Some people may tell you that health care IT will solve many of the quality and cost problems in health care.

I don’t believe them.

I know a 70-year old man named Carlos (not his real name) who was hospitalized following a bout of hydrocephalus.  Hydrocephalus is a build-up of fluid in the skull, which affects the brain.  Among other things, people with hydrocephalus can be confused, irritable, and nauseous.  Carlos had all of these symptoms.

Carlos’ problem was fixable by inserting a special kind of drain in his head called a “shunt.”  This kind of shunt is, essentially, a series of catheters that runs from the brain into the abdomen, and which drain the excess fluid.  You can’t see it from the outside, so it’s meant to stay inside of you for a very long time.

For a week after Carlos’ shunt was installed, his symptoms completely disappeared.  But they soon started to re-emerge.  Worried, his family took him to the hospital.  Doctors found that his hydrocephalus was back – the shunt wasn’t draining properly.  They admitted him to the hospital, and the next day they put in a new shunt.  The surgery went well.

But again, about a day later, he started to have the same kinds of symptoms.  The doctors sent him for a CT scan, which showed, to their surprise, no problems with the shunt.  Unsure of what to do, they decided to wait and see if the symptoms resolved.  It was possible, they thought, that the symptoms were from the quick drainage of fluid through the shunt.

A doctor friend (call him Lou, although that’s not his name) happened to be visiting while this was going on.  Lou knew that Carlos had been on certain medications for a number of years.

“Has he been getting those medications in the days he’s been in the hospital?”  he asked the attending physician.

“I don’t know,” said the doctor, who went to check the hospital’s electronic medical record system.  “The meds are listed here, but it doesn’t look like he has been getting them.”

It wasn’t entirely shocking, since Carlos had been in and out of surgery.  Still, it was curious.  “What are you getting at?” asked the doctor.

Lou pulled out his iPhone, and looked up the symptoms of withdrawal from the medications Carlos was taking.  He showed them to the attending doctor:  confusion, irritability, nausea.  Very much what they were seeing and trying to explain as having to do with the shunt.

The attending spoke with the other physicians treating Carlos, and decided to put Carlos back on the medications.  His symptoms went away a short time after that.

One doctor I talked to about this story told me the doctors did the right thing.  Carlos had a history of hydrocephalus and problems with his shunt.  It made sense to rule out issues with the shunt before looking for other answers.

But another doctor I spoke with didn’t agree.  Given that the shunt had just been replaced, it seemed premature to try to explain everything by looking for another defective shunt.  Instead, this doctor told me, someone should have taken a look at his history before sending him for another CT scan.  As this doctor put it, “a patient is not a shunt.”

Carlos’ story tells us something about health care information technology that we may not want to hear.

The IT systems in the hospital were mainly used to order the CT scan, and to keep track of…medications that weren’t being given.  The single most important piece of equipment in Carlos’ case was the brain of the doctor treating him. .

Being a good doctor means asking good questions.  Until a computer can do more than come up with good Jeopardy questions, we must realize that we are still reliant on human judgment, and human fallibility.

And we will be for a very long time.

Evan Falchuk is President and Chief Strategy Officer of Best Doctors, Inc. Prior to joining Best Doctors, Inc., in 1999, he was an attorney at the Washington, DC, office of Fried, Frank, Harris, Shriver and Jacobson, where he worked on SEC enforcement cases.

Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedIn

24 Responses for “A Patient is Not a Shunt”

  1. You have written many good articles. Keep it up, because I often come here to read them ;)

  2. Jared Houck says:

    I don’t think anyone is saying that Heath IT will or should replace a physician’s ability to trouleshoot and diagnose. From reading your article, the IT side performed exactly as it should – providing timely info including: patient history, EHR documentation/MAR, and home med list. Not performing med reconciliation at the time of admission or instantly jumping to surgical conclusions are purely human error. Health IT is meant to support, not supplant common sense or good judgement.

  3. Hydrocephalus is a severe handicap. I am teaching handicapped children with hydrocephalus. They are sometimes quite lazy and not very motivated. They forget things they had learned and it is not easy to work with. The shunt just solves the physical problems. The psychological and educational problems are unfortunately not positively affected by a shunt.

