The HIT Job

The HIT Job

44
SHARE

I’m well aware that a good fraction of the people in this country – let’s call them Rush fans – spend their lives furious at the New York Times. I am not one of them. I love the Grey Lady; it would be high on my list of things to bring to a desert island. But every now and then, the paper screws up, and it did so in a big way in its recent piece on the federal program to promote healthcare information technology (HIT).

Let’s stipulate that the Federal government’s $20 billion incentive program (called “HITECH”), designed to drive the adoption of electronic health records, is not perfect. Medicare’s “Meaningful Use” rules – the standards that hospitals’ and clinics’ EHRs must meet to qualify for bonus payments – have been criticized as both too soft and too restrictive. (You know the rules are probably about right when the critiques come from both directions.) Interoperability remains a Holy Grail. And everybody appreciates that today’s healthcare information technology (HIT) systems remain clunky and relatively user-unfriendly. Even Epic, the Golden Child among electronic medical record systems, has been characterized as the “Cream of the Crap.”

Moreover, in the last few years we’ve gained a deeper understanding of the hazards of HIT, including new kinds of errors created by the stormy marriage of imperfect computer systems and fallible humans. We’re also becoming familiar with subtler problems, such as the copy and paste phenomenon now plaguing progress notes and the degree to which computers can distance us from our patients (Abraham Verghese’s “iPatient”). These problems are all-the-more irritating since IT was hyped – overhyped – as the solution to so many of healthcare’s woes.

So it’s natural to be disappointed in the present state of HIT, and even to wonder whether HITECH is on target. But overall, I believe that we are on the right track, that there is no better way to get to an HIT Promised Land than the path we are taking, and that the Federal government should be commended for getting involved in a reasonable way.

In the February 19th issue of The New York Times, one of the most off-base, unbalanced articles in my recent memory paints a very different picture.The paper’s lead article – yes, Page 1, Column 1, above the fold, called, “A Digital Shift on Health Data Swells Profits in an Industry,” focuses particularly on the consequences of HITECH. The reporter, Julie Creswell, does raise some new and legitimate concerns, such as the degree to which HIT vendors have jumped into Washington’s toxic swamp of money and politics.

Yet the tone of the article is inordinately conspiratorial about HITECH, and downright dismissive regarding the overall value of HIT. Creswell’s sources are disproportionally slanted to HIT skeptics, including her choice to quote my UCSF colleague Mike Callaham, who pronounced our Epic system “lousy.” (I’d venture to say that most people at UCSF – including me – find the system imperfect but pretty good. Moreover, we switched to Epic, a company that Creswell paints as a Goliath, only after a dismal experience with a different system built by GE, a $240 billion behemoth of a company whose electronic health record product is foundering in the HITECH world.) After reading the Times piece, I found myself in full agreement with Mark Hagland, editor of Healthcare Informatics magazine, who wrote,

The opening one-sentence paragraph says it all. Creswell writes, of a presentation by the Chicago-based Allscripts to physicians in 2009, “It was a tantalizing pitch: come get a piece of a $19 billion government ‘giveaway.’” First of all, characterizing HITECH as a “giveaway,” without in any way mentioning the penalties embedded in the law for providers who haven’t implemented electronic records (EHRs) by the end of 2015, is simply irresponsible journalism….

Ms. Creswell could easily have noted that no other large industry in the United States remains even remotely as paper-based as healthcare… or that study after study has confirmed the benefits to patient safety, care coordination, and cost-effectiveness of the automation of patient records. [The story] will undoubtedly be read by many thousands of laypeople who may or may not have any sense of how misguided and distorted its core thesis is.

The Times article ends as skewed as it begins: with a backhanded reference to the “gold-rush mentality” of today’s HIT players.

Let’s pause to ask a few questions: Does anyone honestly believe that computerizing American healthcare is wrongheaded? Or that the correct strategy was to continue toe-tapping, waiting for “the market” to promote IT adoption when, in 2009, only 16 percent of both US hospitals and doctors’ offices had functioning clinical IT systems? Or that they would like to be a patient, or a clinician, in a paper-and-pencil hospital?

I didn’t think so.

