Hacker News Reader

Hacker News

Comment by phren0logy | Hacker News Reader

Full thread

phren0logy

I'm in medicine, and I desperately want this model for electronic health records.

The current state of lock-in has been effectively maintained by the entrenched players, and it's really bad.

View on HN

Closest we have so far:

http://hl7.org/fhir/

I really like their documentation. My team has moved a lot of our stuff to produce FHIR (or something close to it), for interoperating with other teams and clients.

Since HLX has become increasingly required, I've seen that the lock-in doesn't mean diddly squat. Now we're not "locked in", but for my new vendor to drop-in means I have to pay an extra "API Fee" for them to whip up the API interface to pull everything from the old EMR and into the new EMR.

So either we get to the point where we are legislating perfect compatibility (and I can't imagine how good EMRs will get once the federal government has to outline every individual data field, and update them through, what, the rulemaking process?); or we'll always be paying up for this transition, and lock-in is beside the point.

Maybe we need a coalition of second-tier EMRs to band together and commit to supporting an API and a group to keep it open and updated? I agree that the current approach isn't really working, and doing more of the same is probably not going to help.

The problem, from my perspective, is that we're fighting the battle on two fronts: I need to get all my data from my old EMR into my new EMR, and I need my new EMR to slot in where my old EMR was with respect to feeding data to my data warehouse. Part of the difficulty there is the API, and part of it is that a bunch of shit is done as a black box in-EMR (e.g., my EMR will feed my warehouse some financial data, but my vendor is opaque as to how it's calculated).

It's gotten to the point where I don't want a legal requirement to be HLX compatible: I want a legal requirement separating back-end data from front-end UI, so that we can shop for each of those independently. Once all the front-end shops (which are more valuable than back-end - as a healthcare org, I care about documentation and billing and error prevention) can't lock you in via data, I imagine there will be a fucking quick race to be the universally compatible back-end. And the back-end is ultimately the stuff that affects patients (portability of records) and loosen the bindings on provider organizations (because... portability of records).

Which of course is why even things like HLX didn't really start working until major orgs like CMS and NYS Medicaid came along and said "you will find a way to be compatible with HLX voluntarily, or you will do it via regulation. One way or another it's going to happen within the next 12 months." (I was at a major conference where that was laid out pretty much that explicitly. It was wonderful.)

I saw that you doubled down on [this comment](https://news.ycombinator.com/item?id=26181228). Week 5 excess mortality for 65+ in Israel is now up to 7.4 times above normal. It increased every update for the last three weeks. Unbelievable that people like you take 1 minute to read something, think you are an expert and try and proclaim victory. I'm 100% sure you will double down again instead of taking the honorable path of admitting you were wrong. People like you never get held accountable.

I made my comment two weeks ago, describing a fall in mortality that began three weeks prior. The z-score of mortality for Israel peaked as I described in my last post, and has continued to fall. The 65+ group in particular is well within historical norms.

The aggressiveness of your statement requires me to include an image of the current mortality graph for anyone not willing to take the time to dig it out themselves: https://ibb.co/1vCsD7Y

People like you never get held accountable. Please continue thread-stalking me, I'm happy to keep this up.

Reply on news.ycombinator.com