When someone comes down with a new health symptom, they are typically advised not to self-diagnose using the internet. Resorting to Dr. Google could lead to inaccurate diagnosis, incorrect treatments or increased anxiety.

But new research published in JAMA Network Open suggests that online investigations for health information actually lead to slightly more accurate diagnoses.

TOPLINE DATA

Participants’ diagnosis accuracy improved by 4.2% before and after searching the internet for health information related to a provided case description. Before the internet search, 49.8% of participants correctly guessed the health condition being described. Afterward, that figure jumped to 54%.

Improvements in diagnostic accuracy occurred across all the forms of triage categories provided: emergent (3.1%), same-day (3.5%), same-week (6.4%) and self-care (3.7%).

“Results of this survey study challenge the common belief among clinicians and policymakers that using the internet to search for health information is harmful,” researchers said in the study. “We found that performing an internet search was associated with improved diagnosis.”

The participants spent an average of 12 minutes researching the symptoms before making a concluding diagnosis.

Although diagnostic accuracy improved across the board, triage accuracy and anxiety levels remained unchanged from before and after the internet search.

Participants correctly identified the necessary triage option 74.5% of the time before the internet search and 74.1% of the time afterward. Their reported anxiety levels remained at a 3.3 on a five-point scale throughout the intervention.

Finding useful information was “slightly difficult” for participants, but they said the most helpful online resources were search engines (48.2%) and health specialty sites (42.9%).

The study also found that most people’s diagnosis (85.1%) and triage (87.2%) selection remained the same before and after searching the web. Of the participants who switched their diagnosis, 9.6% changed from incorrect to correct and 5.4% changed from correct to incorrect. Similarly for triage, 6.2% went from incorrect to correct and 6.6% changed from correct to incorrect.

“Another explanation for the small before-after changes that we observed is anchoring,” the researchers said. “Only a small fraction of respondents changed their diagnosis or triage decision after the internet search. Consistent with the theory of reinforcement seeking, internet searchers may simply look for information to justify their initial decision rather than being open to all recommendations.”

METHODS

Five thousand participants with a mean age of 45 years enrolled in the study. Of these, 51% were female, 76.4% were white, 49.7% had private insurance, 14.7% had poor or fair perceived health, and 44.1% had at least one chronic condition.

The researchers created 48 different case vignettes across the four different triage categories. The conditions included in the descriptions ranged from common illnesses like a viral infection to severe conditions like a heart attack.

Participants were each given a case vignette and told to select the necessary level of triage, list and rank the three most likely diagnoses, and rank their anxiety and confidence levels based on if the health problem was happening to a close family or friend.

From there, they were instructed to search the internet for the correct diagnostic and triage designations. Their anxiety and confidence levels were then re-measured along with their perceived difficulty in finding useful information, their trust in the information they found and the kinds of websites that they deemed to be most helpful.

THE LARGER TREND

Online symptom-checking is a fairly common practice among U.S. adults. Between 2012 and 2013, 72% had taken to the internet to look up health information, and 35% were classified as “online diagnosers,” according to The Pew Research Center.

Luckily for these online sleuths, websites such as WebMD have taken steps to make their symptom-checkers more accurate, and there are numerous apps on the market that can help people diagnose and triage their symptoms.

The heavily funded symptom consultation app Babylon Health recently sold off its Canadian operations to former partner Telus Health in a licensing deal projected to be worth roughly $70 million.

Other services include K Health, which closed a $132 million Series E in January, and Infermedica, which landed $10.25 million in its Series A round last summer.

Although these apps provide a convenient way to pinpoint what condition a person’s symptoms could be a sign for, a 2018 review of direct-to-consumer self-diagnostic apps found that they vary widely in functionality, accuracy, safety and effectiveness.