In 2004, then-president George W. Bush asked David Brailer – a brilliant MD and PhD in economics – to direct a new federal office of healthcare information technology, whose primary goal was to promote IT adoption. Brailer and his staff realized that a top-down program that had Washington forcing computer purchases on doctors and hospitals would be a disaster. (In fact, such a strategy was adopted by the UK’s National Health Service – a centralized, command-and-control initiative that, in 2011, was deemed a fiasco and junked, at a cost to British taxpayers of $19 billion.)

Instead, Brailer began planning a program consistent with American values, one that would allow physicians and hospital leaders to choose their own vendors, and encourage market competition. The plan that he and subsequent heads of the Office of the National Coordinator for Health Information Technology (ONCHIT) developed was to fashion a set of standards – relatively easy to reach at first and progressively more ambitious over time – and then to find the cash to fuel a national incentive program. They found the booty (about $20 billion worth) in 2009 when Congress and the president were seeking “shovel-ready” projects to include in the $700 billion federal stimulus package.

That, my friends, is the crux of the story. The meaningful use standards were developed and disseminated after extensive public comment. The IT vendors, a sleepy industry of true survivors (many of them barely maintained their pulses for over a decade, just hoping that the day would come when HIT adoption finally tipped to their side), did what all U.S. businesses would do when the feds were considering tossing money in their direction: they hired lobbyists and made campaign contributions. I don’t love this (and there is a risk that the Epics of the world will succeed in thwarting competition by scrappy upstarts), but that is our system, and any responsible business would have done the same thing – in fact, they’d be stupid not to.

It’s not a conspiracy. It’s America.

And trumping everything, the program has worked. The HIT adoption curve, previously stuck on flat, is now extraordinarily brisk: by 2011, 35 percent of US hospitals had functioning electronic health records, more than double the percentage of 2009; a similar surge has been seen in outpatient practices. The literature continues to demonstrate that, overall, these systems do reduce medical errors and harm. The promise of easier data collection to fuel transparency, pay for performance, and quality improvement activities will soon be realized, and we will ultimately enter the long-awaited world of “Big Data” in healthcare – one in which we can aggregate patient-level data on millions of patients, paving the way for more efficient methods of determining best practices and risk factors.

There have been problems. The systems are not great, and the vendors – including Epic – are putting most of their energy into keeping up with the insatiable demand for installations, and relatively little into improvements. Studies have demonstrated that the promised productivity gains have been weak to nonexistent. We have the aforementioned problems with IT-based errors and new challenges to clinician-patient communication. But the history of IT innovation is one in which systems become optimized only after many cycles of user feedback and vendor improvements. That cycle is beginning to play out, and the result is sure to be better, more mature HIT over time. There is no shortcut.

The Times (full disclosure, my wife Katie Hafner, writes about healthcare and technology for the newspaper) has covered HIT, including its glacial pace of adoption, responsibly, up to now. The Creswell piece, by failing to acknowledge the value of healthcare IT, the absolute necessity of wiring our healthcare system, and the fact that a federal program to kick start this process was a perfectly reasonable policy approach, was unbalanced and unfair. While a cautionary note is welcome, one could come out of reading this article clamoring for a Congressional investigation of the HITECH program and of the activities of Epic, Cerner, and Allscripts. If our goal is to find ways to create an improved, and ultimately less expensive, healthcare system, such a response would be unwise, even counterproductive.

The HIT industry, and those who regulate it, don’t need any special favors, and the Fourth Estate should keep a close eye on things, particularly now that there is gold in them thar hills. But as journalists are drawn to the increasingly vibrant world of healthcare information technology, it will be important that they do their homework and strike a balanced tone. The Times piece, I’m afraid, was a HIT job.

Robert Wachter, MD, professor of medicine at UCSF, is widely regarded as a leading figure in the patient safety and quality movements. He edits the federal government’s two leading safety websites, and the second edition of his book, “Understanding Patient Safety,” was recently published by McGraw-Hill. In addition, he coined the term “hospitalist” in an influential 1996 essay in The New England Journal of Medicine and is chair of the American Board of Internal Medicine.  His posts appear semi-regularly on THCB and on his own blog, Wachter’s World.

Leave a Reply

44 Comments on "The HIT Job"


Guest
Problem Child
Mar 2, 2013

How, may I ask, that you people have been so duped by the HIT industry, its trade group HIMSS, and its offspring, such as CCHIT? There has got to be big bucks flowing into grants, consultancies, “safety” programs, faculty plans, CIO pockets, etc. Dr. Wachter has lost his objectivity. Why?

