A Tale of Two Studies: What Are the Actual Costs of an EHR?

Does anyone in their right mind believe that these are the best of times in healthcare or health IT?

Scratch that.

Does anyone besides Judy Faulkner and Neal Patterson believe these are the best of times? (I mean, everyone knows that Dramatic Transition + Industry-wide Upheaval + Piles of Cash = Satisfaction / Contentment, proving the point mathematically.)

The question: At what cost to overall healthcare improvement do Epic and Cerner (and others, to be fair … except you, Allscripts) reap massive profits?

The short answer: We don’t really know.

While it is generally acknowledged by most (certainly not all, which you know if you’ve spent any time on HIStalk) that the ready availability and automated cross-checking of electronic health records improves care, there is no definitive study showing dramatic clinical improvement, demonstrable return on investment, etc.

Indeed, we now have a number of studies suggesting exactly the opposite:

  • The implementation of an EHR upends organizational structure and often slows down the provision of care.
  • The introduction of an EHR into a dysfunctional organization tends to exacerbate, not alleviate, said dysfunction.
  • Much of the promise of health IT is in interoperability, and the industry is a long way from reaching that goal.
  • Physicians generally dislike most health IT solutions.
  • Patients would rather the doctor look at them instead of the monitor.

This is not to say that healthcare should bring the EHR train to a screeching halt. We know how technology has transformed other industries. We know that paper records are archaic and put patients at risk while asking them to maintain endless patience when the same test has be performed a third time. And we know that electronically is the only way information can be shared in a timely manner.

So, while we may not know what the overall cost of corporate profits are to healthcare, we do know that they are really, really high. You’ve seen the figures associated with Epic contracts.

The truly important point is that the initial value of Epic and Cerner contracts isn’t even a reliable indicator of overall cost. According to a recent study by the consulting firm Katalus Advisors, hospitals that adopt Epic can expect to pay an additional 40-49 percent of initial contract value for “varying upgrade costs.” For Cerner, estimates were a slightly more reasonable 30-35 percent of contract value.

Based on these figures, Duke University Health System and Partners Healthcare can expect to pay an additional $350 million to Epic on top of the $700 million contracts they already signed. UC San Francisco will probably pay an additional $75 million for their Epic relationship.

Generally speaking, what they will get for that investment is not lower costs and greater efficiency. According to a report by the RAND Corporation that evaluates predictions made by a 2005 vendor-financed RAND study, expected cost savings and productivity benefits associated with EHR implementation have not materialized.

Why not?

In a nutshell: Sluggish adoption. Clinician intransigence. Poor planning and change management. Lack of interoperability.

Other than interoperability, these are organizational constraints, which are the constant in the EHR adoption equation. Which begs the question, why spend multi-millions of dollars—plus as much as 50 percent of contract value on top of that—for systems that are not interoperable and may threaten the financial viability of your hospital and organization?

The simple truth is that EHR systems do not currently offer cost savings equal to purchase price. With some solutions, there’s an uncrossable chasm between sticker price and ROI. And we’re talking about the financial viability of hospitals, here, not breakfast cereal. If those Lucky Charms disappear from the shelves, your kid may throw a tantrum, but nobody will get hurt.

Purchasing an EHR is not like a buying a car that you just get in and drive away. It’s like buying a car that you have to stop and recalibrate every mile with the assistance of the trained experts in the back seat who charge you a fee every time they have to listen to you speak or look under the hood. In this situation, paying less for the car is probably a good idea.

We have the most fractured and expensive healthcare system in the developed world, and the way we’re pursuing health IT adoption is making that worse, not better.  Hospitals and health systems must show some restraint and take control, forcing health IT vendors to behave in a way that at least adds as much value to American healthcare as it takes out in cash.

Edmund Billings, MD, is the chief medical officer for Medsphere Systems Corporation.

30 replies »

  1. Somewhere there is a quip in that story, I just don’t want to go there!

  2. Thanks for the explanation.

    So do the studies that DD references show that the effects of EMR adoption are more positive or more negative than has been assumed to date?

  3. “In statistical regression models the exogeneity of the ‘independent’ variables, or regressors, is assumed. But this may be false and problematic if a regressor is correlated with the error term. The ‘problem of endogeneity’ arises when the factors that are supposed to affect a particular outcome, depend themselves on that outcome. For example, the effect of campaign spending on the chances of electoral success cannot easily be estimated since the level of campaign spending depends itself on the perceived chances.”

