Originally published at: https://boingboing.net/2019/03/18/failure-by-design.html
…
Oh my lord, you enter into a deep dark valley with this topic. Our “cloud-based EMR” was hit with a ransomware attack a couple years ago, which locked us out of all our records. We were trying to practice without access to medical history, allergy history, etc. We do keep paper records for vaccines, at least, so we had that, but it was borderline (or over that line) malpractice the whole time, and if something had occurred we would have been hung out to dry. EMR was never designed for providers or patients, it was designed to make insurance denial of valid claims easier. I hate it with a burning passion.
Exactly. It’s all about analytics. You can track ICD-10 codes, drug usage, etc… then you can make wild claims about X and Y and find magical ways to save the company millions. It’s all bullshit from the MBA folks and it makes practicing a nightmare.
I work with lots of small practices, and one thing I’ve learned over the past decade or so is:
All EHRs suck; they just suck in different ways.
“… certified vendors were allowed to bind the healthcare providers to nondisclosure and nondisparagement clauses.”
Personally, I’d view that as a great warning bell when looking for a vendor; though, I expect that it is buried at the bottom of a mire of boilerplate.
Paper has its numerous problems, but it’s a well-known, almost idiot-proof and remarkably stable medium.
I have yet to have a paper chart crash on me.
There are no good EHRs, just varying degrees of bad. And the lack of record compatibility between systems is downright criminal. To transfer records, the most straightforward way is to print the relevant info, fax it, scan the fax into the new system, and shred the fax. Direct electronic transfers almost never work.
Well known and stable yes, idiot proof definitely not. People fucked up paper records all the time. Many people have died over the years due to errors on the records, misplaced records, or even pulling the wrong records for the patient. One of the big problems with paper records is accessibility. You have to find where they are stored (it might be multiple places) and then retrieve them (which can take hours or days).
Unfortunately, the retrieveability problem has been ported over to the electronic records due to geedy vendors trying to stovepipe ever facility. This is a huge failure on the government’s part. Interoperability should have been specified right from the start, but all of the vendors hide behind HIPPA and claim that interoperability is some impossible dream.
As someone who’s helped debug and write some EHR related apps, I’ll confirm it’s a mess. For me, I say just scan the old paper documents and leave it at that. Frankly, I’m glad I’m out of that business. I have no plans of ever returning to it.
Transition to computer records is doable. Banks did this decades ago, and I’m sure there were some hiccups, but you don’t see banks with stacks of paper ledgers anymore. And while the banking sector has had all kinds of missteps and has failed consumers in many ways, having electronic records didn’t cause that. The healthcare industry has been much slower to move over, and it sounds like they’ve done it in the worse way possible.
Socialized Universal Healthcare would fix this by its nature yeah?
Anyone from our more enlightened necks ever run into these problems?
Not sure these are directly comparable. Banks are both the consumers and beneficiaries of electronic records. In medicine, the insurance companies are the beneficiaries, the guy (us) who actually has to use it has virtually no input into the design or functionality of the system. As was mentioned upthread, interoperability should have been a top-line priority in initial design, but no, not an option at all. We are about to change our system from AllScripts to Epic, and we are dreading the changeover. Not least because there is no way to port the records to the new system and we will have to work out of 2 different systems for a good long time. Ineffeciency at its worst.
I’ve talked about drkoop.com before, but it’s relevant here as well. In 1999 this was their real business plan–the portable personal medical record. Two of us were busily writing the HTML front-end pages every day, and an army of developers were building the back end. Then one day a VP came in and told us to stop working on all of it and delete all the files we already created. And I never heard another word about the Personal Medical Record again in the two more years that the company limped along.
Rightfully so. Epic sucks.
I use EHR in my Emergency Department, and parts of it are truly great: no more hunting for charts, no more waiting an hour and hoping medical records can find unresponsive Mr Smith’s chart, or wondering if the resident took the clipboard, or why the order sheet dissapeared from the chart, and most importantly
no more reading coworker’s impossible handwriting.
But the problems pointed out in the article and in the comments are real, everyday occurances that exist because people haven’t bothered to fix them. And because the people with the power to fix them don’t seem to care.
This morning we had mandatory training on how to slave our IV pumps into the charting system. The computer training program locked up 3/4 of the way through it. The scanners don’t reliably scan the barcodes. It now takes 11 steps to start an IV pump, instead of 2.
EHR doesn’t stop errors - the orders can still be written for the wrong patient, orders for incompatible medications can still be written, orders for the wrong dose or the wrong rate can still be written… Then good luck inputting the correct data when the bedside practitioner realizes the computer is incorrect!
EHRs do a lot of good things. But they don’t stop errors - they just change the type of errors.
So since everything about this is obviously awful, have there been any attempts at producing an open source solution?
That scares the shit out of me.
As a patient, I hate it with a passion. Either there’s a huge laptop between me and the physician, or there’s another person in the room transcribing everything, and not getting everything correct.
Or you use my approach, which is to leave the computer outside the room and converse normally, then try to enter from memory after the visit. None of it is ideal, but I think the personal interaction is vital. I practice general pediatrics and so much of what I do is counseling nervous parents, and you just can’t do that from behind a computer.
I never had a paper chart crash, but have had them illegible or unfindable. The perfect solution has not been introduced yet, but EMR as it is currently known is certainly not it.
I wouldn’t dread it that much. Every system has its pros and cons - but by in large - Epic is one of the better ones out there.
I’ve been deploying Epic on behalf of the government for many years - and they are both good to work with and getting alot more interoperable.