The Failed Promise of Electronic Health Records


Credit: Tim Martin | Dreamstime.com

Health care only got two mentions in President Obama's first inaugural address: One was a declaration that "our health care is too costly," the other was a promise to "wield technology's wonders to raise health care's quality and lower its costs."

A few weeks after giving the speech, Obama would follow through on the promise by signing the American Recovery and Reinvestment Act—the $833 billion law commonly known as the stimulus.  The stimulus included about $20 billion in funding for incentives designed to spur the adoption of electronic health records.

The hope was that converting the nation's health records to electronic systems would make health care cheaper and more efficient. Analysts at the RAND Corporation estimated that adoption of the new computerized systems could save about $80 billion over a decade. Similar studies concluded that not only would adoption of the new systems result in savings, they would promote health by enabling better preventive care and chronic disease management.

But four years later, the projected benefits have failed to materialize. Earlier this year, RAND researchers released a follow up study concluding that the savings just aren't there.

One big problem: Small practices, older practitioners, and specialists have been slow to adopt the systems, despite the existence of taxpayer-backed incentive payments of up to $44,000. Simply installing those systems—which can be quite expensive, costing even more than the incentive itself—isn't actually enough. In order to get the incentive payment, providers must meet "meaningful use" standards. And according to the New England Journal of Medicine, fewer than 10 percent of specialists and 18 percent of primary care providers were using the technology well enough to get the federal bonuses.

Another problem is that a lot of doctors don't seem to particularly like the new systems. Among the doctors who are using new records technology, many complain that it actually slows them down or makes it harder to interact effectively with patients because they're stuck typing on keyboards and staring at computer screens.

Interoperability, or rather the lack of it, is also a big problem—and perhaps the biggest snafu of all.

Photo credit: The Doctr / Foter.com / CC BY-NC-ND

The systems were supposed to produce digital records that could be shared across providers and move with patients as easily as email. But RAND's follow up study states flatly that "the health IT systems that currently dominate the market are not designed to talk to each other. A big part of the issue is that IT vendors have sold providers systems designed to lock them into business with a single vendor, not easily communicate with records created and maintained on other systems. RAND's study likens the current system to frequent flier cards usable only with a single company rather than ATM cards usable anywhere.

You can imagine a number of reasons why this might have happened: The stimulus was passed fairly quickly, so mandatory interoperability slipped through the cracks; hospitals and other health providers each bought and managed their own systems rather than attempting to coordinate; doctors are independent operators focused on health more than on technology and administrative efficiency.

But why it happened is in some ways less important that the simple fact that it did. We blew a sizable chunk of taxpayer money on getting providers to install these systems, and we're not seeing the results we were promised. In fact, if anything, we're seeing the opposite: The new systems actually appear to help doctors bill better navigate Medicare's complex billing system, making it easier for them to charge the federal government more.

This is why I'm so skeptical of Obama's more recent promises to control Medicare spending through "modest reforms" that tweak incentives and payment systems. We've been trying modest reforms for a while. And in many cases they just don't work.

(Thanks to Aaron Caroll for pointing out the NEJM letter.)

NEXT: Noah Berlatsky on America's Imperial Ad Men

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of Reason.com or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  1. The stimulus was passed fairly quickly, so mandatory interoperability slipped through the cracks…

    Those would have to be pretty big fucking cracks, considering the stated goals. It would have only taken consulting their office’s IT guy once to get a heads up on the importance of compatibility.

    1. Look, do you want a politician who’s going to stand around jibberjabbering over the minutia of how a law will work or how to pay for it or whether its constitutional, or do you want legislators who Do Something?

    2. “So did you hit ESC and CTRL R? And then….MOVE….”

      /IT guy

  2. Here is what confirms for me that Obama is an evil piece of dog shit. He signed Obamacare into laws and yet he has the temerity to say that we still need healthcare reform. WTF?

    Then, he raises our taxes to “make things fair” and having done that, says we need to raise taxes to make things fair.

    Now I would never advocate killing the president, but one can’t help but wonder if there would have been a single signatory to the Declaration of Independence who wouldn’t have shot the fucker on-sight for treason.

    1. Treason may be the most retarded of crimes.

      1. How so?

        1. Because there is no actual harm and no actual victim. You’re just acting against the state and the people who operate it.

          1. I always KNEW you loved teh terrrrrrstzzz.

  3. It failed because it is impossible. I have work hospital/research IT for 20 years. This was way oversold. Just for a huge problem you glance over interoperability is HUGE. Next doctors visit look at all the different electronic gizmos they have (temp/blood pressure/hemoglobin readers/etc.) that is just an office imagine an OR. All these instruments have to be interfaced with the LIS. then add in all the different manufacturers. Then add in any new instrument they buy. It is a one time thing it is on going thing. Interoperability did not slip through the cracks, morons that know nothing about interoperability wrote the law. Usability always an issue, writing on paper is always easier, but using the data is impossible (basically a stupid complaint) – however for a small practice the benefit can not possibly justify the cost. the nonsense that has been spewed over EMRs boggles the mind.

    1. morons that know nothing about interoperability wrote the law.

      “We never claimed that it was a perfect law.”

    2. It’s not close to impossible, though it may be impossible for something driven from the top down.

      1. yes it is impossible. there are many layers of interoperability. I just outlined hospital level. then you have hospitals sharing data. I have worked on public health lab implementations of this software (lots of money in it 10 years thanks to bioterrorism grants). Do you know how much granularity there is to this data, how many hospitals call things differently or have different tests that give the same result? somehow you have to translate it all, so x is x even when it is called y. For a fun example when coming up with options for gender do you know how many options we had? If you think 3 you are way off. If you think 15, you getting closer.

