EMR Features we actually need.

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

RustedFox

The mouse police never sleeps.
Lifetime Donor
15+ Year Member
Joined
Aug 21, 2007
Messages
8,292
Reaction score
14,608
I sketched this out after lunch. Some ideas about EMR features that would actually be useful to us, instead of a screen cluttered with nonsense that doesn't matter to us:


20180809_145010.jpg


Take it away, SDN. All suggestions will receive either consideration or criticism.

Members don't see this ad.
 
I often wondered why most EMR solutions are either insanely expensive or ridiculously cluttered. It shouldn't be too hard to build your model above from open source software. Maybe the people who build EMR need to take more feedback from the people that actually use it?

A good article I came across Quora the other day.
https://www.quora.com/profile/Jae-Won-Joh/Posts/How-to-build-a-good-EMR-part-1
 
Because EMR was never designed for physicians. It’s more of an administrative tool, for nurse to document and for administration to show they’re following all the regulations. I’ve always loved when they (instructors) give me more than three ways to get to one piece of important information. On top of that, they all need to, at some point, be compatible with each other. So there are certain standards all of them have to adhere to.

You can at best optimize the screen for your daily use. Have your favorite plan/notes/template be handy. You have the big companies which are now use by big health care systems, bigger the company, more grease they pour in to lub the machine.
 
Members don't see this ad :)
I often wondered why most EMR solutions are either insanely expensive or ridiculously cluttered. It shouldn't be too hard to build your model above from open source software. Maybe the people who build EMR need to take more feedback from the people that actually use it?

A good article I came across Quora the other day.
https://www.quora.com/profile/Jae-Won-Joh/Posts/How-to-build-a-good-EMR-part-1

There are already several open source EMRs out there that have some (but not all) of the capabilities of proprietary EMRs (VistA, OpenMRS, OpenEMR, ...); however, none of them are as user friendly as need be and are almost always standalone solutions.

Most hospitals and hospital systems want one system to handle everything for them (EPIC, Allscripts) and they pay dearly for that "feature".

In order to build a robust, user friendly, open source EMR you will need $1-10 million of funding to make it happen as a start-up company that is running lean. You will also need to sell this system to the hospitals which have EPIC entrenched and are currently paying them $$$ for annual maintenance. The pain of switching from one system to another will be off the charts (it's 20 out of 10 pain!).
 
Genius.

I would like an actual list of allergies rather than a laundry list of bull*** that makes me have to go to uptodate and search out the 4th line antibiotic for condition X.

Oh, and if someone is going to document an "allergy" it should hard force you to specify what the "allergy" is. And no, "anxiety" after getting morphine is not an allergy.
 
I would like an actual list of allergies rather than a laundry list of bull*** that makes me have to go to uptodate and search out the 4th line antibiotic for condition X.

Oh, and if someone is going to document an "allergy" it should hard force you to specify what the "allergy" is. And no, "anxiety" after getting morphine is not an allergy.

I actually take the :30 seconds to dispel allergies with the patient, and tell them to stop listing "codeine, iodine, morphine" as an allergy if its bogus.
 
All the hate for Meditech makes me chuckle, a little bit, because, the first time I used it, someone asked me if I had used it before, because I was so facile with it. If anyone was ever natural with anything, it was me with that.

Otherwise, the job was a piece of ****.
 
There are already several open source EMRs out there that have some (but not all) of the capabilities of proprietary EMRs (VistA, OpenMRS, OpenEMR, ...); however, none of them are as user friendly as need be and are almost always standalone solutions.

Most hospitals and hospital systems want one system to handle everything for them (EPIC, Allscripts) and they pay dearly for that "feature".

In order to build a robust, user friendly, open source EMR you will need $1-10 million of funding to make it happen as a start-up company that is running lean. You will also need to sell this system to the hospitals which have EPIC entrenched and are currently paying them $$$ for annual maintenance. The pain of switching from one system to another will be off the charts (it's 20 out of 10 pain!).