  4. David Dranove,

    Can you translate from jargon to English?

    What is “endogeneity bias”?
    What is “heterogeneity of the treatment effect”?

    Also, what is the link to your study and that of McCullough et al?

  5. There are two current studies using state of the art econometric methods that shed substantial light on the effects of EMR adoption on costs and outcomes. These studies address the serious problem of endogeneity bias that plagues most research on EMR. They also address the heterogeneity of the treatment effect (adoption of EMR is the research “treatment”) that is overlooked by prior studies. My own study on costs can be found at NBER. A study on outcomes by McCullough, Parente, and Town (faculty at University of Minnesota and Wharton) is also making the rounds.

  6. Ha! Se non e vero, e ben trovato.

    Expect to see lots more of that when ICD-10 finally rolls out: coding will be much LESS accurate.

  7. Had a discussion with another physician about what the EHR “hath wrought”. We both agreed that the clinical information available to us when interpreting X-Rays (CTs, MRI, etc.) is worse than it was when we used to get it on paper.

    GIGO – garbage in, garbage out.

    Who have been the beneficiaries of the push for the EHR?

    Patients – no – or at least no evidence (for all those who love evidence)

    Doctors – no – unpaid data entry clerks

    Hospitals – no – extra expenses and difficulty

    Government – no. As has been commented on before, EHR don’t lower costs, but do make for easier “upcoding”. Data that can be “mined” is probably bad data.

    – Cerner, Epic and other vendors of EHR – YES!!!!!!

    I wonder how much in campaign contributions these corporations made?

    (For an interesting anecdote about data mining with an EHR, there is a post on SERMO – which may be apocryphal. Apparently a new EHR had been installed at a military clinic. Physicians had to code the reason for the visit for each encounter. The EHR had a long list of reasons – arranged alphabetically. Since it was a pain to go down the entire list for the correct reasons most of the physicians merely checked the first box – “artifical vagina”. After several months the commander reviewed the data and found out that 75% of all clinic visits were for “artificial vagina”. Now that is some valuable data !?!?!)

  8. Right. iPod? There’s spell check on iPod?

    I have an iPod.

    And an iPad, and iPhone.

    You are a real piece of work, “doctor.” Awesome.

  9. iPod spell check put it that way. Why don’t you go to their site and yuck it up with them? I guess that was the best inaccuracy I could offer, eh?

    Interesting you picked on that word too. Hit close to home?!

  10. Wonderful argument. Two wrongs make a right. Nancy Pelosi & Harry Reid have made that their boilerplate defense of indefensible actions these last 4 years.

    Hey, your king is now telling us the debt ceiling is not an issue to be debated.

    Do all you Democrat hacks think you can call your personal banks holding your loans and claim this as well?

    Wonderful leadershit, er, leadership these last 12 years. First we have a person who had the Supreme Court declare him President, then put our military in TWO unsustainable wars for 10 or more years, then after 6 years of one sided party rule, the idiots, er, electorate then turned over to the other side of the one party rule and allowed them to pass a domestic law that insidiously enslaves and will chip away at undesirables for years to come, and then had the Supreme Court declare it law as tax. Which the proponents said in passing it the law wasn’t. Like the war in Iraq was related to terrorism and to get rid of WMD.

    The more you examine Republocrat strategy, it is the same thing. Politicians who want power and control, who will kill anyone in their way.

    And you people vote for these antisocial Narcicistic cretins with joyful glee every 2-4 years. Who needs suicide and mass murder when, you have American voters! That is why there has been no further terrorist attack in the past 10 years. The Muslim radicals have you doing their work!

    Beautiful and ugly, ain’t it!?

  11. This HIT thing all started with an Executive Order from George W. Bush, who, in May, 2004, declared HIT to be safe and effective and established ONC with entrepreneur D. Brailer as the first head honcho. Speaking of tyranny and domination, and experimentation without consent, just look back at the HIT program.

  12. Those of you who have any interest in autonomy and independence should be paying very close attention to how Obama “solves” the gun control and debt ceiling limit issues in the coming weeks.

    Executive orders on issues of these magnitudes are not about branches of government, it is about the beginning of tyranny, and if left to partisan rationalizing and minimizing, watch what this Current Occupant of the White House will do if either of the other two branches of government try to interfere with his Frankenstein monster of PPACA.

