Major Health Organization Stops Forcing Doctors To Adopt New Technology (internalmedicinenews.com)

nbauman writes: The administrator of the Centers for Medicare & Medicaid Services, told an investors' conference that they will be backing off the unpopular requirement that doctors show "meaningful use" of their new computer systems. Andy Slavitt, acting administrator, admitted that "physician burden and frustration levels are real. Programs that are designed to improve often distract. Done poorly, measures are divorced from how physicians practice and add to the cynicism that the people who build these programs just don't get it."

Dr. James L. Madara, CEO of the American Medical Association, agreed that EHRs were having a negative impact on physicians' practices. Many physicians are spending at least two hours each workday using their EHR and may click up to 4,000 times per 8-hour shift, he said. Instead, CMS will reward health care providers for patient outcomes through the merit-based incentive pay systems created by last year's Medicare Access and CHIP Reauthorization Act (MACRA) legislation.CMS is calling on the private sector to create apps and analytic tools that will keep data secure while fostering true and widespread interoperability.

Comments

Doctors: Whiny bitches, all of 'em.

[mythosaz]

I spent about a decade doing high-level end-user compute management for a large healthcare organization.

There are two major forces at play.

Doctors just want fancy equipment so they can keep up status.
Doctors are lazy and entitled, and can't be bothered to do anything beneath them.

I've been on countless projects for SSO or (reduced signon, anyway) and context management. I've had to make sure countless pretty-boy doctors could get the new device that the OTHER hospital gave THEIR doctors. It's **all** about physician satisfaction. It's a seller's market, and if you don't give the doctors every last thing they demand they will go to work at the other hospital down the street. Of course, doctors know EVERYTHING, so there's no negotiating with them at any level. Site managers know they're fucked, and we know site managers are fucked, so we bend over and take it.

The context management systems (that keep patients synchronized across multiple clinical apps -- your EMR, or your radiology app, or your bed placement app, or your 10 other non-integrated apps) all suck and are fantastic boondoggles. SSO works for major systems, but unless you're AMAZING and have every last system in Cerner (or whatever you use), your docs will fuck that up too and blame IT.

Whiny bitches, all of 'em.

Re:Doctors: Whiny bitches, all of 'em.

[Thud457]

I'm a Doctor Jim, not some e-paper pushing records clerk!

Re:Doctors: Whiny bitches, all of 'em.

[Rei]

I've also worked in the industry, and I'm of the opposite view: a lot of interface designers have given doctors crappy interfaces that don't take into account real-world use cases.

My particular field was psychiatry, so a lot of the software was tablet-based and focused on asking subjects questions and recording observations of the subjects. The important thing was to realize that these aren't some sort of web-poll - the real world is complex. Maybe the subject will throw a fit and walk out partway through or refuse to answer questions, or only give answers that don't make sense or aren't clear. Perhaps a question's answer choices don't reflect all of the nuances of the situation, something the form designer didn't think of. Perhaps something important or unusual happens in the interview that the doctor needs to note. It's important that software be as flexible as pencil and paper - that they can "pick it up" and "set it down" whenever they want, that they can add answers or scribble notes wherever, etc, and all of this gets recorded, is available to others, and doesn't just "disappear" on them.

Much of modern data-collection interface design is about trying to constrain people - you must do X, Y, and Z, in this order, with some nicely laid out plan of how everything's supposed to be done, etc. But sometimes that's just not practical in the real world. We found that when we made the software have the same "features" as paper, while still collecting data, acceptance was quite good.

Be nice to your users. You can point out possible errors or omissions (so long as you're not being a pest about it), but don't constrain them, don't try to *make* the data be "perfect". Just trust that they'll record the data as best they can. And be ready to handle any imperfect or incomplete data because well, congrats, we live in the real world so sometimes data is just simply going to be imperfect.

Re:Doctors: Whiny bitches, all of 'em.

[ColdWetDog]

Well mine are somewhat different. I'm a physician and have been working with EHRs since the late 1990s. When we had 386 processors and liked it. Yep, there are asshole entitled doctors (and $your_favorite_whipping_person). Lots of physicians would like an EHR that would, you know, help out. Instead we get systems that are designed to 1) help the billing department (an important aspect of medicine, but not the most important) and 2) get little gold stars by following the Meaningless Abuse, er, Meaningful Use "guidelines'.

Medicine is not an easy subject to computerize. For one thing, the old saying 'whey you computerize chaos you end up with computerized chaos" is quite true. Much of medicine is still hunches, witchcraft and showmanship - things most computer systems really don't deal with well. The rest is completely driven to insanity by several decades of Medicare and Medicaid rulemaking on top of often completely contradictory rules by Congress. In it's current state, you can't possibly do everything correctly because you would run afoul of something along the way. I've often thought that if you tried to program an AI to follow Medicare rules it would eventually just unplug itself as the only rational approach.