With open source, I was referring to the use of open source content management systems (like Drupal) for building custom hospital-specific EMRs. That way they can build it the way they want, have the features exactly how they want it, and best of all shouldn't cost millions of dollars to develop. Just a thought. Scalability and security are usually the issues with such an arrangement, but then CMSs such as "Drupal" in particular is really good at those aspects.

When you say "one system to handle everything" do you mean flexible enough to act as an interface at the level of registration, patient encounter, inpatient, labs, and admin? Or something that can be shared across networks and have inter-system compatibility? The latter is the hard part.
 
Your system makes too much sense so no one will be able to make money off proprietary bs. Therefore it will never work.
 
EMRs are built for 1 thing. Charge capture. Any other use of it is secondary or tertiary.
 
Open source is all well and good, but it also needs to be able to integrate with the hospital and local labs, radiology services, etc, etc. My understanding is that a good portion of the 8-9 figure price tags for EMR implementation for hospital systems is getting that interoperability at least somewhat functioning.
 
Members don't see this ad :)
But in all seriousness though, EMR could be vastly improved. Its primarily designed to let the billers collect data to increase billing and find what new metrics can be set. I would agree with your system. Also, way better than I could have drawn it!
 
With open source, I was referring to the use of open source content management systems (like Drupal) for building custom hospital-specific EMRs. That way they can build it the way they want, have the features exactly how they want it, and best of all shouldn't cost millions of dollars to develop. Just a thought. Scalability and security are usually the issues with such an arrangement, but then CMSs such as "Drupal" in particular is really good at those aspects.

Maybe the concept of a CMS has completely changed in the last decade, but I fail to see how it is in any way related to an EMR system.
 
Maybe the concept of a CMS has completely changed in the last decade, but I fail to see how it is in any way related to an EMR system.
It has. Drupal is no longer just a content management system, it has a strong backend that can support any form of webapps, with powerful database management features.
 
With open source, I was referring to the use of open source content management systems (like Drupal) for building custom hospital-specific EMRs. That way they can build it the way they want, have the features exactly how they want it, and best of all shouldn't cost millions of dollars to develop. Just a thought. Scalability and security are usually the issues with such an arrangement, but then CMSs such as "Drupal" in particular is really good at those aspects.

When you say "one system to handle everything" do you mean flexible enough to act as an interface at the level of registration, patient encounter, inpatient, labs, and admin? Or something that can be shared across networks and have inter-system compatibility? The latter is the hard part.

Drupal and the others like it (Wordpress being the other) are made up of multiple components that may be prone to security issues but that's also true for any software that is written. It may be difficult convincing someone to go along with it for this reason among many others.

EMRs need to be web applications which would solve the whole shared across network and inter-system compatibilities. The "one system" to handle everything is what hospitals want and exactly as you describe it.
 
Your system makes too much sense so no one will be able to make money off proprietary bs. Therefore it will never work.

You can make money from open source software and systems. Just not as much as the proprietary companies.
 
Open source is all well and good, but it also needs to be able to integrate with the hospital and local labs, radiology services, etc, etc. My understanding is that a good portion of the 8-9 figure price tags for EMR implementation for hospital systems is getting that interoperability at least somewhat functioning.

EPIC, Allscripts and their ilk charge the 8-9 figure price tag because they have the hospitals over a barrel. Once you go with them then you are dependent on them to provide annual maintenance and ongoing updates. Inter-operability between radiology, labs, registration, etc. is based on HL7 / FHIR which is a standard. Switching to a competing proprietary system is just as expensive because they dictate the price as to "what the market will bear".
 
Naive med student here, but why can’t an EMR do both - be good at charge capture & user friendly?

Seems like it wouldn’t be that hard to have all the features of cerner/epic on the backend then put a customizable Doctor/nurse/etc specific home screen on the front end.
 
