Although online information can be useful, medical advice found on the Web is not always trustworthy, even when using symptom checkers.
Many people used to check their symptoms online in the hope of avoiding a trip to the doctor or as a way of finding out how they can self-threat whatever was ailing them, but many times the more people looked the farther the true answer was.
In an effort to help several schools, medical institutions and insurance companies came up with symptom checking software. These could be used by patients in order to find out what illness they suffered from more accurately than using normal internet check.
But a recent large-scale study from the Harvard Medical School analyzed the accuracy of these programs and established that 2 out of three times the first result displayed was wrong.
The research team used symptoms belonging to 45 specific medical scenarios, and tested the accuracy of 23 of the most commonly used symptom checkers wich were developed in the U.S. and Europe.
The results were more than disappointing, when the team found out that only 51 percent of the time the correct diagnosis was even included in the top three results.
Not only do the symptom checkers display inaccurate results 49 percent of the time but many times they fail to even include the right ones in the top 20. Thus 42% of the time it is possible that your right diagnosis will be displayed as the 20th result or even lower.
However the team stated that the symptom checking software was about as accurate as most medical telephone services used by private practices. So while an actual trip to the doctor is still the best way to receive a correct diagnosis, symptom checkers could be a more cost effective temporary solution.
The symptom checkers were also proven to be a bit too prudent, and risky. The software recommended that the user report to a doctor when the symptoms could of been treated with self-adminitered medication.
But in the long run, software development can vastly improve, meaning that it might one day even be more accurate that some doctors.
Machine learning, computer data base interactions and other new software techniques are now being developed to improve existing applications and computer results. If search engines like Google can develop techniques which improve search accuracy the same advancements could power symptom checkers.
Specialists, including the Harvard Medical School research team that conducted the recent study, do continue to recommend the use of symptom checkers but do advice caution and patience.
Even the recent study offers very good feedback which can be used to improve the results displayed by symptom checking software, meaning that as we learn more about what is wrong with them, they become better.
Image Source: mhealthtalk.com
