I talked to a doctor at Stanford and he mentioned to me one reason why the main Stanford Health and Stanford Children's Health have separate patient portals (not unified) that you have to create separate logins for is because of the huge amount Epic wanted to charge to merge them. (They both run separate Epic instances as far as I can tell)

I used to go to NYU medical, which uses Epic. As janky as their interface is, it was nice enough to use and see all my charts, etc, as well as communicate with doctors over email. So when I moved to another city I specifically chose a medical system that used Epic. Once I got my login I then logged into my NYU Epic to initiate a data transfer and wouldn't you know it... it was impossible. After digging around to no avail I contacted my new provider and they told me the best thing to do is print out a PDF of my medical history in NYU's Epic and they would scan it back into their Epic... with a complete loss in structured historical data (bloodwork, vitals etc). Insane.

How are leeches like this tolerated? The sooner they are driven out of business, the better.

They're not entirely tolerated fortunately. The US government has been taking action to tear down EHR walled gardens via the Cures Act[1] which is manifesting in public access APIs called FHIR.

[1] https://en.wikipedia.org/wiki/21st_Century_Cures_Act

FHIR is nice but doesn’t completely reduce the complexity of integrating the underlying data itself considering the mess that is HL7 and mess that occurred with v3.

The situation is sort of like moving from SOAP+XML Schemas to REST/JSON. Easier to develop, but the underlying data structures are still extremely complicated and tricky to get right.

This is especially true when dealing with ancient systems written in niche languages/database hybrids (MUMPS comes to mind): https://en.m.wikipedia.org/wiki/MUMPS

Do you feel OMOP helps to resolve this?

YES! If only people in healthcare IT would adopt things in years rather than decades...

Just a small correction my original post (better late than never). It's also (perhaps to a larger extent) due to the HITECH Act: https://en.wikipedia.org/wiki/Health_Information_Technology_...

From what I understand this is the norm in the US. For another example, see credit reporting agencies. Control of critical infrastructure is given to private companies under the assumption that competition will make them do good. In practice, most competition goes away immediately and the maximization of profit destroys any remaining incentives to not be horrible for everyone involved.

Same way any systems integrator or ERP vendor like SAP is tolerated? Integrating systems takes time and money.

I suppose because when tech, like Google, comes along Epic gaslight, obstructs, and projects their way in the media to becoming the good guy.

Healthcare IT is full of deeply committed professionals who deeply dislike Epic.

The functionality of the competition isn't much better (google notwithstanding, hypothetically)

That suggests that they believe Amazon will never be a competitor/threat. A lot of dead companies have thought that.

As a former hospital IT employee, as it was told to me by my IT management, you don't get to pick Epic as your EMR; Epic somehow "decides" you're worthy enough to buy it.

Then once you do, you're told who to hire, what to buy, what to train on, what to overall do, and if you don't do that they hold your support costs hostage under the guise of a "quality installation score". Not to mention the no-competes they slap on hospital staff so they can't bail as soon as Epic goes live.