Wikimedica
Medical doctors spend a lot of their time doing pattern matching. Specifically, matching symptoms (combined with other data, like age and sex) against maladies to produce a diagnosis, and from that determining a treatment. Much of a med student’s studying is essentially memorization of long lists of this stuff: symptoms -> malady -> treatment. Much of the work of an interned doctor is trying to improve their recall of these patterns in real-world situations. And finally, much of the work of practicing doctors is applying these patterns to their patients over and over again.
This is pretty dumb.
The process here is really nothing more than pattern-matching with an extremely large and complex database. But humans aren’t particularly good at either storing data or searching it. The intuitive leaps that a human can make are useful in a minority of cases (the show House M.D. revolves around this exception). Yet, most patients would be better off with the brute-force search capability of a powerful computer.
Why hasn’t anyone made a giant database of symptoms cross-referenced by demographics, which offers a simple interface for entering patient parameters and producing a list of results, weighted by probability? This is not a terribly hard programming task, and although keying in the data would be a challenge, all of the necessary knowledge already exists and is well-documented in medical texts (the same ones those med students spend their youth committing to memory). Certainly the value to humanity would be enormous, and it has a profit potential to match.
My guess is that the regulations surrounding dispensing of medical advice are rigid enough to deter anyone from trying to build such a tool. Moreover, the established medical orthodoxy (mainly the AMA) has proven itself to be extremely resistant to anything which would replace activities currently performed by a doctor.
My view is that freeing up doctors’ time to focus on the truly challenging cases that can’t easily be diagnosed with a simple pattern match would be a net win for everyone. Given the shortage of doctors in the last decade or two, any such technology becomes even more useful.
Maybe it’s just that no one has gotten around to doing it yet, in which case you might argue, shouldn’t I stop complaining and start building the thing? Well, I just might - though, having just started a new venture, it would have to be a few years down the road.
Or maybe it does already exist, and I just don’t know about it. If that’s true, then I have two questions. First, why isn’t it a standard part of everyone’s life? The first thing anyone should do before seeing a doctor is to search the database. This would serve as a screening system that would prevent people from wasting a doctor’s time with an easy-to-diagnose issue. And the second question: what’s the URL?
November 26th, 2007 12:59 am
This is something similar to what you are talking about… but they seem to have changed it into something less useful. I used to use it all the time when it made sense.
http://symptoms.wrongdiagnosis.com/
November 26th, 2007 6:00 pm
Patients are tending to take things into their own hands these days, investigating their symptoms and developing their own second opinions, refining their search for the best care. Today, it’s not uncommon for a doctor to meet a patient with a ream of printouts describing their condition, and requesting specific drugs.
Even if a central database were to exist, a human will still need to sign off on the treatment it recommends. Who will be sued if the diagnosis is incorrect? If the answer is “the doctor”, then it behooves the doctor to cross reference the system against their education and other resources, and I wonder if the savings in time would be so great. A nice feature might be a networking tool that would allow other professionals to see who has searched a set of symptoms most recently, or connect generalists to specialists.
My suspicion is that such a reference would certainly be handy (most likely for medical students, or in cases where the symptoms are especially obscure) but can’t and shouldn’t diminish the time a doctor spends interacting with any patient. Except the annoying ones who, you know, have printed up a website about lymphatic cancer when really what they have is the flu.
November 19th, 2020 12:18 pm
erectile dysfunction medications
Adam @ Dusk » Wikimedica
November 19th, 2020 12:18 pm
erectile dysfunction medications
Adam @ Dusk » Wikimedica