  4. “Instead, this doctor told me, someone should have taken a look at his history before sending him for another CT scan. As this doctor put it, “a patient is not a shunt.””
    So? If all hospital documentation took place on paper, would the story been any different?
    I know it’s fashionable to blame HIT for everything that goes wrong wherever HIT is used to any extent, but in this case it is all user negligence.
    Perhaps you should pay attention to the HIT that Dr. Lou used to solve the puzzle. Quite lovely, isn’t it?

  5. MD as HELL says:

    It was no different. That is the point

  6. However, if you put together what the iPhone was able to provide with what the EHR already has, it’s a short way to see how improvement is evolutionary, perhaps even inevitable. Not so with paper.

  7. Tom Leith says:

    @MD as HELL: It was no different. That is the point
    Oh c’mon MD. What was it that showed Carlos had been prescribed the meds, and what was it that showed he wasn’t getting them? A paper chart? No. Where would the doc have looked in a paper system? Most likely he wouldn’t have looked — it would have taken too long. He’d go ask a nurse who wouldn’t know any better than he did and she’d spend howsoever long finding the answer, then tracking down the attending to deliver it, who’d turn around and grump some grump at her and give a verbal order she’d document in a paper chart and then maybe it’d get communicated properly to the pharmacy…
    What was it that showed the withdrawal symptoms? A book? No, an immediately-available authority evidently trusted by both docs.
    I think the point is that it WAS different: the two docs could confirm Dr. Lou’s suspicion right away. They could probably correct the problem right away too.
    If you want to complain that the system didn’t flag a mismatch between “patient’s current meds” and “in-hospital Rx orders” and “in-hospital dispense events” making up for the frailties of the attending (and the nursing staff) you’re spot-on. Minimally-implemented tech can’t make up for human inattention. We all know that. When docs start demanding to have systems point out their frailties we’ll start getting them. This is a Guild issue, not a technical issue.
    It is entirely true as you point out over and over again that data is not information is not knowledge and knowledge isn’t the cause of action, but can you see no value whatever in having (to borrow a phrase) “information at your fingertips”?
    From my point of view it was different but not different enough. Fine. I don’t understand for a minute how it is a drug succeeds when it is better than a placebo but HIT fails when it doesn’t achieve a standard of perfection never, ever, anywhere reached in a paper charting/ordering/billing system, much less being better than nothing.

  8. HITechxpence says:

    Margalit is highly conflicted and ought to state exactly what her conflicts are as a promoter and defender of HIT.
    Was the redo shunt necessary? Was the patient not getting the correct medications prior to the second operation? Was the patient sent home from the hospital on the correct medications the first time?
    What you describe with excellent clarity is a typical error that paradoxically, is promoted by complex HIT systems. It is why the Health Care Renewal blogger, Dr. Scot Silverstein refers to these devices as errorgenic (error promoting).
    It is why these devices must undergo safety and efficacy vetting by the FDA.
    Evan, have you reported this to MedWatch of the FDA?

  9. imdoc says:

    To me this story is more about payment dysfunction. ” The single most important piece of equipment in Carlos’ case was the brain of the doctor treating him…”
    The financial reward however is to do procedures. I wonder if non-surgeons were consulted to evaluate the confusion.

  10. Nancy Sweeny, MD says:

    Clearly, this case, and thousands like it,exemplify the failed panacea of HIT, EHR, and CPOE. In this case, the patient was the computer and the shunt. The devices are so cumbersome, error promoting, and cognitively disrupting that the toxicity you report here is widespread. We do not hear about it because of several reasons:
    1. Do not disclose contractual agreements
    2. Hold harmless clauses
    3. The policy of blame the user as the learned intermediary
    4. Medical staff “code of conduct” policies that give the hospital administration and its paid HIT champions carte blanche to intimidate and punish any doctor at any time who says something the admin does not like to hear.
    5. Disruptive administrators who are in the proverbial financial bed with the vendor.
    6. Administrators who went to HIT charm school and recite the refrain when the computers fail and all patient care records vanish in one fell swoop: “we are not aware of any patient safety issues” or “patient safety was not affeted”.
    Take a look at the mayhem wrought by EHR and CPOE at Dallas’ Parkland Hospital as described in a series of reports in the Dallas News.
    No onewants to know the start reality of what happens when CPOE systems are deployed to run patient care without having any assessment for safety and efficacy.