        1. I worked with integrating electronic insurance documents from numerous sources into a common structure for years. And this was frickin’ IBML OCR before you could even touch the data. Same with mortgage documents.

          It’s not close to impossible. It’s not trivial or cheap, but it’s not impossible.

          1. Oh you did it with mortgage documents that are at most 50 pages and 90% legal boiler plate, so no problem. Let me let you in on a secret there are codes for tests for billing medicare and all or most hospitals have standardize on these. Now you with your mortgage document integrating experience are probably saying just use those, there is no problem. And you would be right, I used those on the public health project even used HL7 another industry standard and it worked. Now a couple of things, those codes do aggregate tests that are similar. Now in public health when dealing with trends and aggregate that does not matter. But would you want your doctor to be making decisions based on a result that could vary by 10% based on the method? what if you were on the operating room table? Now come up with code for every test method, simple right only a few 100,000. Now what about units (we aren’t just using dollars here). And of course the data itself can run in the gigabytes (genome, mri, xrays, etc.) so you will need the infrastructure to support that. Mortgage document please.

            1. I have to agree with mb.

              I am a developer and currently in the midst of working on an EMR project for a private practice in a particular field. Their practice consists of 1 doctor and perhaps 10,000 patients over the many years. They’ve used an EMR system and are switching to another due to stagnation of development on the old one.

              They will end up spending over $100K on this. That may sound ridiculous, but the current system – which is woefully under featured and is designed specifically for this field – has hundreds of interrelated tables and forms. Even just tracking the results, compliance, and measurements for a single protocol (of a dozen choices) to treat a particular issue (of hundreds if not thousands) is staggering.

              EMR is the way of the future and could offer substantial benefits, but its complexity is frequently underestimated even by those intimately involved in it.

    3. “Goddamnit, mb, I pay you to SOLVE fucking problems, not tell me you HAVE problems. FIX IT…”

      /President Obama

    4. If they’d wanted to ensure interoperability they should’ve just said that everything would have to be moved into alignment with VistA (the VA developed a robust EMR solution on the taxpayer dime so it’s open source via foia request).

      I work with a bunch of older doctors. The front desk has taken to telling our head physician that I’m ordering him a laptop. The comes close to tears each time.

      1. Yes, making the VA EMR nationwide would have made a lot of sense, wouldn’t it? So many of us said so at the time.

        But then big crony types like Siemens couldn’t have made billions for their piece of shit forays into EMRs.

        So logic failed again, and instead everyone got pushed in to expensive de novo systems that weren’t ready, so they wouldn’t miss out on the stimulus money.

        The clusterfuck that we got at our medical center can’t even do basic things that you could do on Microsoft Word with Windows 95. When we asked programmers about things like “autocorrect” they looked at us dumbfoundedly, as if we were asking for rocket ships to Mars.

        Our system doesn’t even tell nurses when you have entered an order for a patient. We work in the goddam ER! Some patients have been going hours without their meds because no one knows there is an order waiting unless they open up the patient’s name on the screen to specifically look for it. We asked if perhaps a patient’s name could change color, or blink, when there was a pending order — the programmers said that was way beyond the system’s capacity. Jeez, a kid in a 1960s science fair, with a dry cell, some copper wire and a couple of Xmas bulbs could do that, but no, not our $60 million system.

      2. I am familiar with the VA system. I have even installed and evaluated it. I was not impressed. One it is not an EMR, it is a part of an EMR (incomplete). Two the VA I ride past everyday is implementing something else (I have been contacted about working on the project), maybe to work alongside the current system, maybe to replace it, I don’t know – I do know their system is not sufficient. The federal government has tried to implement many systems like these, failed on every occasion (google fbi lims, dea lims, nih lims (they are still going but way over budget)), building your own is moronic, installing one is hard enough.

        1. I’m not familiar with the VA’s systems, but I wonder how robust their system is for dealing with procedures that aren’t probably called on by their core constituency. I’m mostly thinking of modern or elective areas of fertility (IVF, etc.) or vanity (LASIK, etc.).

  4. I hereby declare a new category of bogus argumentation: Appeal to Magic.

    “If we just get everybody a _________, this problem will solve itself.”

    1. This is funny because it’s TRUE!

      When I started out in programming (mid-80’s – my first REAL job!) when we had shit that just didn’t work, and we were figuring out how to fix it (if we could), we’d draw a box in the logic chain with the letters “FM” inside it.

      We’d talk through the logic chain, get to the icon, and somone would ask, “What’s ‘FM’?” “Oh! That’s ‘Fuckin’ Magic’, because that’s what has to happen there, cause the program won’t work…”

      I STILL use the “FM Box” at work. Just dropped it the other day re: some intractible issue we were working on. Too funny.

      1. So basically a programming version of:

        1. steal underpants
        2. ?????
        3. profit

  5. Sometimes dude you jsut have to roll with it man.


  6. This was never designed to lower costs. It was designed to give the fedgov an easily searchable database of drug users and gun owners.

    1. That’s what I always assumed – hasn’t the government already decided that electronic records under someone elses’ control don’t fall under the 4ths’ protection against warrantless search and seizure? Those medical records aren’t *your* papers.

  7. Let’s see how many more promises will fulfill & how many fail.
    Certified Financial Advisor

Please to post comments

Comments are closed.