Errors of Omission

I am a rural Family Physician who has been in solo practice for more than 22 years.  I am neither a technophobe nor an information technology Luddite.  I have been using electronic prescribing for over 6 years and am in the market for an EHR that is net-based, scalable, interoperable and linked to a nationwide patient database.

While I wait, over the years I am seeing more and more patient care that is less co-ordinated and even thwarted by the very health information technology (HIT) that is supposed to increase efficiency. In my opinion, this is leading to decreased information transfer that is wasting precious time and putting patients at risk with errors of omission.

I will give anecdotal and real examples of HIT run amok that I suspect are more common than generally appreciated.  Alarmed by the lack of awareness of the potential frequency of these errors, I am writing this hoping that the blogosphere can somehow counter the momentum of an all-powerful HIT cerebrosphere.

1.     e-Prescribing (ERX): While mandated alerts about potential drug interactions in this software is often life-saving, it can also be life impairing.  There are two reasons for this:  1) at point of care, the warnings are just too darn sensitive and I’m being conditioned to ignore 90 percent of them.  I am afraid that this will cause me to click the “ignore” pop-up at the wrong time.  For instance, doxycycline and Dilantin have an interaction and a prescription of one in the presence in the other always prompts a warning.  When I researched this, I found the plasma concentration of doxycycline is decreased by a clinically negligible amount.   2) at the pharmacy window, the warnings can override physician judgment. A colleague of mine described to me how he prescribed a fluoroquinolone antibiotic to a patient on Coumadin and, aware that there was an interaction,  ordered the appropriate follow up testing and dosage modifications.  The pharmacist not only refused to fill the prescription,  they also did not notify him.  Instead, he asked the patient to call the doctor.  Two days later the patient was admitted to an ICU with life-threatening sepsis.  In both cases, needed prescriptions were omitted.

2.     Electronic Health Records (EHR):  In the old days, when I received a consultant’s letter, (unless it was an internist with OCD who was so thorough, he put a 3rd year medical student to shame) I could always count on an easy-to-read last paragraph that clearly stated the impression, plan, and suggestions.  What’s more, if there was something really important, the consultant often gave me a call.  Now I am typically faxed a 15 page electronic note in various and sundry fonts packed with 2 years of medical detritus from previous notes and labs.  I often cannot find an impression or plan.  This is a time consuming and opaque breeding ground for clinical errors of omission.

3.     Centralized Scheduling:  Instead of being able to directly call a consultant’s office for an urgent patient referral, I am now obliged to use centralized computerized scheduling that is basically transforming them into big box specialty centers.  It seems that most of the smaller independent cardiologists, gastroenterologists, orthopedists, vascular surgeons and other specialists have now banded together in large group practices protected by a phalanx of  centralized schedulers, often at an off-site call center.  Now instead of “send the patient right down and we’ll squeeze them in”, it becomes “fax the pertinent records to our intake coordinator and we will call the patient back in 48 hours to schedule an appropriate appointment”.  Aside from the fact, that I now have to call the patient in 48 hours to see if they got an appointment, I also often have to send the patient with suspected appendicitis or acute renal failure to an emergency room .  Last but not least, these systems cannot handle patient-specific nuances.  For example, a patient who may need a carotid stent may not be scheduled with the specialist expert in that procedure.  In other words, poorly functioning scheduling systems can lead to important lapses in connecting the right patient to the right doctor at the right time.
4.     Radiology Precertifications:  Last week I saw a patient with severe flank pain and bloody urine.  Thinking that a kidney stone was the most likely diagnosis, I wanted to order a CT scan to confirm an obvious diagnosis.  After going through a web-based electronic precertification process, I was informed that an answer to my request would be made available in 24 hours.  Of course the hospital refused to schedule the test without an approval number.  My office notified me of the situation and, after additional time consuming calls, I was able to force an expedited review by a live on-call person.  In this instance, a precertification system led to an unnecessary delay in an important test.

5.     Offsite Radiology readings:  Last week I ordered a mammogram and a CT of the pelvis on two different patients .  The mammogram was performed at the local hospital, the radiologist took the time to examine the patient and he called me with the results. That patient in question is now scheduled for a biopsy.  On the same day I sent a woman in for a CT of a very enlarged inguinal lymph node.  This was done at the same hospital but was read by a radiology group across the state.  I still have not received the report but I will be going to the office today  (Sunday) to check the fax machine.  This is an example of a delay in getting the results of an important test.