Guest
Whatsen Williams
Feb 28, 2013

Seechit is an interesting organization, spawned by HIM$$ and EHRVA, to legitamize the CPOE devices and provide the illusion that they were safe, effective, and usable, in their lobbying efforts with Congress.

CPOE systems meet the definition of a medical device in the FD and C Act. The FDA leadership has said that. Hence, they are CPOE devices.

Kreinky is comparing CPOE devices to drugs. Not exactly a good comparison, sir!

Guest
Kendra Williams, RN
Feb 28, 2013

In order for a drug to be sold, it must be approved by the FDA. In order for a prosthetic knee to be sold, it must be approved by the FDA. In order for an EHR and CPOE machine to be sold, it must be approved by no one. Go figure.

What I would like to know is who is getting favors and greenbacks in this scandal?

Guest
Aquifer
Feb 28, 2013

Well i have to admit, i hope that if we set up an approval agency for this stuff, it will operate to higher standards than the FDA …

Guest
Feb 28, 2013

EMRs are not approved by an agency – you can’t do that for software anyway, as it has no actual direct therapeutic effect (nor does it directly harm) a patient the way an agent introduced into the body can. Doctors, nurses and other providers still stand between computers and people, and that should never change. The only gray-area is dosing and automatic medication dispensing via software, and regulating them as devices is something HHS and the FDA is working on now.

EMRs are certified by an independent non-profit http://www.cchit.org/ that works with HHS to establish minimum standards for EMRs that will

Guest
Aquifer
Feb 28, 2013

Well i suppose I don’t quite understand what the point is in establishing “minimum standards” if there is no requirement to meet them …

In any case – my comment was primarily meant as “snark” re the FDA …

Guest
Feb 28, 2013

Certification of an EMR is required, or the federal government will not make the incentive payment to a physician for installing and using it. That might have been a nice thing for the NYT reporter to have pointed out, but space in the paper is limited and it’s hard to fit in a discussion of the broader context for those payments when you have an ax to grind about the industry lobbying.

Guest
Carl Simon, PhD
Feb 28, 2013

The HIT industry is a bunch of vampires, sucking the healthy blood from the medical care system. The New Tork Times opened the door to moor investigation of the flow of money in and out of the organizaions such as HIMSS, AMIA, CCHIT, AHIMA, and hospitals invested in their vendor.

Guest
Curly Harrison, MD
Feb 27, 2013

As in the UK, the costly HIT experiment in the US is failing, with little to show for it. The NY Times report was an excellent description of the slick conduct of the principals of the industry.

Guest
Whatsen Williams
Feb 27, 2013

The current state of HIT is indefensible. Until the devices being used to run the care of the patients are assessed for safety, efficacy, and usability, the entire programme is nothing but an unregulated experiment using the patients, doctors, and nurses as unconsented guinea pigs to enrich the HIT manufacturers.

Guest
Feb 27, 2013

Fantastic return to reality, Bob, though I’m a little surprised and disappointed that your excellent summary didn’t inspire the usual nihilistic tirades against everything wrong with health care, medicine, capitalism, democracy, and the human condition, with EMRs being only a proxy for them. (Though there were a few stabs at it from the usual suspects.)

I will admit that the NYT article wrecked my day, as I personally know all the people involved, and like to see them brought down a notch or two – but I also know the article was slanted, devoid of context, and bordered on muckracking.

As for the predictable parts of the response on here – and at the risk of sounding like an HIT champ’s version of the NRA – I’ll say only that it’s best not to blame a technology for its failures (in this case, little or no working interoperability) when you should be blaming the owners/operators of that technology, i.e., the executives who install it and set priorities about its implementation and use.

Brailer’s ONC knew on day one to create methods, standards and processes for data interchange, based on prior generations of work during the pre-industrial phase of EMR adoption. HL7, CCD, CCR, LOINC, DICOM – the whole alphabet soup, which tastes great intramurally – has been boiled, cooled, strained and thrown all over the walls of this blog since its own day one. And we all know (or should know) that entire careers have been made creating that soup, and twice as many fouling it. None of these standards are perfect but their greatest failures have been inter-murally, as none have been used toward even 5% of the task of moving now-computerized records out of one system and into the system of a competing provider’s system.

Oops – spilled the alphabet soup beans: note the word “competing.”