  11. Evan Falchuk says:

    imdoc -
    I don’t think anyone in Carlos’ case was motivated by the financial rewards of the CT scan, even though it was an avoidable expense.
    Many doctors complain – Atul Gawande may be the most famous – of how undervalued doctors’ judgment is, however. This may be part of the explanation of what we saw happen here.
    More here:
    Evan Falchuk

  12. MD as HELL says:

    Garbage in, garbage out, Mr. Leith. You go ask to see the bottles of current and past medicines. You call the pharmacy(s) the patient uses.
    Do you have the slightest clue how many inaccuracies exist in all medical records? It makes malpractice attorneys lick their chops.
    It matters not whether they are electronic or paper. Records are nothing to stake one’s life on.

  13. anon says:

    Good lord, this is a stupid analysis.
    HIT makes it easier to deliver good care, much more efficiently. But DUH it does not guarantee it. Facebook makes it easier to stay in touch with friends, but it does not make you popular if basically you are unlikable. DUH.
    If you’ve got a hospital whose underlying standard of care is so appalling that they fail to maintain a patient’s medications when he is admitted, it’s unsurprising that HIT alone cannot fix the hospital. But as pointed out above, HIT did make it easier for the family to find and correct the hospital’s error.
    HIT should have been a no-brainer in healthcare. This is an information and communication industry! There should already be a figure like Mike Bloomberg, a guy who got rich by bringing IT to the practice of medicine in a user-friendly, mission-critical way. We should be into the 4th or 5th generation of HIT. But the delivery of healthcare is so radically dysfunctional and untethered from the pursuit of quality that HIT has been left completely underdeveloped and ignored. It’s absolutely amazing. It’s like failing to apply statistics or the development of germ theory.

  14. Evan Falchuk says:

    Hi anon,
    Stupid analysis – unfortunately, no. This happened precisely as described.
    You may be very surprised that the hospital at which this happened is one of the best in the country – and it’s one of the most embracing of health care IT.
    It was probably an error to fail to give Carlos the medications. But if the medication had been, for example, Lipitor, we wouldn’t be having this conversation.
    No, the real problem was the doctors failing to look at Carlos as a whole patient.
    Good medicine requires judgment and insight, neither of which were on particularly good display here, regardless of whether there was a paper file or an electronic one.

  15. Robert says:

    “Stupid analysis – unfortunately, no. This happened precisely as described.”
    Umm . . . does someone want to explain to Evan how an “analysis” differs from an “anecdote”?
    I agree with the commenters who have pointed out that “IT versus human judgment” is a false dichotomy. IT in healthcare is about taking smart, caring people and getting the best out of them. If anyone is claiming IT will allow dumb, careless people to practice good medicine, I’d love to know who they are.

  16. rbar says:

    anon summarized it quite well, but then goes into overstatement overdrive:
    “HIT should have been a no-brainer in healthcare. This is an information and communication industry! There should already be a figure like Mike Bloomberg, a guy who got rich by bringing IT to the practice of medicine in a user-friendly, mission-critical way. We should be into the 4th or 5th generation of HIT. But the delivery of healthcare is so radically dysfunctional and untethered from the pursuit of quality that HIT has been left completely underdeveloped and ignored. It’s absolutely amazing. It’s like failing to apply statistics or the development of germ theory.”
    The fact that EHR are not in their 4th or 5th generation indicates that they are just not that great. They just facilitate the beureaucratic and communication part of healthcare. And one has to consider effort and benefit – how much bang is there for the buck?
    Don’t get me wrong, I am all for it, and I have worked with 3 different EHR since starting residency a decade ago.
    But: the best information does not help when the physician(s) in charge do not review it. Or when the crucial information is from a different hospital with their own EHR and the info is not requested.
    Re that particular case, I have seen many shunt patients with problems. If the neurosurgeons feel it could be shunt dysfunction, the other docs usually follow. A good neurosurgeon would have recognized since him/herself and/or asked for a neurology consult.

