Hacker News Reader

Comment by jbob2000 | Hacker News Reader

jbob2000

I had a conversation with a coworker who used to build software for the criminal justice system. To make a long story short, it turned out to be impossible to make a good software system for the criminal justice system. Between the courts, police departments, and government, nobody was willing to give up any data, fearing they would lose their own value in the system.

I fear that healthcare will suffer the same fate. Between insurance companies, doctors, hospitals, patients, governments, and every middleman inbetween, nobody is going to want a system that makes one of them irrelevant.

View on HN

> Between insurance companies, doctors, hospitals, patients, governments, and every middleman inbetween

You can eliminate middlemen if you can show cost reductions for a player larger than the middleman. The big players are providers, payors and technology (pharma/device) companies. Plenty of middlemen in between those that can be made irrelevant.

You would think that's the case but from experience it's not. I can name two hospitals in NYC that were supposed to see savings from EHR implementations but ended up having multiple competing EHRs that don't speak to each other because of conflicts of interest. You forget there in many cases there are many payors, tech vendors and providers within a single institution.

That's been my experience as well. EHR providers have no incentive to exchange data with each other and will do everything in their power not to. There are regional efforts towards interoperability but there needs to be more political will to mandate cooperation.

What about HL7? When I last worked with an EHR system (2? years ago), implementing HL7-based interop with other providers was all the rage so that they could eliminate fax machines when transferring patient records.

HL7 is a massive farce. It is an impossibly large standard that's difficult to understand and rarely used for interoperability.

Eh, Epic and Cerner interface fine, and offer HL7 interfaces, and Epic in particular bends over backwards to do what their customers want for implementations. My impression was that hospitals are (understandably) reluctant to allow their direct competitors direct access to information they've collected.

I wouldn't go that far with Epic. They always have some unadvertised module they try to push on hospitals the minute they want to interface with a third party.

You can eliminate middlemen if you can show cost reductions for a player larger than the middleman.

You'd be surprised. Weak decision-makers -- and they are legion -- fear what they don't understand. They'll justify paying a more for a known entity than even entertain a small risk of chaos or disruption.

In a rational world, you'd be right. We don't live there, unfortunately.

Firefly's "Malcolm Reynolds", S1E10 "War Stories":

    It wasn't a bad idea, Wash, but eliminating the middleman is never
    as simple as it sounds. ... 'Bout 50% of the human race is middlemen,
    and they don't take kindly to being eliminated.

While it is obviously possible to bypass middlemen, people tend to react strongly against anything that appears to be threatening their livelihood.

You forget that there is more than just the american market. Im sure that the NHS in the UK would love stuff like this.

I think the comment is pretty astute -- this is a big challenge.

Even if Canada, where the state is effectively the sole insurer and determines treatments and pricing for everyone (arguably more centralized than in the UK), the actors in the healthcare system are still wary of connecting to each other.

I learned about this recently from someone I know who is working in "health informatics" at a hospital. Their systems don't share any data with other hospitals, the ministry (the insurer and regulator), etc. There's an initiative to have hospitals in just a single major urban area connected, and despite it being a few years in only a few hospitals are actually participating.

I know Google is making large bets but I am not sure they can make so much of progress that doctors and insurance companies need to worry about their survival. If we ever reach that point, everyone in the population will face the heat.

Doctors and insurance frameworks that act as obstacles domestically may push Google to pursue healthcare developments in other countries with less onerous restrictions. In other words, I'm sure Google will find a path of least resistance to realize the benefits of their investments even if it means going outside of the US market.

Google should partner with Watsi [1], who is already working on funding third world healthcare.

[1] https://watsi.org/

Some folks on the inside think Google has an ambition to replace the insurance companies.

http://medcitynews.com/2016/01/google-biogen-payer-cms/

Big Pharma will pay huge money for this data. Imagine targeted ads based on your medical history. This was exactly the scenario discussed, solving the "who'll pay?" problem, when our corporate overlords were negotiating data sharing with Google Health and Microsoft HealthVault.

Fortunately, at the time, this exploitation of patient data wasn't permitted, so those efforts fizzled out.

I've been out of the game for a few years, so I don't know the current legal and regulatory landscape.

>>nobody was willing to give up any data, fearing they would lose their own value in the system

I suspect this is correct, particularly between the insurance companies and the providers. The coding of what exactly was done directly drives the reimbursement that the providers get. Thus, there's a clear incentive to be opaque and show the insurers only what you need to. Sometimes for good reason (insurers denying coverage solely for cost reasons), sometimes not (billing fraud).

Yup. Only way for the US to have portable medical records is to switch to single payer.

Source: I designed, implemented, supported five healthcare information exchanges.

Healthcare has one advantage and that is that the data ultimately is owned by the patients. So this might actually allow them to get access to the data.

Maybe it's theoretically owned by the patient , but is there any place in the world, that if a patient wants the data , in a common electronic format , he gets it ?

Heck, even the stuff you do get is written intentionally in bad handwriting by doctors(which they do so to protect themselves, and probably at a risk of harming patients).

Back when healthcare information exchanges were a new thing, one of the competing notions was for patients to "carry" around their own medical records, explicitly granting access on a need-to-know basis. Thumb drive, cloud storage, whatever.

I know that your statement isn't suggesting this strategy. I only mention it to note the logical conclusion of a patient-centric system.

Currently, getting access to your own medical records is like getting your own financial records and credit reports, only worse.

Even though I am my data, it is my identity, my self, none of the aggregators and purveyors, who profit from my data, see it that way, and only comply reluctantly.

They can always play the political game , like UBER - maybe building a solution and advertising it straight to patients and/or politicians(and legally bribe them if needs be). Once patients groups and politicians will be in the mix , it will be much harder for the healthcare system to resist change.

That's a good point and one that's starting to change a bit with the rise of integrated health systems (payor and provider integration). Intermountain Healthcare is one such organization.

It's an interesting status, kinda like game theory meeting system theory. Communication but only to a certain degree otherwise the organs die.

Is google big enough to set up an independent "healthcare stack" to just do their own thing?

Reply on news.ycombinator.com