Designing a New Electronic Medical Record
techcrunch.comSince this is going on github, why not expand it a bit to allow people to submit things like stylesheets for different types of media/screen sizes or even JavaScript code for formatting the CCD XML format or maybe doing different internationalization formats? Anyway the next step after the design is for everyone to try to code the same design, so might as well at least be open to that part of it.
I don't really have time but if I wanted my submission to be considered I would try to cheat that way by creating working JavaScript code that processed XML and spit out different formatted versions for phones vs. larger screens and also different languages.
We're having a quick discussion tomorrow - we currently have .HTML. We are going to add .CSS and .JS.
Anything else we are missing?
Please excuse the off-topic rant: It's my mission right now to change something a little different about EMR design. You're looking at the output. That's wonderful and I am really happy about it. We also need to fix input. EMR/EHR interfaces are mostly terrible; that has significant negative consequences for both patients and providers.
You're looking for an output design that is effective as a physical document. Please consider a follow-up challenge for EMR input UI/UX.
Not sure how this thing works but maybe allow people to submit a URL to a working demo?
Do you know where there are any samples a CCD XML file or better yet, a DTD for their format?
Not at all surprised the VA is involved in something like this -- they are one of the best large EMR/EHR deployments in government and a great system by commercial standards too. They seem to be a decade or two ahead of the active duty military medical system on this. (VistA vs. AHLTA).
There's a new effort to merge both systems onto a new one called OSEHRA (Open Source ...), which should be interesting, but I hope it looks more like VistA and less like AHLTA. I suspect MEDCOM is more powerful politically than the VA, though.
Please forgive the semi-shameless plug, but if you're into this we're a team that thinks about this full time, and we're hiring! We're a small startup rethinking the way patients and their caregivers interact with data. Check us out at http://elationemr.com
i'm actually interested in this for different reasons.
ps. got an api?
[edit] forgot to read the profile
Plans for one! My contact info is in my profile. I'd love to get your input on what you'd like to see in an API.
Cool, having some professional experience in this domain, I'm interested in what you folks are doing.
looked up a bit for existing records, and required fields and there is no placeholder for genetics data.
Maybe I'm wrong, or I'm overseeing something but I think that "Patient has inherited malicious BRCA1 mutation" information should be IN patients medical record.
we're using the standard set of fields that all EMRs will be required to produce. you can def submit a design that includes those AND how you would represent genetics data.
You are using the very same set of fields that all EMR will be required to use by the government?
So if I understand correctly, you are basically doing what you have been told to do? Good for you - call that corporate software development, but not design, creation, startup of whatever.
Regarding your reply, it's not about the "right" way to represent genetics data, but having any way at all to input that - which seems "absent" in the current model, but maybe I haven't dug deep enough.
Also, $50k for the design that will be used for 6 millions potential clients? Hmm it's like 0.0083 cent per client. If you make a piece of software that successful, it's like selling it for 1 cent apiece on the applestore.
So I guess I'll pass on this one.
BTW there is something funny in the article: "You might even end up working with big EMR startups like Practice Fusion."
Well, yey :-/
I'd rather like to create a new EMR company - the one that'll put the others out of business :-)
Good luck to all entrants anyway.
EMRs and all related systems (PACS/DICOM in radiology which is what I know best, Lab Info Systems, etc.) all interoperate by open (but usually fairly stupid) standards like HL7. They are basically the worst of OSI, ASN.1, etc. all rolled into one. But, they're standards, and there is a huge amount of workflow deployed around them -- $5-10mm imaging devices that speak these protocols, entire labs, huge numbers of doctors trained around them, etc.
There's no way a records presentation system would be useful if it didn't take standards like HL7, ICD9 codes, etc. on the input. There has been a huge effort over the past few years to get small practices onto EHR/EMR (for a variety of reasons...patient care quality and billing/admin cost savings), so anything which didn't work with those standards would be a huge step backward on those efforts as well.
The nice thing is most of these ugly protocols have decent libraries available (some open source, some commercial), so you can build a HIS/EHR/EMR/PACS/LIS/etc. with general software engineering talent and a limited amount of specialty knowledge, but you really do need some experience in the clinical environment (at least as a tech or doing tech support), I think, to build a credible overall system. And to actuall sell larger systems, an MD on the team is really helpful (at least for PACS, a rad or ortho).
That said, being able to 'pretty print' a medical record (on paper, or maybe some kind of interactive format) for patient use or portability outside the system doesn't require a huge amount of clinical knowledge, but it's not enough to replace an EHR.
They're dedicating a whole $25k for this project. As a prize? Surely the government should be engaging a good UX design shop and spending a reasonable sum of money for a complete design and implementation cycle including user testing.
This is right up my freakin' alley. Awesome.
Oh man, you're the guy that did the cool site face lifts :).
when you're saying design, you don't mean the actually restructuring, but just the way it's represented correct? so whatever is in the txt stays in the txt
great. Just signed up.