  17. Gary O. says:

    The author uses fallacious reasoning to reach an absurd conclusion citing one interesting anecdote to support his thesis stated in the first two sentences that it is false that “IT will solve many of the quality and cost problems in health care.” First, he is attempting to convince us that IT should be rejected because mistakes will still be made even after IT is implemented (the “perfect solution fallacy“); he uses a case guaranteed to arouse the emotions (“misleading vividness fallacy“). Of course, the hallmark of these two related fallacies of argument is the omission of any discussion of how much IT will not work to improve quality and cost. Perhaps the author assumes his audience is quite comfortable solving complex problems by being given only two possible solutions, IT or no IT (the fallacy of “false dichotomy“). Secondly, and perhaps more importantly, his initial premise regarding IT is incorrect. Those who have adopted electronic medical records in a purposeful way have taken great pains to explain that “[s]iloed electronic care information is almost as flawed as siloed paper information…. EMRs need to be a part of a total care package, not a stand alone tool.” (Halvorson, G., Health Care Will Not Reform Itself, p. 58) Thus, Mr. Falchuk wrongly implies that knowledgeable adopters have claimed that IT alone would improve quality and cost. By focusing on one tool used to connect caregivers in an integrated, coordinated, patient-focused approach, the author commits the logical “fallacy of division.”

    • Satin says:

      Re: “citing one interesting anecdote to support his thesis stated in the first two sentences that it is false that IT will solve many of the quality and cost problems in health care.”

      Perhaps you’d believe several Wharton professors better?

      “Information Technology: Not a Cure for the High Cost of Health Care.”

      Or perhaps Abraham Verghese, Professor and Senior Associate Chair for the Theory and Practice of Medicine at Stanford, who wrote:

      … I have similar problems with the way President Obama hopes to pay for the huge and costly health reform package he has in mind that will cover all Americans; he is counting on the “savings” that will come as a result of investing in preventive care and investing in the electronic medical record among other things. It’s a dangerous and probably an incorrect projection.

  18. Evan Falchuk says:

    It’s a hard topic, because it requires thinking about health care differently than most people are used to doing.
    There is a growing body of evidence of the impairment of medical decision-making.
    I wrote about one recent study here: , but there are many others.
    It has something to do with the conditions under which doctors work.
    Doctors don’t have the time to think and reflect about their patients. They work in an environment in which spending the kind of time they need is not valued as highly as it should be. Patients suffer needlessly as a result.

  19. Nice Stuff…nice Blog.

  20. Christy N. says:

    Excellent post. I appreciate it greatly. I don’t usually comment on other peoples post, but I had to thank you for this. Thanks again! Blog On!

  21. だこったのか?のほこりをなだめるためにしくなでるバック、なぜ、いくつかはっている。 Moのクリームはのびは、ビーズのれたのようながオフフックrustledとして、うこ�
    クロエ バッグ

Leave a Reply


Founder & Publisher

Executive Editor

Editor, Business of Healthcare

Contributing Editor

Contributing Editor

Business Development

Editor-At-Large, Wellness

Editor-At-Large, Europe



The Health Care Blog (THCB) is based in San Francisco. We were founded in 2003 by Matthew Holt. John Irvine joined a year later and now runs the site.


Interview Requests + Bookings. We like to talk. E-mail us.

Yes. We're looking for bloggers. Send us your posts.

Breaking health care story? Drop us an e-mail.


We frequently accept crossposts from smaller blogs and major U.S. and International publications. You'll need syndication rights. Email a link to your submission.


Op-eds. Crossposts. Columns. Great ideas for improving the health care system. Pitches for healthcare-focused startups and business.Write ups of original research. Reviews of new healthcare products and startups. Data-driven analysis of health care trends. Policy proposals. E-mail us a copy of your piece in the body of your email or as a Google Doc. No phone calls please!


Healthcare focused e-books and videos for distribution via THCB and other channels like Amazon and Smashwords. Want to get involved? Send us a note telling us what you have in mind. Proposals should be no more than one page in length.

If you've healthcare professional or consumer and have had a recent experience with the U.S. health care system, either for good or bad, that you want the world to know about, tell us about it. Have a good health care story you think we should know about? Send story ideas and tips to

REPRINTS Questions on reprints, permissions and syndication to



Affordable Care Act
Business of Health Care
National health policy
Life on the front lines
Practice management
Hospital managment
Health plans
Specialty practice
Emergency Medicine
Quality, Costs
Medical education
Med School
Public Health

Electronic medical records
Accountable care organizations
Meaningful use
Online Communities
Open Source
Social media
Tips and Tricks


Health 2.0
Log in - Powered by WordPress.