EMR doesn't stand for Electronic Medical Record, it stands for Excruciating Medico-legal Regurgitation [of poorly organized data]

Thankfully, I'm not the only person frustrated with the epic failure of current EMRs. Dr. Pandey has an insightful article on KevinMD:

http://www.kevinmd.com/blog/2012/12/explaining-epic-failure-emrs.html

"I think there are two reasons for such seemingly epic failure. First, how we interface with an EMR. Second how the EMR tries to impose its will on to us instead of the other way around."

 "The constant juggling not only made the patient encounters time consuming and cumbersome, it literally made us curse at the computers and ruin an otherwise perfectly normal day at work. Patient volumes have gone down from 15-16 patients per day to a half of that after EMR adoption."

EPIC EHR: Epic failure in interface

Some of the hospital systems I work in have decided to adopt EPIC as their enterprise EMR/EHR. I am a big advocate of electronic health records; when done correctly, they can be an incredibly powerful tool for protecting patients, improving communication, and improving efficiency.

However, when tens of thousands of features and buttons are thrown together without any clear overall interface vision, with nary a nod to current standard interface practices, without any reasonable understanding of actual workflow practices, what you may get is EPIC.

What Apple taught us is that interface matters -- and if you take the time to build a thoughtful interface, then applications are more useful, have shorter learning curves, and cause much less frustration. Those are all nice features in a gaming app, but they can be potentially life saving or life threatening when they exist or don't exist in an electronic medical record system. I have spoken with providers who have been using EPIC for more than a year, and they are still tearing their hair out and are frustrated. And that's despite countless hours of training. I am looking around at my physician colleagues, some older than 50 or 60 years old, and  I simply can not imagine them interacting proficiently with this system.

This is not to say that EPIC is the only EMR failing in interface design. As far as I've seen, the vast majority of EMRs are convoluted design failures, put together in haste in order to hoodwink hospital executives into thinking they are dream systems, but in reality they slow down physicians by unacceptable degrees and are pushing a large number of us to look for jobs outside of medicine altogether.

Here's what I'd like to say to Judy Faulkner, the founder and billionaire CEO of EPIC:
1) You know you have built your interface correctly when you can reduce your training from 12 hours to 1 hour.
2) Take some of the hundreds of millions of dollars you have earned over the last 2 years and hire a designer to redesign the interface for EPIC. Send the designer to some EPIC training classes that providers are forced to attend and watch new providers try to interact with your system. You will be amazed at how much room there is for improvement..
3) Create an open API for EPIC so that this wonderful world of designers and entrepreneurs who are out there could build a much more beautiful and simple interface.
4) Why are lab results spread out to take up as much space as possible, instead of using long agreed upon standard notations that all doctors use?



5) What is the difference between "chart review", "results review", "review visit", and "snapshot"? Would a new user to your system immediately understand the difference? Why not use terminology that every doctor already understands: "HPI, Exam, Procedures, Results, Prior Records"? Look at the left hand column iTunes, or any other modern interface, and you will see an organized interface that makes sense.

Here are some actual quotes from our training class:

If you want to document a "call-in" (when a doctor calls the ED to tell us he/she is sending us a patient), don't click on "phone calls" (why would you do that?)

If you want to admit a patient, don't click on admit radio button, it doesn't do anything.  

If you don't know a patient's date of birth for a call-in, just make one up. [That's not going to cause any confusion later!]

Clicking "Active" will make a macro either present or hides it. (????) 

I would love to hear other people's experiences with PICIS and other EMRs. Is there any enterprise system that is actually functional and easy to use?

Health Maintenance - Jiffy Lube for Humans?

The average adult spends $30-150 every 3 months to maintain their car, but what do they do to maintain their body, which presumably is more important?

When's the last time you checked your cholesterol? (compare that with your last oil change)

Blood pressure?

blood sugar?

triglycerides?

percent body fat?

Think about this: jiffy lube took the really cumbersome and long task of car maintenance and made it  routine, focused, and fast. Maybe the same could be done with health maintenance?

Just like with cars, if there's something actively wrong, you have to go the standard route (see a mechanic/doctor/hospital), but if you're just doing routine maintenance, you don't need all that. 

What if regular health maintenance check-ups raise your health awareness and end up motivating you to eat/act healthier? What if insurance companies offer cheaper insurance rates if you show a record of regular maintenance and good numbers?