6.     Offsite cardiology studies:  Whenever I can, I try to send my stress tests to the local cardiologist because, if there is an abnormality, I am called and if necessary they will even (gasp) see them for an urgent formal consult after the exam. In contrast, the large institutions often act as technicians only.  It has been more than once that I see a patient 2 weeks after a stress test, call medical records to get a report, and find an abnormal result.  Luckily, there have been no fatalities to date, technically no error was made, but I also consider this an error of omission.

In effect, what I am seeing is not an increase of medical errors.  In fact every prescription, test and consultation is of the same if not better quality.  However, the workflow and communication supporting them seem are being hindered by the HIT.

This may seem counterintuitive to the administrative, policy, informatics, politicians and the academic types but it is my practicing opinion, forged in the real world.

This post originally appeared at Disease Management Care Blog.

17 replies »

  1. I really love your blog.. Very nice colors & theme. Did you build this amazing site yourself? Please reply back as I’m wanting to create my very own blog and would love to learn where you got this from or just what the theme is named. Thank you!

  2. Good post. I learn something new and challenging on websites I stumbleupon on a daily basis. It will always be helpful to read content from other writers and use something from their websites.

  3. Love E-Fax just wish the pricing would stabalize, we have been through 3-4 companies

  4. Thank you. I’ve heard some of these things and a thousand other things many times. I’m sure you know that many of your issues can be fixed rather easily, yet they are not, and the more “regulations” we have, the less real problems that will be addressed.

    Fore example, the LFTs presence (or not) in a long list can be easily fixed with a large hover that shows the contents before you click. It is so simple it’s infuriating. Won’t be exactly as flipping pages, but darn close and the technology exists and is applied in other software products.
    When a doctor is about to add “his” HTN, the current problem list should provide autocomplete from existing Dx, and same for selecting codes. Same type of solution with small variations will fix the lisinopril issue.
    The changing of templates should not be needed unless you use documentation by exception, or pre-filled templates, which is playing with fire. If CMS wouldn’t require these crazy things for payment, I doubt anybody would do that.
    And so forth….

    The bottom line is that there isn’t enough input from a broad enough spectrum of users that goes into EMR design. Now that the government decides what is meaningful, there is even less consideration for users.

    Paper is a malleable and very forgiving type of media. Computers do have their strengths, but with today’s technology, paper is superior for ad-hoc, hectic and erratic note taking. We will have to wait for the day when you can talk to the computer and it will understand and abstract what you say into an appropriate computable structure, and the even more distant day when it will understand your verbal questions and instructions (show me the last lipid panel with LFTs). Multi-touch will be obsolete technology.

    Until that future materializes, it will remain a tradeoff between the well appreciated immediate capabilities of paper and the not yet realized long term abilities of computers. I know it seems like a lot, but there really isn’t much that needs to be done to reach an even point, and then tilt the balance in favor of computers.

  5. Since you asked….
    They are slower for one. It is so incredibly fast to flip open the outpatient chart to find information – the EHR is just as prone to human error and reports have a tendency to be mislabeled or the title is incomplete and you don’t know if the lab report titled “lipid panel” includes a set of LFTs or not and you have to click to find out – oops no it doesn’t, lets try the next one…. but turning pages is so very fast. And when you have had the EHR for awhile the lists get really really long. Hard to “thin” an EHR.
    Keeping a medication list or problem list was so much easier with a paper chart. Pencil and eraser work really well. The number of clicks required to stop one medication and replace it with another is unreal. I could write out the prescription twice and update the pencil list in the same amount of time. And duplicate entries are such a pain. One provider will enter “hypertension, 401.9” and other “hypertension, 401.1” and now they are both on the list. Or someone will bump the lisinopril up to 20mg but forget to put a stop on the 10mg dose, and now both are there for someone to commit a medical error by choosing the wrong dose.
    It is so easy to forget to change the parts of the template that don’t match that visit. When you’re in a hurry there are so many ways to make a mistake and leave in inappropriate parts of the template there for all to read and wonder at. I know I am not the only one who occasionally makes mistakes because I see the notes from others that make you scratch your head, but then you tell yourself, “Oh, they have an EHR too.”
    The technology required to maintain a paper chart will still exist in 5years or even 100 years. Will my current EHR be around in 5 years? How many times am I going to have to port over my data?
    Using a mouse is so yesterday. I want my multi-touch ready software NOW! But it has to be written by someone who understands what it is like to try to multitask in the exam room. I have yet to see such a product.

