5 Least Accurate Online Symptom Checkers

Patients are surfing the Internet to self-diagnose their symptoms more than ever.

While some online symptom checkers are reasonably accurate, others appear to provide incorrect diagnostic advice nearly two-thirds of the time.

It is helpful for pharmacists to know which checkers are the least reliable in order to remind patients that the tools are not generally suitable substitutes for medical advice from a qualified health care professional.

A recent study published in BMJ audited 23 online symptom checkers using 45 standardized patient vignettes reflecting conditions that ranged from common to less common and low risk to life threatening.

All symptom checkers examined were free, publicly available, in English, and did not focus on a single type of condition.

Those that failed to list the correct diagnosis within the first 3 results they provided were ranked as follows:

1. BetterMedicine — 71% inaccurate

The BetterMedicine Symptom Checker is a product of HealthGrades, an online service that aids consumers in researching, comparing, and connecting with physicians and other health care professionals.

2. EarlyDoc — 67% inaccurate

EarlyDoc uses guidelines and established scientific research submitted by physicians and nurses to develop an algorithm that analyzes the “complaint check” entered by the user.

3. Esagil — 66% inaccurate

Esagil finds matches between diseases and user-reported symptoms supported by analyses submitted by physicians. After providing a diagnosis, the symptom checker offers drug advice, treatments, and health care centers.

4. Symptomate — 66% inaccurate

Symptomate is based on artificial intelligence algorithms that use medical knowledge submitted by physicians.

5. Symptify — 64% inaccurate

Symptify is managed by emergency department physicians and software engineers that uses a customized algorithmic engine to offer diagnostic advice. Other services include a facility locator, user profile, check-in for facilities, emergency contact list, and consultation history.