HACKER Q&A
📣 cuuupid

Why do all electronic health records suck so bad?


I go to 6 practices (counting urgent cares, non-primary care, specialists), and they all have their own system with some provider I've never heard of but sounds like any of the others.

This provider always has a hidden subdomain to access the practice's patient portal. There's no way to get to it unless you have memorized it or have a link in your email handy and strip out the end of the URL.

You could go to the homepage of the provider, and they will have a login button, but only for doctors. If you click the "Patient Login" button it either prompts you to ask your doctor, or some of them will present a login that doesn't work and I've absolutely no clue who its for.

Usually it asks for a password, when I have never made a password, and to reset the password it needs my email, but I have no idea which one.

Then, two factors dictate my fate: if the UI is pretty OR if it is a big provider like AthenaHealth (naming & shaming), it is unlikely that anything will work. Usually I have to login 50 times, create an account a bunch, call the practice 2-3 times, and then finally magically the correct sequence of the above will unlock my patient portal. Each time it fails it will, of course, tell me it can't find my account, only to also email me "did you recently try to log in?"

I then try to navigate the maze of their interface which ultimately only needs to provide me with records & let me schedule appointments but, for some reason, has a dozen widgets that appear empty and I can't configure, and prompts me constantly to add profile information. Of course this is necessary and critical, as without a profile picture, how will I use the future social networking + tiktok-copy feature?

After sacrificing my firstborn in hopes of preventing the application from crashing, I might be able to find the records (so far I am 2/4 in getting to the records, taking into account I could only login to 4/6 EHRs), but they are rendered terribly as if this was an afterthought. Some offenders like AthenaHealth actually render your single blood panel into 10 different tests that navigate to different pages (and pressing back in your browser brings you all the way back to Home).

So above, on a commercial level, I was able to use an electronic health record to access my health records (the SOLE and ONLY functionality it absolutely needs to have) a shocking 33% of the time across multiple providers.

On a larger enterprise level or government level, this is even worse. EHRs are regularly costing enterprises millions and the government $10-20MM a pop [1] and yet fail to deliver results with any consistency or stability.

What gives? Why do EHRs suck? This seems easy enough to hack together in a day and still be HIPAA compliant, and it is clearly a competitive market if there are this many providers.

[1] https://sam.gov/opp/1e345d6c414d455897ea4e6292e94b1b/view


  👤 analogj Accepted Answer ✓
So this is partially the reason why I built my own open-source Personal Health Record (PHR) Fasten Health [1][2]

In my experience patient portals vary from incredibly functional to almost worthless, which was a huge problem for me given that some of my important specialists were in the latter bucket.

Honestly, everything that I've read comes to the conclusion that EHR's aren't designed for patients, nor practitioners, they're built for the accounting dept. Patient portals have been tacked on-top to comply with govt regulation & certification programs [3], but UX/usability is almost universally lacking.

Thankfully the FHIR API's that Fasten leverages seem to be fairly consistent:

- the interoperability standard (FHIR) ensures that patient medical records are (somewhat) consistent.

- EHRs APIs are tested against an automated test suite before they are "approved" - its not comprehensive, but its better than the subjective UX rules.

[1] - Marketing - https://www.fastenhealth.com/

[2] - Source Code - https://github.com/fastenhealth/fasten-onprem

[3] - https://chpl.healthit.gov/


👤 jonahbenton
It's the original Microsoft Windows story. Microsoft figured out that the features that mattered were those that their buyers- IT admins- cared about, not users.

Same deal. The EHR doesn't have to be good for patients. You can't pick your doctor based on their EHR. Patients (insurers) pay for health care delivery, not for information delivery. It's just a cost center and the features that matter to buyers have nothing to do with communicating effectively to patients. Here the buyer is the hospital admin who has to have a system that interoperates with their other systems. Both doctors and patients can hate it. Doesn't matter.


👤 gregjor
> This seems easy enough to hack together in a day and still be HIPAA compliant

Even if the problem stems from incompetent developers I think you grossly underestimate the scale by multiple orders of magnitude.

This reads more like a "Tell HN" rather than "Ask HN" because you pose a rhetorical question. Lack of incentive, competition among providers, and no regulatory framework can explain the "why." If you think of health care providers as companies at least partly in the data collection and brokering business it makes more sense why they treat health records as their proprietary property.