Re: What Do We Really Know For Sure?

I believe the term that's used for symptoms that help you narrow down the possible causes is "differential". "A tummy ache is not terribly differential. The presence of a screwdriver extending from the wound is pretty differential." I think there are degrees of how differential symptoms are.

If you have multiple symptoms, they can help to narrow down the possible cause by being considered together. Weakness isn't terribly differential, nor is nausea on its own, but if you combine the two, you move in the direction of narrowing things down. Indeed, the lack of symptoms can be used to help diagnose, too. Sensitivity to light suggests a lot of things, and lacking a malar rash helps to start to rule out lupus, for example.

What's called for is an expert system or similar such computer programs for diagnosing disease. One example: http://www.rawstory.com/rs/2013/02/12/ai-system-diagnoses-illnesses-better-th...


But I think this only really works for data that the doctors have already validated and doesn't approach handling diseases that aren't well understood or documented.

This article points out a few online diagnostic sites, but none of them really seems to be doing a good job. I recently saw mention of a company that makes an app for diagnosis, Isabel. I don't know how good it is, haven't looked at it yet.

I have ideas for handling illnesses that aren't well document or understood.

I think it should be possible to create a program, a website, where you can enter your symptoms and then test out different remedies and report your success or failure, and thus collectively, with many users, generate statistics that help diagnose and treat illness.

"Your symptoms match very well with this particular cluster of symptoms. Other people who matched well with this cluster of symptoms found improvement with the following treatments..." Tweet