  6. I am. What do you see as the main downsides? And what do you think will improve and/or eliminate those downsides?

    BTW, instead of scanning, you could get an e-Fax number and all the faxes will be sitting in the computer. Not sure if PF has this built in or not, but even an outside e-Fax program is better than scanning. Your scanning person could just move files into charts. Of course, any paper that comes in through means other than fax, will need to be scanned, but it shouldn’t be that much.

  7. Actually, I don’t pay anything directly. I do have to maintain my computers but fortunately I haven’t met a computer yet that I couldn’t fix and I actually enjoy fixing computers. With Practice fusion I have no server which simplifies things greatly.
    As to value – I do like having access to the charts wherever I go. I do like seeing the trends in vitals over time. I like the built in coding and the integration with scheduling and billing. We are going to be getting integration with lab and I believe that will be more efficient. I like not having a chart room. The medical records clerk has been replaced by a scanning clerk so no savings there. Hopefully some day all the local clinics will have products that talk to each other, but that appears to be a long way off.
    There are definite downsides to an EHR but no one ever seems interested in hearing about those….

  8. “But you have to use, it will make you better.”

    Make that:

    “But you have to pay for it and use it . . .”

    You say:

    ” I do see value in some areas”

    What are those areas?

  9. You have several valid points here. Technology should not get in the way of person-to-person communication. No matter how efficient or accurate technology can make your data, it cannot stress urgency or explain things in the same way as another human. Personally, I believe that the EHR meaningful use incentives might be a little bit premature. There are several great EHRs that I’ve researched, but there are just as many poorly designed systems being used.

  10. Medicine is a complex decision making process involving at least two (maybe more) humans. Communications is another complex human interaction. All organizations have varying degrees of complexity aka bureaucracy because one person can not know and do everything. In health care these interaction and processes are in transition and transformation … we will get through this transition, we’ve learned to use cell phones, email, and ATMs even electronic voting.
    What I do appreciate is Dr. Sidorov’s thoughtful review and assessment of the current state of the art. I am worried that I never did learn how to program my VCR and I don’t get Tivo either buy I’ve mastered other technology like the ATM. I also can’t program my digital watch, thank goodness the ecowatches don’t require programing.

  11. I was on a demo webinar the other day for the the new release of Praxis EMR version 5, their ONC certified release. We only have one REC client using it. It’s intriguing. I have to say that the user interface presents by far the lightest “cognitive load” of any of the platforms I have to work with, and navigating around seemed to be a snap. They tout a different approach — no templates.

    Dunno. They rank high in the user surveys, but…

  12. “Look, here’s this great new technology”
    “I don’t want it.”
    “But you have to use, it will make you better.”
    “Oh, all right. But look, it caused these problems.”
    “That’s because you’re incompetent”

    The point isn’t that the technology is bad (“guns don’t kill people, gun-owners do”) it is that it has been mandated for reasons that remain largely unproven and without consideration of its potential for harm. You would have to be very ignorant indeed to argue that HIT has not “over-promised and under-delivered.” Give me a reason for each and every piece of HIT other than some mandate. And that reason had better be grounded in reality. I have used Practice Partner, NextGen, SOAPWare, Meditech, and now Practice Fusion and I can most assuredly tell you that they do not deliver even half the benefits promised. And yet I still will continue to use them because I do see value in some areas (and mainly because I don’t have the space for a chart room) but don’t try and sell me the BS that they make me more efficient or that they improve patient care. They are simply a tool – a very imperfect tool. If someone doesn’t want to see their precous HIT criticized, then stop making promises about it that simply are not true. Any product can become associated with some poster child of success (“ACME Clinic saved thousands using our product!”) but that does not make those claims applicable to others. The strongest champions of HIT have unfortunately been those who aren’t stuck using it.
    Many predict a round of consolidation in the HIT industry, and I hope it is brutal. Maybe when the dust settles we will start to see products that allow us to use our touch screens effectively and share data in some common format. And fix HL7 already – it is a travesty that after all this time getting one product to talk to another still requires a “custom solution”

  13. So many points to discuss, but here are a few:

    1. This post was not written by Dr. Sidorov

    2. We tossed out e-prescribing the day I had to override an “interaction” between Prozac and Proctofoam. All studies show no reduction in errors with e-prescribing.