And no, Obamacare didn't really change much - just added a few more insane rules to the giant pile.

Meaningful Use was one of those things that might have been a good idea if one person set it up and left it in a corner. But it morphed into a giant committee that had inordinate power over EHRs and singlehandedly did more to screw up the advancement of electronic health records than any other single decision by the US government.

There is a god. I will sacrifice a whole box of Rigatoni to His Noodliness in thanks.

Re:Doctors: Whiny bitches, all of 'em.

[Solandri]

There are two major forces at play.

Doctors just want fancy equipment so they can keep up status.
Doctors are lazy and entitled, and can't be bothered to do anything beneath them.

You've just broken the cardinal rule of User Interface design. The user does not exist to use the device; the device exists to be used by the user. If the user is unable or unwilling to quickly adapt to the device's UI, the fault is in the UI, not the user. It doesn't matter if they're lazy, entitled, stuck-up, whatever. If you want your device to be successful, you have to make them want to use it.

I've been helping several doctors set up and transition over the EHR systems. The thing I keep hearing over and over again is, why do they have to do this when paper records were working just fine? In other words, the cost of computerizing their patient records is exceeding the benefit they're seeing. And this isn't doctors and nurses who are trying to learn a new EHR system. Most of them have been using a EHR for 2-3 years now. They know how to use the systems, they systems are just so convoluted that it's impeding their workflow compared to paper records. That's a massive failure of user interface and software design.

Why do you think Apple is so successful despite selling technically inferior products? Because they get this - they make their devices dirt simple to use.

Re:Doctors: Whiny bitches, all of 'em.

[ColdWetDog]

Yep, I spent all summer of 2013, and I mean all summer, trying to beat our POS EHR into some semblance of utility. It was a total failure and it looks like I will be spending the entire summer of 2016 trying to figure out how we disentangle ourselves from this mess. A complete was of a lot of time and money that could have been spent doing something useful.

And "Meaningful Use" was a big part of the reason that our vendor screwed the pooch. It wasn;t a very good system to begin with (Healthland Centriq), had been developed before Meaningful Use was a gleam in the committee's eye and the vendor spent precious (well, cheap Indian) developer hours trying to shoehorn MU requirements into the system instead of getting it to just work. And that little game was repeated all over the country.

Interoperability starts at the server

[xxxJonBoyxxx]

>> CMS is calling on the private sector to create apps and analytic tools that will keep data secure while fostering true and widespread interoperability.

If they were serious about interoperability, the Feds would go after Epic Systems, GE and every other provider of incompatible and ridiculously expensive health care software first. Interoperability ain't a problem to be solved with the next crappy Fitbit clone...

Who knew?

[gstoddart]

Instead, CMS will reward health care providers for patient outcomes

Which is the only meaningful way to do it.

All of this bullshit about forcing people to use bad software is just pointless. I only wish more organizations would do this.

On more than one occasion I've been pushed to "contribute" to SharePoint or otherwise use a piece of software which in no way actually helps me do my actual job. Because someone was more concerned with showing how a useless piece of software was being adopted than understanding why it's not being adopted.

Yawn, you're going to give me a fucking badge for posting to a forum which nobody is reading and which won't solve my problem, because you stupidly believe "teh soshul networking" is going to solve all your problems, when all it's doing is creating new ones.

And I've seen far too many systems intended to replace something already in use, which clearly are written by people who just don't get it. It's an often ignored dirty little secret that absolutely crappy interfaces don't get people to use the software because you go through far too much garbage to do anything.

I've seen stuff which tried to replace custom software, with well written GUIs, for crap which mapped everything to try to look like a spreadsheet ... and which was utterly un-usable. It was like some moron wrote the software with no consideration for what it was being used for.

Useless Metrics

[ranton]

EHRs were having a negative impact on physicians' practices. Many physicians are spending at least two hours each workday using their EHR and may click up to 4,000 times per 8-hour shift, he said. -- Dr. James L. Madara, CEO of the American Medical Association

How does this metric identify a negative impact in any way? If those clicks are keyboard clicks it doesn't even sound high at all. How about something like "doctors among the top 20% of EHR adoption misdiagnosed 10% more often", or something similar? I'm have no idea if pushing adoption of EHRs is beneficial, but based on the metrics Dr. Madara chose to use they don't seem to have any idea either.

