is hope that the next generation of online symptom checkers will demonstrate greater benefit.
– Jack Klausner