    3. Some posters here love the analogy of airlines with their checklists. What we’ve created is a system in which the pilot has to go back in the cabin during take-off to find who ordered the fruit plate!

    If we want to reduce errors, set up systems that reduce physician distraction. It seems that IT as used in medicine is more successful at increasing distraction than at any other function.

  14. This is a great post and describes reality very well. However, if you read carefully, it seems to me that at least half of the issues described here are due to things other than technology per se, such as consolidation into big practices and the addition of various bureaucracies which are presumed to provide those “economies of scale”, but instead provide anything but a so called “:patient-centered” experience.
    Unfortunately EHR and HIT in general are now mentally associated with this deterioration in actual patient care.

    That said, there are very legitimate complaints regarding EHRs, such as those long and practically useless notes, as well as the overly sensitive alerts (BTW, the user should have the ability to change severity and get rid of most alerts that are not useful).

  15. This trend is really there, of course, but to attribute all these data points to one monolithic thing called “HIT”, “EHR”, is to further confuse. As physicians you know very well, multiple symptoms might be from one pathological process, or more than one. There is no lesion called “HIT”.

    Your enemy is not technology, but a random combination of the trend toward bureaucracy (which is increasing) and stupidity (which is old.) About stupidity, nothing; but about bureaucracy: it is an inexorable feature of size. Repeat: inexorable. (Max Weber conclusively demonstrated this more than century ago; the central planners have not yet gotten that memo.)

    A key feature of bureaucracy is the proliferation of mindless process and the subsuming of situational judgement under guidelines. (Weber demonstrated that this organizational trend is not just a possible feature, but has all the force of natural law. )

    It is simply true that we use computers to attempt to manage the complexities bureaucracies exist to emit. And they fail.

    The eprescribing software is not capable of handling the complexity, as you say. (I’m not sure it ever will be.) But the government has decided you need to use it, and that it will help you and your patients. This is not a computer problem; this is a bureaucracy problem.

    The pharmacist was stupid. He/she would have been negligent without a computer. He’s had a phone for 100 years.

    The 15 page fax is simply laziness on the part of the sending office, and apathy on the part of the physician. Any good EHR will produce that same single paragraph summary you used to enjoy. It’s become fashionable to blame the computer for what the people told it to do. If the physician whose name is on the note said “I want a single paragraph summary to go to all referring docs, and I want it to look like this…” it would get done. It might be that the office has gotten so big he is separated from the senders by many layers. He would be separated without a computer.

    The existence of treatment algorithms in the payers’ pre-cert departments is simply their effort to get rid of “unnecessary care”. I don’t know if you’ve noticed, but your profession has been indicted by the Dartmouth Atlas, and that is one convenient indictment. The physician community has responded by agreeing with the wonks that… those OTHER specialties need guidelines.

    The payer is not getting you a quick answer because they don’t have to; they do not get sued for malpractice, you do. (They behaved this way before computers.)

    Poorly functioning central scheduling? Yes, but this is a feature of size, not technology. The schedulers are many layers removed from the doctors whose daily lives they are controlling. To mitigate these potential disasters, the schedulers are given highly complex protocols, and like the e-prescribing software, the computers aren’t good enough. (I manage physicians. We are used to such rules as “Dr. Jones will see one Worker’s comp new patient on Wednesday afternoons at 2:04, if it is a carpal tunnel syndrome with EMG attached.”

    The scheduling would be slower without the computer.

    Large medical practices? The government has decided that large organizations will deliver better care. (This contradicts the accumulated wisdom of millenia and all human experience, but… there are studies.) As you know, medical practices are integrating in order to be able to negotiate with payers and interact with the government. To the government, “HIT” means “control and command.” To the providers, “HIT” means “cope with the rules”. The technology is in the middle, and a red herring in what you are seeing happen to medicine.

  16. Absolutely correct – each and every one of the authors examples is happening hundreds if not thousands of times a day all across this country. But for the most part they seem to be the type of problems that lay beyond the understanding of those creating our health care future.
    I think the push for the EHR is the most obvious example of the backwards thinking that leads to these problems. We were told that if physicians and health care providers would only embrace this technology that we would see better communication and co-ordination of care. This has not happened, and for reasons that are obvious to anyone who is thinking. We should have urged the formation of communication standards (both by clarifying the laws surrounding the sharing of information and implementation of a common digital format), and encouraged the players to utililize these standards, and adoption of EHR would have followed.