    Orders of mandating EHR? Easy. Orders of controlling private insurance interventions? Already doing, but easier with executive order. Making choice outside of government mandates? Felony charges!

    You know, it is both humorous and frightening to watch supports scream at you how right they are and how wrong dissenters are, until, the alleged leadership these partisan idiots bow to then feed on their own.

    Oh, look who’s shouting loudest once the consequences play out, the outraged former partisans. Hypocrisy could have so many poster children, but I still enjoy seeing Obama’s face at the top of the stack.

    Hmm, but he’s also in line for the latest dictator -to -be as things play out.

  13. OK, I get your drift now. But if we are to really go back to the subject of my initial post, the physicians can make the consolidated data base work, my personal physicians are very laudatory of their software, and would not think about going back. Reluctant (maybe I should say obstinate) docs are dinosaurs of the worst kind. We have levers to bring the corporate entities into compliance (we can speak to corporations through bottom line impact) but without the cooperation of the physician it will be a long laborious task.

  14. No Roy, I don’t think we will have a single payer system. I wish we could have a single payer system, but we won’t unless this corporate welfare program crashes and burns.
    To go back to the subject of this post, why do you think it is that interoperability is not available just yet? It is very convenient to fault EHR vendors for it, but they don’t call the shots. Not even the successful ones.
    Portability of information has no ROI for health corporations, and quite the opposite is true. Portability of data across corporate boundaries is detrimental to customer retention and customer management.
    Yes, there will be inclusive access to something, but it won’t be access to your “Board Certified” physicians, or to what that used to mean, unless of course you happen to be a corporation or an equally well endowed individual.

  15. “It’s all about you isn’t it!?”

    Nah, it’s about the evidence.

  16. So you think we will have a single payer system? If so the financial benefits will be wonderful for the nation. But the administration has stated flatly that we are not capable of moving to that model. Much of the nation has long been confined to the providers dictated by the intermediary. Margalit, I have a feeling that you are a supporter of the slippery slope philosophy. Change is a vital part of life, Health Care Reform will be touted as one of the most beneficial social changes in the years to come. I urge people to open their minds to the new reality of inclusive access with a payer involved.

  17. Unfortunately, Roy (may I call you Roy?), pretty soon you will use the physician your insurer tells you to use, assuming they let you “use” a physician at all…. and don’t worry, it will have an electronic file….

  18. Mr. Williams, you couldn’t be more right. No one is connecting the dots. The EMRs and HIEs aren’t only inadequate. There are staggering costs that are being ignored: patient pain and suffering, deaths, lost productivity and enormous financial costs. Yet, we casually accept that docs won’t be able to exchange records until 2015 — if then! You may want to read more on this point at http://medkaz.com/wheres-the-urgency-the-anger-the-outrage/

  19. It’s all about you isn’t it!? I stand by my comments.

    Change is difficult and physicians that are over the age of 40 are the most difficult to bring into the modern era.

  20. “We all have know that physician resistance to change would be a problem…it always is.”

    If you’re talking about change that is expensive, creates many more problems than it solves, and makes it more difficult to do our job, you’re 100% correct.

    Otherwise, you’re just spouting bullshit.

  21. I found it interesting that Cerner and GE funded the 2005 RAND study that presented grossly exaggerated estimates of savings to the Congress. When the CBO stated they were excessive, Congress voted for HITECH amy way. Sad to say, this HIT project is headed for doom as in the UK.

  22. http://www.nytimes.com/2013/01/11/business/electronic-records-systems-have-not-reduced-health-costs-report-says.html?pagewanted=1&_r=0&hp


    Media coverage above.

    The experiment is failing. Costs are extensive and savings and improved outcomes are non-existent.

    Thus far, no one has factored in the unexpected deaths, EHR crashes when all records of all patients vanish at once, near misses, and other adverse events.

    No one is counting the patient carnage in this experiment, and it is not trivial.

  23. An interactive patient data base is critical to the future of health care. It is the fundamental building block of future outcome research. That of course does not even mention the convenience to patients that move through the care continuum.

    We all have know that physician resistance to change would be a problem…it always is. I now am at the point that I will not use a physician who is not Board Certified, will not allow me to call them by their first name, and does not have an electronic patient file.