An Insider's View on Upcoding and the Center For Public Integrity's Article on 'Cracking the Code'

The Center for Public Integrity has a piece titled "Cracking the Codes" which is making waves in the medical world over the last several weeks, as it has revealed data indicating a massive increase in Medicare billing by hospitals and physicians ever since the federal government began incentivizing physicians and hospitals to start using Electronic Medical Records (EMRs).
The immediate (and understandable) fear is that doctors are abusing the built-in efficiencies of EMRs to over-document and/or over-bill. For novice readers unfamiliar with ER billing, here is a quick primer: ER charts are coded into 5 separate categories (plus a sixth for critical care), based on the complexity and risk of the case. The complexity and risk is derived from the ultimate diagnoses, the number of history (the patient's version of his complaints) elements involved, number of physical exam elements (number of body systems examined) examined, and the medical decision making (MDM). MDM is complex, but essentially describes how much thinking, researching, communication, and risk that the case involved. Most ER physicians document the chart, while a separate coder codes the chart; the coder reviews the chart and assigns a billing level to it.
As a physician who began his training and practice during the not-too-distant era before the ubiquity of EMRs, I can tell you that as reimbursement rates have plummeted, and the number of uninsured has sky-rocketed, the focus on documenting fully and accurately has intensified dramatically. There used to be a time when graduating ER residents knew little about RVUs ("Relative Value Units", essentially the number of points assigned to each type of billing level or procedure, which translates directly into dollars). But now, every graduating high-energy resident is looking for an ER job with "RVU based pay" instead of an hourly rate. On an RVU based pay structure, the more patients you see and the better you document, the more you earn. Before the cynics jump on this idea and butcher it, let me just explain that RVU based pay is part of the reason you can walk into any one of our EDs and be seen in less than 30 minutes these days, whereas waits of several hours were the norm in the past. In the past there was no financial incentive for ER doctors to see the next patient, only additional work and risk. Now, ER doctors are eager to not only see the next patient, but also take care of the high risk (and therefore higher RVU) cases. RVU based pay, I argue very strongly, has played a significant role in improving patient safety by clearing out waiting rooms, and incentivizing doctors to see the higher risk cases.
With this growing push to see more patients, see them more quickly, and achieve stellar patient satisfaction results, EMRs have focused on reducing the documentation inefficiencies present with handwritten charts. Imagine hand-writing or typing the same 2 page letter 25 times a day, with just 5-10 sentences different in each letter. How much time would that give you to actually talk with patients and families and to focus on medical decision making? I have a normal full medical exam that I perform on every medical patient I see ("medical patient" refers to complaints such as abdominal pain and chest pain), the same way a baker prepares his dough the same way each day, or a housekeeper has his cleaning sequence, or a teacher gives the same Magna Carta lecture each year. Rhythm is a very powerful force that gives assurance of quality and consistency (two things *every* industry seeks). EMRs have been incredibly helpful in  recognizing that a lot of our documentation is redundant, and creating solutions to help reduce wasted time and effort in reproducing the redundancies. This has resulted in *tremendous* increased productivity by ER physicians, allowing them to see on average 2-2.5 patients per hour in the ER, where 10-15 years ago they were seeing barely 1.5. You can not underestimate how important this increased efficiency is: just talk to families of patients who crashed or died while waiting for hours in waiting rooms of yore.
What is the cost of productivity based pay and the increased efficiency brought about partially by more advanced EMRs? Charts are much more complete than they used to be (RVU based pay incentivizing doctors to documenting everything they do) and they look a lot more similar than they used to (EMRs allowing us to easily replicate the 90% that is the same on many patients). Whereas in the hand-written past there was some variability in the documentation of the 90% that I do which is the same on each medical patient, now there is very little to none. This, the Center for Public Integrity attacks as chart cloning. I see it as the byproduct of yet still inefficient documentation systems trying to relieve some of the ever-growing *massive* documentation pressures on doctors that distract from what is actually important: sitting with my patients and families, listening, and providing care and reassurance.

Time Shifting Emergency Department Demand

Om Malek, posting on GigaOM, has an interesting article on a startup which time shifts power consumption of appliances attached to its smart grid system from peak hours to off-peak hours. Unused energy is sold back to a local grid operator, creating revenue, which is shared with the customer.

Managing Emergency Department volume is surprisingly similar to managing electricity demand. There are massive fluctuations in demand at different hours of the day. Both require massive amounts of infrastructure in order to manage peak demand. Both systems have a potential for life threatening consequences if peak demand is not met. Peak demand is affected by weather and day of the week (well documented in ED literature).

In addition to the obvious similarities listed above, what is most important is that total demand is determined by the sum of independent events, which are often convenience choices by consumers: "turn on dishwasher now or in an hour", "go to the ER now for my cough or wait until the morning."

The inherent complexity of managing ED patients makes managing ED volume even more challenging than managing electricity consumption. Whereas electricity use is either "on" or "off", and all the consumer needs is either more or less of one resource, each ED patient may require a multitude of different resources, some very scarce or hard or expensive to perform concurrently, such as ultrasound or CT scan.

And when dealing with complex systems, it is well documented that efficiency often declines dramatically once capacity reaches 70-80% of total maximum capacity.

Therefore, it becomes even *more* imperative for EDs to employ techniques to reduce demand during peak hours, just as power companies are learning to do the same with electricity demand. It is *not* sufficient to just "call in the on-call doctor", because ED patients require more than just a doctor; they require nurses, techs secretaries, beds, x-ray techs, etc. Unless you are calling in additional ancillary staff, your additional doctor is most likely not going to be very efficient. The concept is "don't spend money on increasing infrastructure to meet peak demand; rather, spend resources on smoothing out demand.".

