Comment by j_lev | Hacker News Reader

Comment by j_lev | Hacker News Reader

Parent Full thread

Some questions:

1) How far along are we in terms of those computers that were predicted in the 80s to one day replace doctors? Not trying to antagonize but I can't see any harm in having a row of booths with terminals that patients can log into and start answering questions about symptoms, being led through the "troubleshooting" steps, etc. There's a lot of time spent just waiting around in hospitals, waiting rooms, etc. Maybe you could have a swipe card that saves your progress on the terminal, combines it with the lab results. The system wouldn't replace a GP just yet but it would instead work alongside the GP.

2) Are there any crowd-sourced solutions available yet for these cases where doctors clearly have no idea? Someone enters all their symptoms and people discuss online and try to come up with some possible options.

3) How much are insurance companies responsible for the sorry state of health care in the US? I was speaking to my friend who says that he pays $850 bucks a month (up from $800 a month last year) on top of what his company pays. I've heard other stories about people basically being a slave to their company because they can't leave due to the generous benefits. I've also heard that insurance companies pay back only 60 cents in the dollar, so to make up the same $100 a doctor was previously earning he now has to charge about $170 (the other option is to be cut from the network of doctors used by the insurance company). And is this the logical conclusion for health care? Will Canada/Australia head the same direction? Seriously, there is a massive opportunity for the US to fix some of these problems now, yet it just seems that you guys are heading in entirely the opposite direction.

(working in a health IT startup)

On 1:

So, the big thing is that you never actually want to replace the doctors. Their egos won't allow that, they'll see it as an existential threat (which it is), and most importantly: you don't want to give that kind of power to a device which doesn't function as a moral agent.

On 2:

Sometimes, you'll be getting advice from somebody like tptacek...sometimes, you'll be getting advice from losethos. I wouldn't trust my healthcare to random bozos on the internet.

On 3:

Almost all of it is their problem, I'd say.

The increased focus on billing and everything else means that the doctors have to deal with madness they wouldn't have to otherwise. The fighting and billing cycle can take literally a year, so no honest doctor is at an advantage not billing everything they possibly can, and any institution (read: hospital) is going to force that sort of hand-wavy arithmetic (because of the opposition of insurance providers, in turn because of previous bad actors).

The amount of money thrown into these maws anyways (from both insurance and Medicare, a weird issue unto itself) serves to raise the price floors so that you can't even pay for yourself if you want to, because you're literally on a different marketplace.

The companies also go and force you to use doctors that aren't necessarily convenient to you, maybe many miles away. If you try to use a local physician, well, tough shit, you might have to pay for them anyways.

I could go on and on and on and on...

It's a fractal of sadness.

Thanks for the insights

Much of what doctors do could and has been written as expert rule-based systems. You might find the 1980s Mycin program for diagnosing different kinds of infections interesting; it was said to outperform members of Stanford medical school faculty in the domain that it addresses.

One problem with this style of tech is that your program is only as smart as the rules you give it, and you need seasoned doctors to help put together this knowledge. Even if they had the time, what doctor wants to facilitate the commoditization and automation of his or her profession, the undermining of their voice of authority?

I also question whether regular people have enough common sense and medical vocabulary necessary to safely operate such a system on their own behalf. Seems like it would be a fertile ground for lawsuits.

http://en.wikipedia.org/wiki/Mycin http://en.wikipedia.org/wiki/Expert_system

Thanks everyone for the responses.

> what doctor wants to facilitate the commoditization and automation of his or her profession

Agree in principle but it just takes one incumbent to go rogue and become rich selling his knowledge, and the entire market is disrupted. It happened in the securities industry - the knowledge of old-school pit traders has been almost completely transferred to algorithms. We're seeing the start of it in drug stores now too. angersock mentions the same concerns below with doctors' egos, but the thing to remember is that initially these systems are brought in to be operated by a trader/pharmacist; complete replacement is several steps down the line.

1. There are already systems that before your appointment , gather detailed data and history and display it to the doctor is a condensed and efficient manner. For more details, see[1]. In general : " history taking in this manner is more efficient than the traditional method and allows more time for discussing the assessment and plan. Best of all, it is also associated with better clinical outcomes."

But they aren't commonly used. It's probably because of medical conservatism coupled with business and status reasons.

And in some cases it does help to have the medical interview with a person. But wouldn't a cheaper nurse armed with a great software tool, offering a 20/30 minutes appointment is better than 7/10 minutes of a doctor's time? In some places in the british system that's what happens and research says the results are comparable to a doctor , with patients being happier.

2. There are crowd sourced solutions like you specify. The one i know is a ycombinator company called crowdmed.com . At least according to the media they do bring something very good to the table, but they're very new so it'll probably some mote time to see how they function and scale.

As an aside, there's an an expert system by a company called isabel healthcare, that in the hands of a regular family physician ,can enable regular physicians to diagnose rare and complex diseases at the level of top diagnosticians.It's available via cloud.I'm sure almost no physician of anybody in this thread will have it.

3. Not sure who is to blame, but i think many of those problems are common globally to some extent. Maybe the british system is somewhat better with the policy of employing lower level providers across many roles.

[1]http://www.aafp.org/fpm/2007/0700/p39.html

As an outsider but with intimate knowledge of both systems, I agree with you that British system is better than US system for a large percentage of population.

"2) Are there any crowd-sourced solutions available yet for these cases where doctors clearly have no idea? Someone enters all their symptoms and people discuss online and try to come up with some possible options."

A while back I was looking for info on some medical "issue" that I had trouble defining the correct search terms for. Anywho, I eventually came to some sort of medical site where users could ask questions. And then doctors would comment/answer/respond on it. Now, I presume that these doctors were paid and/or verified, but I never looked too far into it. They sure sounded like they knew what they were talking about, and it wasn't as if they copy-pasted something from the net / a book, the responses were tailored to the questions.

Partially addressing 1 and 2: As someone who has had Lyme and writes software I can't imagine how we would even get 10% of the way there on #1, and #2 just sounds like a recipe for disaster. Imagine if Reddit took on patient diagnosis.

For lyme specifically, the blood tests are generally inconclusive, symptoms vary from person to person and are inconsistent, and even doctors who specialize in lyme seem to disagree on core details of the diseases' effects.

If consensus can't be reached even at that level, and if that level of complexity is present in many other ailments (which I assume it is), how would be able to build a beneficial piece of software around that information?

Ok but I guess the question is, is 10% better or worse than the diagnosis rate of a group of specialists?

Also, I imagine there are massive benefits if people all around the world were pumping data into a system. Correlations that a group of specialists may not think to look for may start to appear.

The problem with medicine is a problem of specificity and sensitivity. The most important statistical Technique in Medicine is RR (relative risk). In sum, we are not smart enough nor do we have the data to increase the efficiency of medicine.

1- If this happens to be like all those websites with symptom matching and finding possible causes, i'll pass. Thanks.

I have found out that I, along with my wife and my cat most likely had cancer whenever I looked some symptoms up.

Reply on news.ycombinator.com