I sketched this out after lunch. Some ideas about EMR features that would actually be useful to us, instead of a screen cluttered with nonsense that doesn't matter to us:


View attachment 238571

Take it away, SDN. All suggestions will receive either consideration or criticism.

If you're serious about making this happen, the most similar EHR I've seen to the picture you drew is Larry Nathanson et al's home-brewed EHR created at BIDMC in Boston. The name escapes me. As of 2014 when I did my audition there as a med stud, was in use in BIDMC ER and a couple other outlying hospitals. Beautiful, beautiful system.

I was a codemonkey/startup guy in my first career and have dreamed of seeing a system like this in common use in ERs for a long time, ideally open source. As others have noted, the bureaucratic and capitalistic hurdles to making that happen are significant. Maybe I'll be in a place to put some sweat equity into it in a few years...
 
What you seem to be asking for is more accurate info in the EHR. It won’t matter which system you use if people put poor information in. No program is useful if the data is terrible. Bad data in = bad data out.

You can design a new system, but if you don’t spend the money to interface it with a treasure trove of good data, it will be useless. The problem with most of these systems is that the EHR companies have most of the negotiating power and once purchased, the hospitals don’t want to spend additional money to make them work properly.
 
EPIC, Allscripts and their ilk charge the 8-9 figure price tag because they have the hospitals over a barrel. Once you go with them then you are dependent on them to provide annual maintenance and ongoing updates. Inter-operability between radiology, labs, registration, etc. is based on HL7 / FHIR which is a standard. Switching to a competing proprietary system is just as expensive because they dictate the price as to "what the market will bear".
As I understand it, the standard is poorly implemented and anything but standard from system to system. From what I can see at various hospitals I've worked at and the idiosyncracies of how labs ordered may or may not show up in the system, I'd believe it.

But it sounds like you have more experience than I do.
 
What you seem to be asking for is more accurate info in the EHR. It won’t matter which system you use if people put poor information in. No program is useful if the data is terrible. Bad data in = bad data out.

You can design a new system, but if you don’t spend the money to interface it with a treasure trove of good data, it will be useless. The problem with most of these systems is that the EHR companies have most of the negotiating power and once purchased, the hospitals don’t want to spend additional money to make them work properly.

I like my EMRs filled with important data entered by nurses such as whether the patient has been to east africa lately or their preferred pronoun as well as chaplaincy resident oh excuse me I meant fellow progress notes.
 
Naive med student here, but why can’t an EMR do both - be good at charge capture & user friendly?

Many reasons, but to me it boils down to prioritizing coding ease by needlessly dividing up the note. Whereas I would prefer to write or dictate "65M w/ hx CAD p/w 3 hrs of sudden L-sided dull chest pain after mowing the lawn..." it would be easier for coding if I clicked things individually such as "onset: sudden" "provocation: physical activity" "quality: dull", etc.

Clicking seems fast, and it is if you have to input a small amount of data with limited choices, but given the amount and variety of information we have to input it makes little sense. Having to click 30 things on 15 different lines spread over 4-5 pages is immensely inefficient: click, move, click, unclick, click the correct box this time, search, search, click, move, scroll, search, click, scroll back up, move, click.
 
For me, an ideal EMR would have minimum verbosity, allow for viewing relevant data without having to leave the patient's chart, and allow for ordering without navigating to other screens. Superfluous data could also be hidden via folding regions of text or collapsible text areas similar to what can be seen in a text editor for computer coding (e.g. ).

Even epic, which has better charting than other EMR's has way too much verbosity.

E.g.
Vital signs in "prose" format imported through a .phrase
VITAL SIGNS: BP: (!) 168/77 Pulse: 91 Resp: (!) 32 SpO2: 98 % Weight: 122 lb (55.3 kg)
More concise information:
HR 91, BP 168/77, RR 32, SaO2 98 % on room air.

