Well, now officially: The results of a new study showed that diagnosis via the Internet is in most cases erroneous. Of course, this may seem obvious to someone, but quite a large number of people are diagnosed using Google. That's because searching for symptoms on the Internet — when you really feel bad — is one of the easiest and fastest ways to find information and understand what’s happening. As it turned out, in 67% of cases the diagnosis on the Internet is wrong. We hope that after reading this article you will think twice before you google the symptoms and diagnose yourself.
What is wrong with online diagnostics?
If you do not really understand whySince the researchers generally dealt with this question, the answer is quite simple: in trying to understand what is happening, people google the symptoms and go to the doctor. And this, as you know, should be done if a person is sick. Unfortunately, many neglect this rule and engage in self-medication, which in some cases ends in death. That's why all the symptom information Google gives out must be reliable. But what sites does the search engine produce and can you believe what is said there?
So how to look for symptoms individually and thento summarize their occupation is pretty pointless, most often people use diagnostic sites and applications. The most famous English-language diagnostic service in the world is WebMD, and some Russian-language sites and applications claim that the diagnostic accuracy is almost 70%. The mechanism of operation of most of these services is the use of a large array of data with the conclusions of doctors and statistically make a probable diagnosis.
To always be on top of the latest news from the world of popular science and high technology, subscribe to our channel in Telegram!
In a study published in a journalThe Medical Journal of Australia, scientists from Edith Cowan University (ECU) in Australia, tested the symptoms of 1170 medical reports on 36 available online diagnostic sites to analyze the accuracy of the diagnosis. The results showed that sites showed the correct diagnosis in only 36% of cases. The authors of the paper note that the quality of their diagnostic recommendations varies greatly.
It should be noted that diagnostics - this is not just a one-time assessment, but a process,which requires knowledge, experience, clinical examination, testing and time. If you watched the series “House MD”, then you probably remember how difficult it is to make the right diagnosis. Of course, the series is not real life, but many of the cases shown in it are based on the medical histories of real patients, so a good diagnostician in any case is worth its weight in gold.
The reality is that these websites and applications should be treated very carefully, as they do not see the whole picture. Interestingly, a study published by fiveyears ago, a group of scientists from Harvard University showed almost exactly the same figure - diagnostic services and applications offered the correct diagnosis in only 34% of cases. In general, over the past five years, online symptom testing services have not become more reliable, at least from the point of view of accurately identifying the only, most likely disease that a patient can have.
Do you google the symptoms of various diseases when you feel bad? You can share your medical records or just talk heart to heart here.
Science Alert Cites LeadMichelle Hill, the author of the study: “recommendations for seeking medical care in emergency and urgent cases were appropriate in about 60% of cases, but for emergency situations the result was reduced to 30-40%. Also, diagnostic sites and applications that use artificial intelligence algorithms are more reliable than everyone else. ”
You will be interested: In Russia they create AI based on big data for making diagnoses
Speaking about machine algorithms, it should be notedthe successes that the developers have achieved. So, today medical AI makes 100 diagnoses in 4.8 seconds! Read more in the fascinating material of my colleague Vladimir Kuznetsov. Moreover, modern machine algorithms accurately diagnose such diseases as lung cancer, depression and dementia.
But despite the flaws, the authors of the studyThey believe that, in general, diagnostic services are useful, but only if people use them as an educational resource in combination with the advice of doctors. Scientists note that it is actually very important to teach the algorithms the correct diagnosis - the correct diagnosis will reduce the unnecessary burden on doctors (in cases where you can do it yourself), and also not put a person's life in danger in situations where you can’t do without specialist help . But no matter what advice these algorithms give us, it is important that in the end the doctor himself is not mistaken, do you agree?