Don't tell Europeans/Scandinavians about that. In Finland there is now massive Epic rollout in the capital region (costwise it looks like $1bn or more) and most likely also huge installation coming up In Central Finland (Its either Epic or Cerner).
Samething in Norway, central Norweigian health district chose epic.
Oh and almost forgot capital area in Denmark. Huge epic installation.
It's garbage because the interface sucks. It looks like it was designed in 1998, and is full of clunky non sequiturs like, "You cannot proceed. Proceed anyway?" There are untold hours and dollars spent on training users because of it.
It's garbage on the backend because the update process is ridiculously manual. Every minor change sends out a ticket with incredibly detailed instructions, e.g. to update a single element somewhere. The role of the analyst is essentially to execute a script, by hand. Not only is this more costly and slower, it's more error prone. But it creates an ecosystem of these Epic-trained and -approved technicians which helps lock in customers and ensures another steady stream of income to epic.
It's a billing system with clinical related functionality tacked on. It's pretty frustrating to use from the perspective of a "boots on the ground" user.
That's a pretty pessimistic and quite frankly not accurate assessment of the software. They have a ton of clinical functionality, and that's what the majority of the dev/design spends their time on at their users group meeting, among other places. You might not like it, but many do.
You haven't spoken to many doctors then. I don't mean to be a jerk or snarky. It really is an awful piece of software. Among its worst flaws is that in most implementations, neither Cerner not Epic encourage structured data recording except for billing codes. This means that if a patient comes through the medical system frequently, doctors have to read pages and pages of unstructured text to get a sense of what's going on for the patient. The shittiness of the software (and I'm normally sorry to curse) is unquestionably leading to worse outcomes for patients.
I'm not really sure what to say to that, since you seem to be implying that no one likes it, which is pretty patently false.
A lot of it does depend on the organization you're working with, as some are more dysfunctional than others when it comes to setting up best practices and build for their physicians. Others actually listen to their clinical users and tailor the system for them.
I'm not saying no one likes it, I'm saying none of the several dozen providers and nurses I know (between two different hospitals in the area using Epic) like it.
I also don't disagree with the fact that some of the issues arise from the implementation requirements. But at some point, it's still the system's fault if it allows its users to be burdened like that. It shouldn't take clicking across three different pages and who knows how many modals to triage one patient in an emergency room. It's silly that I've had to learn which order to provide my transfer of care report so the nurse doesn't have to keep clicking back and forth between different pages...
I fundamentally disagree with the idea that the system can somehow overcome poor implementation; guardrails can only do so much. At the end of the day you can set up almost any piece of software in a way that hinders rather than helps a user. At some point the organization needs to take some responsibility for that.
At any rate, this all started with a glib "garbage software" comment, so I suppose I should happy that you acknowledge that the implementation requirements set by the organization have at least something to do with overall user satisfaction.
I used to work in IT where I worked with sizeable Epic installations... and now I’m a clinician amongst other things.
I’m going to agree with the garbage software sentiment.
I will say that Epic implementations tend to be liked better than alternatives... but when it replaced some piece of shit Meditech implementation nursed from 1980 that’s not really high praise dude.
> I fundamentally disagree with the idea that the system can somehow overcome poor implementation
If very large amounts of 'implementation' have to be done on top of the software, then that's also a sign of bad design. It should be handling more of the implementation and making it more streamlined.
Having used a dozen EMR systems and Jira, I'd say Jira is much better at its job than most EMRs are at theirs. In fact, now that you mention it: you could easily model each patient as an epic, assignments to various people on various teams (ICU, pharmacy, lab, etc), to-do/in-progress/done. Holy cricky, you may have just cracked the EMR nut.
They already have an option like that. The problem is a lot of organizations still want to do their own thing or have their own requirements they want to impose on their users, which is where the large amounts of "implementation" comes from.
I'll be the first, then. I've used two versions of Cerner (the first was heavily modified by the University where I did residency and the second is a more recent version), Soarian, Epic, CPRS, and had brushings with half a dozen outpatient EMRs. Epic is bad, but it's unfortunately still miles better than anything else I have ever used so far.
I think you’re right. I would like/love it if my hospital dropped Cerner for Epic, as I could get my job done at least 20% faster, which would allow more time with family.
My main point I wanted to make was 1) that I also hated Epic the first time I used it in residency (we switched from Cerner to Epic for outpatient only). 2) I have so far never met someone who prefers another EMR over Epic, which really says something since Epic is also bad.
In 2005, when I started working there, the Epic codebase was MUMPS and Visual Basic 6, which left mainstream support before I was even allowed to touch it. There was no migration plan for expiration of extended support in 2008.
As for the MUMPS code, there was a node size limitation for the code. Which included variable names and comments. So each code segment (identified by inscrutable five-letter (or less) names like "^ZHMRG") was jam-packed as tightly as possible with one-letter variable names and zero comments. Basically unmaintainable, but I bet they're still using it.
Garbage software. I left in 2007.
They like to hire straight out of universities. It was the second job for me, so I didn't really know to jump ship earlier.
Total garbage software that essentially functions as a billing tool.