Re:Useless Metrics

[freeze128]

Medical management software is not designed for speed or ease of use. For instance, you work in the Emergency Room and need to log in a new patient, and list symptoms. What you're doing is data entry. You are using the keyboard to enter this information. If you need to click on other pages or tabs, you are taking your hands off the keyboard and moving to the mouse. Then you are moving the mouse to the precise location, and clicking a button. Then you are moving your hands back to the keyboard so you can enter more text. Now do that hundreds (if not thousands) of times a day. This is a ***HUGE*** slowdown, all because the software devs don't bother to integrate keyboard shortcuts or alt-key assignments in their software. They just don't get it.

Re:Studies have shown computers distract doctors

[gstoddart]

Exactly, trying to force the tasks required for a specialized task to fit into a GUI designed by someone who has no idea of what that task actually entails is madness.

Would you do complex engineering with a checklist which looks like it was written as a first year project and which imposes the process on you, but can't me made to actually match the real world?

Hell, on numerous occasions I've been on the receiving end of some bloody accountant trying to apply his idiotic metrics to something which can't be quantified readily ... why, no, I can't quantify the way in which I will find and fix bugs in a way which is meaningful to an accountant ... and, no, your standard template document has nothing to do with be solving a tricky problem of semantics.

One size really doesn't fit all. Some sizes don't fit anybody.

Just solve the problem already

[Taxman415a]

EHR systems are a horrible burden on healthcare providers and as they are currently implemented they offer very little of the benefits to the patient that they could. The UI of the EHR system is implemented essentially only for back office use and the provider interface is bolted on as an after thought. It's extremely clear from even a cursory look at the EHR systems that there was little if any thought given to optimize the workflow for the provider. In a given patient appointment, the provider has to click through various functions each of which requires descending 8 levels of menus to click, then wait for the several second delay and back out 8 levels and decent 5 or more levels for the next round. Patient report not being happy that their doctors are staring at a screen the whole appointment, but with the inefficiencies built into the UI it's literally impossible not to. In addition one of the main theoretical benefits of EHR systems that providers can pull up your health history and make decisions based on all of the information doesn't work because the different systems don't really interoperate as they were supposedly required to do. If you see a specialist that's on a different EHR system you either can't actually access the information without sending IT a request for that information and waiting for it to be made available or it will be in some even more horribly inaccessible format such as an image. Instead of wasting time on apps and analytic tools there should be some real teeth implemented into the interoperability requirement. Instead of being paid Billions of dollars to make systems that have only fake compatibility, they should be required to come up with systems that interoperate seamlessly. I'm going to take a bet that if there were some real, serious teeth implemented such as no government payments to the EHR providers anymore, the interoperability problem will suddenly vanish. I'm not a fan of heavy regulation in general, but when the companies have taken Billions to meet a requirement and they have managed to implement it in name only, then it's time to pull out the big guns. Don't get involved in the details of fixing the interoperability unless they fail again after being faced with serious consequences. Thing is they probably won't, the problem isn't really that hard to solve given the amounts of money spent. The companies currently don't want there to be interoperability because the current lock in benefits them. When that benefit is eliminated they'll fix the problem quickly.

EHR Developers are not EHR Daily Drivers

[ErichTheRed]

The state university health system that most of my doctors belong to started using EHR software in earnest about 6 or 7 years ago. It amazes me that the designers and developers of EHR software seem like they design stuff that's intentionally frustrating to use. I've seen worse UIs, but they tend to be for things like buzzword-compliant ITIL based service desk ticketing software, or things that are so proprietary that a functional GUI is not something the customers will pay for. Every time I've gone for an appointment, especially when I'm a new patient (even within the same health system,) the first 10 minutes of the appointment is a frustrated doctor asking question after question, followed by 6 keystrokes, 20 clicks, dropdown here, expand button there, etc. etc. etc. It's as if an offshore code factory was handed a spec, coded exactly to that, and no integration work was done to ensure it would be usable -- and I wouldn't be surprised if that was the case. You might say doctors are a pampered, privileged class who are used to having nurses and medical assistants to do all the "work" but from what I've seen the software is a mess. My dermatologist gave me a "tour" when he found out I was an IT guy -- if I were a doctor I'd be running back to the paper charts in a flash.

Contrast this with the industry I work in -- airlines. Yes, it's old, proprietary, ancient, slow dinosaur technology.at the core, but the GUIs are designed for maximum throughput. An experienced reservation agent can do a booking in under a minute without taking their hands off the keyboard, and everything in the application is actually designed to minimize cognitive load. As an example, I've never worked behind the counter on real passengers, but I can sit down in front of the GUI and understand the flow, look stuff up, etc. That's because the reservation system companies do actual time-and-motion studies and watch real people use the product. I highly doubt the EHR companies do this, nor do they have anyone on staff who uses their software regularly.