Even better is if the vital sign line could be linked (e.g. hyperlink) to the underlying vital sign data. Fine granular customization for generating note specific to certain areas of patient care would be ideal.

Similarly, displaying thumbnails of key imaging data and/or ECGs with the ability to click on the item and view the full image would be ideal. Currently one enters a text description of the item - EKG: SR XX beats per minute, biphasic T waves in V2, V3, no further abnormalities noted. While this is nice it, would be better to have access to the underlying data without having to click through multiple aspects of the chart.

Not surprisingly, I have many, many more ideas. Unfortunately, all of the EMR companies lockdown their interfaces so strongly one cannot meaningfully customize their interaction with the data.
 
Not surprisingly, I have many, many more ideas. Unfortunately, all of the EMR companies lockdown their interfaces so strongly one cannot meaningfully customize their interaction with the data.

This... this opacity is the main cause of doctors' EHR problems. And it is even more infuriating to those of us who have been programmers and can appreciate the simplicity and stupidity of what is probably actually going on under the EHRs' hoods in most cases.

I often think about working my way up to CMIO primarily just so I can open up the hood and take a look. (And I'd work fewer nights as well I imagine.)
 
Never understood why folk hate on EMRs so much, me I love Citrix and the rest all pretty okay
 
I thought Citrix was the plugin that enables you to remotely access the EMR, not an EMR itself? Also "the rest all pretty okay"????? I want what you are taking.

Cerner then ?

But yeah, I've used Epic, medtronic think it's called, t-sheets (paper and electronic versions, also something super jank called Healthland. All pretty much okay, really, in the grand scheme of things.

Currently drinking some kombucha midshift if that helps ?

(haven't read the health literature onnit any but I do like the taste )
 
meditech is NOT 'all okay', I don't care what anyone says. That thing is an absolute travesty. I mean, no scroll bar? Are we still in the 80s?
 
meditech is NOT 'all okay', I don't care what anyone says. That thing is an absolute travesty. I mean, no scroll bar? Are we still in the 80s?

When the docs using Meditech are finding ways to edit in MS Word and then just copy/paste the note, it should be clear to admin just how awful that thing is.
Meditech is the glue-eating mouthbreather of the EMR world.
 
When the docs using Meditech are finding ways to edit in MS Word and then just copy/paste the note, it should be clear to admin just how awful that thing is.
Meditech is the glue-eating mouthbreather of the EMR world.

Hey don't knock it! The geniuses at HCA signed an exclusive 100 bazillion year contract to use Meditech. These guys are great, and never make bad decisions....

The only thing I like is it does import all my lab and imaging data automatically......CERNER! I'm looking at you..........
 
Cerner then ?

But yeah, I've used Epic, medtronic think it's called, t-sheets (paper and electronic versions, also something super jank called Healthland. All pretty much okay, really, in the grand scheme of things.

Currently drinking some kombucha midshift if that helps ?

(haven't read the health literature onnit any but I do like the taste )
HealthLand is a steaming piece of ****, and we're the largest hospital to use it (with 7k visits/year). Absolute garbage.
 
Hey don't knock it! The geniuses at HCA signed an exclusive 100 bazillion year contract to use Meditech. These guys are great, and never make bad decisions....

The only thing I like is it does import all my lab and imaging data automatically......CERNER! I'm looking at you..........

Yeah, Meditech's MHx/lab/imaging import function is pretty great all things considered. Soarian could never do that.

Man, I hate Soarian...
 
meditech is NOT 'all okay', I don't care what anyone says. That thing is an absolute travesty. I mean, no scroll bar? Are we still in the 80s?

Mine has a scrollbar. The trick is that you need to click on the scrollbar to get the mousewheel to actually scroll.

You need to click the scrollbar every single time, for every single frickin' Meditech window you open, if you want to use that stupid mousewheel. Man, I hate Meditech...
 
I saved the pic and will try to convince allscripts to mirror it for me 🙂
 
Top