Slashdot Mirror

Even a simple db that cross references diseases to symptoms / blood work results (and other test results) doesn't seem to exist.

They do exist. See the Problem / Knowledge Coupler.

you'd think that doctors could order up a set of tests based on their initial thoughts, input the results to a program, and have the program guide them with possibilities to try and narrow down the search of what may be wrong.

That assumes that we have enough data to link a test or a series of tests / symptoms / findings to a disease. That also assumes that we know the disease the patient has. Neither assumption is true. We've really just cleaned off the low hanging fruit in this respect.

IF Watson has enough data programmed into it then it might be the kind of system that could answer your question. However, it seems to be attempting the same thing that PKC dose and that system hasn't really been the breakthrough that it's inventors had hoped for.

There is just a lot of stuff about the human body that we don't know....

Sounds like you're thinking of Larry Weed's Problem-Knowledge couplers. These have been around for literally decades - they were the next big thing when I was in medical school (early 80's). Apparently it hasn't died yet, but every time I look at them, I find that they are 1) slow 2) SLOW 3) tend to look for unusual / uncommon diseases and 4) not any better than hitting a textbook (or these days Google).

The problem with computers sorting on symptoms is that symptoms are common, symptoms overlap diseases and there are a lot of unconscious filters that medical professionals use to get the diagnosis in the ball park that have yet to be fully rationalized. They could be useful as an adjunct to jog your memory or in a training, but they aren't going to replace anyone.

IMHO, the biggest problem with these sorts of programs is that humans aren't rational and aren't strictly deterministic. Applying a strictly rational, deterministic approach to "fixing" people only goes so far.

However, I googled him recently out of curiosity to see what had happened to him and his ideas, 20+ years on. I have to say that the results were remarkably disappointing. It seems that Larry Weed was too far ahead of his time and may still be. Check out his company at http://www.pkc.com./ One rather dramatic quote of his: if physicians ran airports they wouldn't have radar, just lots of intensive care units around the periphery.

Go to http://www.pkc.com to find out even more about these knowledge couplers.

For one, these things are Windows software, with a web access version. Also, you subscribe to these things to get the semi-annual updates.

Another problem that I actually do sympathize with is that this is grounds for serious lawsuits.

Yeah, I'm surprised how much this has been discounted. Although PKC makes a point about calling recommendations "primary options" to distinguish from hard advice and transfer responsibility to the provider, there is still a high degree of editorial decisions that the software publisher is making.

It's unclear from the story or their website as to how conflicting research is treated. Are different journals or studies weighed differently? How about incorporating common clinical practice that might not show up in the published articles? Or academic theories that are widely used for diagnosis and treatment but have not been conclusively proven?

In their FAQ they state that "Every question contained in a Coupler is there because an action recommended by the medical literature depends upon its answer". What if there are no conclusive actionable recommendations, but results can be inferred from related studies? Does that mean that there are any gaping holes in their content? Is this counter to a systemic approach?

Good software should be able to handle all of these questions, provide a usable and adaptive user interface, avoid unnecessary data entry, provide for local input and overrides, and maintain primary responsibility with the MD. So far, besides some good results, we have no evidence as to the quality of their program (although a tour is available on their site). And what happens if PKC is so widely adopted that it becomes the Microsoft of medical software or the AOL of medical content?

As they are currently positioned they shouldn't so causally brush off their legal exposure. They might be better off repositioning the same service as a faster and more accurate method of accessing research and best practise rather than as a guided decision making tool. This might alleviate the fears of certain doctors while also stressing that the decision lies with the practioner and the advice comes from third parties.

Another problem that I actually do sympathize with is that this is grounds for serious lawsuits.

Yeah, I'm surprised how much this has been discounted. Although PKC makes a point about calling recommendations "primary options" to distinguish from hard advice and transfer responsibility to the provider, there is still a high degree of editorial decisions that the software publisher is making.

It's unclear from the story or their website as to how conflicting research is treated. Are different journals or studies weighed differently? How about incorporating common clinical practice that might not show up in the published articles? Or academic theories that are widely used for diagnosis and treatment but have not been conclusively proven?

In their FAQ they state that "Every question contained in a Coupler is there because an action recommended by the medical literature depends upon its answer". What if there are no conclusive actionable recommendations, but results can be inferred from related studies? Does that mean that there are any gaping holes in their content? Is this counter to a systemic approach?

Good software should be able to handle all of these questions, provide a usable and adaptive user interface, avoid unnecessary data entry, provide for local input and overrides, and maintain primary responsibility with the MD. So far, besides some good results, we have no evidence as to the quality of their program (although a tour is available on their site). And what happens if PKC is so widely adopted that it becomes the Microsoft of medical software or the AOL of medical content?

As they are currently positioned they shouldn't so causally brush off their legal exposure. They might be better off repositioning the same service as a faster and more accurate method of accessing research and best practise rather than as a guided decision making tool. This might alleviate the fears of certain doctors while also stressing that the decision lies with the practioner and the advice comes from third parties.

Another problem that I actually do sympathize with is that this is grounds for serious lawsuits.

Yeah, I'm surprised how much this has been discounted. Although PKC makes a point about calling recommendations "primary options" to distinguish from hard advice and transfer responsibility to the provider, there is still a high degree of editorial decisions that the software publisher is making.

It's unclear from the story or their website as to how conflicting research is treated. Are different journals or studies weighed differently? How about incorporating common clinical practice that might not show up in the published articles? Or academic theories that are widely used for diagnosis and treatment but have not been conclusively proven?

In their FAQ they state that "Every question contained in a Coupler is there because an action recommended by the medical literature depends upon its answer". What if there are no conclusive actionable recommendations, but results can be inferred from related studies? Does that mean that there are any gaping holes in their content? Is this counter to a systemic approach?

Good software should be able to handle all of these questions, provide a usable and adaptive user interface, avoid unnecessary data entry, provide for local input and overrides, and maintain primary responsibility with the MD. So far, besides some good results, we have no evidence as to the quality of their program (although a tour is available on their site). And what happens if PKC is so widely adopted that it becomes the Microsoft of medical software or the AOL of medical content?

As they are currently positioned they shouldn't so causally brush off their legal exposure. They might be better off repositioning the same service as a faster and more accurate method of accessing research and best practise rather than as a guided decision making tool. This might alleviate the fears of certain doctors while also stressing that the decision lies with the practioner and the advice comes from third parties.