Blame it on RHIO? Not by the hair on my rinny-CHIN-CHIN. (For the millenials: what we now call HIEs were called CHINs, before they were called RHIOs.) The acronym thus perfected, the HIEs were finally funded, kinda, by HITECH, but then the money ran out before the lights went on, and now there’s almost nobody home.

Who REC’ked the HIE ball? Adam Smith did, with his invisible hand. Seems like everybody is now off working on the HIXs, now that they’re done not finishing the HIEs. This is not an accident, lobbying dollars driving the process or not. The failure to reach any kind of critical mass, so far, on true interoperability is an act of willful self-interest.

To wit…hospital CEO and CFO, thinking out loud, in meeting with CTO and CIO: Our patients’ data is our customers’ information, so explain why we should invest our limited time and money – diverting fixed resources from your aptly named Epic install – to make it easy for that information to cross the street to our competitor?

We do not have a lack of EMR interoperability because we cannot figure out how. Where there is an economic will, there is a technical and operational way. The real reason: hospitals and modern group practices are individual rational economic agents with limited resources, and they know (if only intuitively) that there is not only no good reason to cooperate and interoperate, but a countervailing reason NOT to. Sure, they have HL7 in there, so they can check off the interoperability box on the meaningful use form. But you don’t think any of them are actually going to help those CHIN / RHIO / HIE people, do you? They’re stuck in ENOUGH meetings with the Epic install!

It really is that simple. If you don’t believe me, see how easy it is to move your account history data from Comcast to TimeWarner, Hertz to Avis, Amazon to Barnesandnoble.com.

Guest
Dr. Rick Lippin
Feb 27, 2013

” I, for one, am ready to try something new.” says Dr. Wachter

I doubt it

Dr. Rick Lippin
Southampton,Pa

Guest
TBMD
Feb 27, 2013

Ron: Well said. I don’t have the technical background to know the next steps to bring us to “future state” but I agree that what we are doing now is not it. I’m all for disruptive innovation but it has to be innovative. Various versions of non-interoperable ehr’s have been around for at least 10 years. Implementation has accelerated over the last 3-5 years yet the systems are propitiatory, have poor user interfaces, hamstring the physicians, are massively expensive and drive up the cost of healthcare, do not support team care, adversely impact primary care (already in peril) and are unsafe for patient care. Please tell me what is good about this?

Dr. George Margelis says we are 1/2 way through and disillusioned. Any technology should have early wins as it is being developed. Where are the early wins? Where is the improvement with each new upgrade or version? The basic flaws keep getting perpetuated as more and more money is shoveled into the developers.

It’s time to take a breath and start aligning incentives that leverage technology that is out there but just not yet incorporated into ehr. I’m glad we are trying something new and I believe most clinicians are READY. But don’t push crap on us and expect us to say “Thank you master, may I have more gruel”.

Guest
Feb 27, 2013

“I believe that we need a strategy that will get us to a new place in our healthcare system in the next 5-7 years: interoperable electronic health records, accessible to both clinicians and patients, that can serve as a scaffolding for levels of decision support, advanced analytics, patient engagement, trainee education, and quality improvement that we can only dream about today.”

But that’s exactly what’s NOT being implemented today as a result of HITECH.

Systems that weren’t designed to be interoperable, to be usable, to support the kinds of complex analytics for useful CDS, to use common metadata, to support flexible workflow, to support team based care, etc., etc. can’t magically or incrementally morph into systems that can.

And that’s the problem. Much of the expense, time, effort and political capital spent on HITECH’s promised ROI, promoting, buying and implementing yesterday’s systems (and their closed architectures) is not well aligned with the goals.

You just can’t get there from here without MORE massive expenditure to rip and replace stovepipe systems, buy and implement new ones.

It’s as if the government provided incentives to purchase Dynamos (http://en.wikipedia.org/wiki/Dynamo) to provide electricity for each and every hospital in America…and then wondered why there isn’t an electrical grid.

Guest
Feb 27, 2013

“interoperable” is a misnomer that has become a cliche. Don’t expect to “operate” an Amazing Charts install from an e-ClinicalWorks client, etc. What would then be the point of different EHR “products”?

What we need is a master data dictionary standard. ONE standard. You “certify” to that comprehensive data transparency standard and then innovate away (or not) around it; features, functionality, efficiency, usability– which become your value/price differentiators.