Although we discharge ~80% of our patients, not all of those discharged should have their care delayed (e.g. acute fractures). However, every ED provider will attest to the vast number of patients whom we treat who could have been safely managed either several hours earlier or later.

This leads us to the obvious point: in this era of growing disruptive innovations in healthcare, it is time to develop ways to reduce peak demand on emergency departments. Companies which can assist patients in seeking non-ED based solutions for their medical care, or can safely redirect non-emergent ED patients to return during non-peak hours, will have great demand for their services.

Health Datapalooza 2012: Healthcare innovation explosion

The Health Datapalooza in Washington DC showcased a number of innovative companies, technologies, and people who are leading the healthcare innovation revolution. These innovators are the tip of the iceberg of disruptive innovation piercing the titanic healthcare system, overladen with the burden of poor access, asymmetric information, lack of interoperability, and misaligned incentives. These innovations may transform the future of emergency medicine, often considered the eye of the storm of wasted health care dollars.

Symcat, the winner of the conference's $100,000 Robert Wood Johnson Foundation Aligning Forces Challenge, directly addresses the bane of the emergency medicine physician, what co-founder Craig Monsen called the "cyberchondriac" patient - the patient who is convinced he has a spinal cord tumor because he strained his back this morning and "on Google it said... and I want my MRI now!" Symcat is more than a beautifully designed interface - it algorithmically calculates the likelihood of each differential diagnosis based on data mining vast sources of symptomology correlated with disease occurrence.

The algorithm result improves as it leads the patient through relevant data points, such as age, sex, and additional historical factors. More importantly, Symcat helps avoid the unnecessary and expensive ED visit by then guiding the patient through his healthcare options and the anticipated cost of each option. Symcat even links the patient to an available appointment with a physician through ZocDoc!



ZocDoc is taking the physician appointment world by storm. If you haven't seen it or used it yet, you will soon. ZocDoc is an online appointment scheduler which allows patients to directly sign up for available appointments with listed physicians. Growing rapidly in major metropolitan centers, ZocDoc corrects the mismatch between physicians who have available appointment slots and patients who don't have an efficient mechanism to find an available appointment (so they end up in your ER). Oliver Kharraz, cofounder of ZocDoc, was a panelist at the Datapalooza and I got to speak with him about the relevance of his service to Emergency Medicine. Not only does his service help prevent unnecessary ED visits, but it is an invaluable tool for helping ED physicians arrange follow up appointments for patients after hours. Concrete follow up, we all know, is essential to reducing our liability.



Understanding the need for the growing number of ACOs and TPR hospitals to prevent unnecessary ED visits, triage.me, by founders Dan Wilson and Mark Olschesky, is a wonderful little web app which helps direct appropriate patients to local clinics instead of the ED using a simple clean interface. The app even helps you make an appointment, find directions to the clinic, and get prescription discounts. It is easy to imagine handing out brochures to your under-insured ED patients with helpful resources such as triage.me.

Another important tool every ED physician needs to learn about is iBlueButton from Humetrix. Blue Button was established by the VA, Department of Defense and CMS, and it allows patients and physicians immediate access to patient records. Blue Button "has already been adopted by over 900,000 Veterans, service members, and Medicare beneficiaries" and "more than 200 private insurers participating in the Federal Employees Health Benefit Program (FEHBP) are now required to offer a Blue Button functionality to the over 8 million individuals covered by the FEHBP program."  Blue button is a consumer-mediate health information exchange which should allow for better healthcare communication and decision making, allowing increased and better informed participation by patients in their healthcare decisions. It is available as an iOS app. For a video demonstration of this product, watch the video:

Stanford Med using Khan Academy tactic to improve physicians collaboration and learning

Techcrunch has a great piece on Stanford medical school flipping its didactic model so that students listen to lectures at home and collaborate on problem solving projects in the classroom. Techcrunch posits that this is more innovative than Harvard and MIT posting their lectures online; I agree.

Having sat through hundreds of laborious hours of monotone recitations at the University of Maryland Medical School, I can attest that lectures are not an efficient nor effective method for transferring information to students like me. Watching videos at home allows each student to play, pause, and repeat any portion at any time without any embarrassment or guilt. If there is a word or an idea that isn't clear, the student can pause and research something immediately on the web, and gain the appropriate contextual understanding that allows for real learning - this isn't possible in a live lecture.

Using classroom time learning to work with other students on complicated medical issues instills a very important skill which is all too lacking in current medicine: collaboration. Too often patients come to me in my ER and it is evident to me that the patient's four physicians (eg internist, cardiologist, gastroenterologist, oncologist) have not really been talking to each other, that each of them is changing the patient's medicines without consulting the others in real time. The patient is left alone trying to make sense out of sometimes conflicting recommendations. Mayo clinic understands the importance of real time collaboration of physicians, and it's major success, I would argue, is based on this core feature.

Some companies recognize the critical nature of the current broken communication system for physicians, not made easier with the tremendous boundaries erected by HIPAA. It is great to see innovative companies like TigerText creating new and improved methods of communication for doctors.