The way we’re going now is to “promulgate a breadth of competing standards” so that, in effect, we really have none. Lotta busywork going on. The is ONE standard in the U.S. for 120AC there-prong grounded interfaces. You don’t have 4,000 ONC-CHPL “certified” sizes and shapes of those interfaces (the male & female plugs and receptacles themselves, excluding the differentiating cosmetics surrounding them on the racks and in the bins at Lowe’s).

Guest
Aquifer
Feb 28, 2013

I cannot speak from experience – but i have heard, here and there, that the VA system is pretty good – does anyone have an opinion on whether that might serve as a template for a standard?

Guest
J. Stefan Walker, M.D.
Feb 27, 2013

Very apt analogy. We need this single standard; ideally, from a convening of appropriate bodies of both HIT and the various specialty groups in the U.S. Why this has not yet occurred is a mystery to me.

Guest
southern doc
Feb 27, 2013

Follow the money.

Guest
Feb 27, 2013

It is natural for a degree of disillusion to foster half way through a major project, and I suspect that is what we are seeing here. At the outset everyone is excited. Half way through we get frustrated by the challenges, and see some people profiting up front whilst those doing the hard work are not seeing any significant rewards. However it is important to stay the course and focus on the end goal. This requires clinician collaboration and strong leadership. You seem to have that in the USA, Keep in mind the rest of the world is watching you, and look forward to leveraging off your success.

Guest
Feb 26, 2013

Bob & Bev, forgive my metaphor, but assuming that the government had to launch this entire HIT thing because it was not going anywhere on its own, and assuming that in the long run it doesn’t really matter if the launching mechanism was perfect or not quite so, the HIT craft is now soaring.
It is a disastrous mistake to keep the rocket boosters and the clumsy fuel tank attached to the craft, just so its trajectory can be better controlled from Washington. These things (i.e. meaningful use and such) are slowing natural progress now, serve no further purpose, and need to be discarded before they cause the entire contraption to crash into the ground.

Guest
bev M.D.
Feb 26, 2013

Bob, I definitely agree with your comment about insanity, and that also applies to the health care law. In both case the status quo just had to be disturbed, even if the disturbance was painful and not quite on target, it got the ball rolling.
I am not sanguine, however, that policies alone can get us where we need to go. Good old customer pressure for a better product is the only thing that will get us what we and patients need, and that’s hard to bring to bear when you can’t boycott the available product because the government says you can’t. Someone desperately needs to come up with a product that works, so everyone will flock to it, and the companies that lose out will quickly follow suit. How do we make that happen?

Guest
Feb 26, 2013

Thanks for all the thoughtful comments on my post – this is understandably an issue that people are passionate about. While I appreciate the concerns about the existing HIT systems and federal policies, I believe that we need a strategy that will get us to a new place in our healthcare system in the next 5-7 years: interoperable electronic health records, accessible to both clinicians and patients, that can serve as a scaffolding for levels of decision support, advanced analytics, patient engagement, trainee education, and quality improvement that we can only dream about today.

If you accept that as a goal, then the question is not whether Epic is perfect (it’s not) or HITECH is flawless (it’s not), but rather: What is the set of policies that are most likely to get us to that place, with the least pain and cost? I believe that the present strategy is about right. (Of course, like any policy, it’ll have to be tweaked over time as we get smarter, as will the HIT systems themselves, but that would be true of any path that we might choose).

For those who disagree, I’m looking forward to hearing alternatives that aren’t a blend of status quo and wishful thinking. Remember what they say about insanity and doing the same thing… We’ve tried the same thing in healthcare for 20 years – while every other industry computerized, to great effect – and we see where it has gotten us: low quality, insufficient safety, poor reliability, confused patients, massive waste, and backbreaking cost. I, for one, am ready to try something new.

Thanks again.

Guest
TBMD
Feb 26, 2013

If Epic is the best there is, we are in big trouble. I think it is a bit of “the emperor has no clothes”. Large systems spend $millions (billions!) on the system and cannot go back. It is the end users that suffer because the user interfaces are not build for clinicians and the system is clunky and cumbersome. Is this the best we can do? I agree with Bev that we must go forward but if we can’t be honest about the mess HIT is in, we can’t crawl out.

So far I do not see “better,safer care” with Epic ehr. It impacts productivity, safety and patient satisfaction. When the experts tell us “you will never get back to 100% productivity with ehr” one has to wonder who it is really serving.

Epic is neither innovative nor intuitive and was not developed for good clinical care. Period!