EHR Design: It’s a Matter of Time

Screen Shot 2014-09-22 at 3.00.23 PM

As reported last year at HIMSS and by many online news and opinion sources since, physician dissatisfaction with EHRs is growing. Indeed, while this blog post doesn’t focus on the broader picture, general physician career dissatisfaction is disconcertingly high.

The breakneck push for more and better EHR use as a component of regular medical care is a significant part of that malaise, but it is insufficient as an explanation. For the most part, doctors really don’t like what the health IT industry is giving them to work with. The HIMSS survey proves it, showing that around 40 percent of physicians would not recommend their EHR to a colleague.

One would expect an industry to develop better products and improve usability, acceptance and satisfaction over time. In health IT, the opposite has occurred, with most pointing fingers at Meaningful Use as the culprit for awkward workflows and Rube Goldberg solutions cobbled together so everyone can get paid in a timely manner.

It seems EHRs are taking more time to use rather than less, which was the original goal.

According to a survey published this week in JAMA online, this is exactly the case. In an article called, “Use of Internist’s Free Time by Ambulatory Care Electronic Medical Record Systems,” a team led by Clement J. McDonald, MD, looked at the amount of time EHRs carved out of what physicians consider their free time.

Here is a summary of their findings:

  • 411 physician respondents
  • 61 different EHRs used
  • 9 EHRs used by 20 or more physicians
  • 59.4 percent said they lost time after moving to an EHR from paper
  • 63.9 percent said note writing took longer with an EHR
  • 33.9 percent said it took longer to review EHR charts than paper
  • 32.2 percent were slower to read other clinicians’ notes

So, how much time did physicians lose to EHRs?

On average, they lost 78 minutes a day or 6.5 hours per week. If a clinic workday is 8 hours, these physicians were spending almost a full extra day each week on EHR-oriented stuff. Because responsibilities and lurking error potential are relentless, physicians operate in a state of constant urgency. (This is also a potential explanation for unreadable handwriting.) It’s no surprise, really, that losing the better part of a day to the EHR will cause significant frustration.

Naturally, since 78 minutes extra per day was the average, some EHRs took much longer and others took less time. The EHR that took the least amount of time was VA VistA, which averaged about 20 minutes per day and 100 minutes per week—80 percent less than the overall average of 6.5 hours.

Think about that. How much more would your level of job dissatisfaction rise if technological change added 78 minutes to your day versus 20?

Understanding why VistA requires so much less time means understanding the evolution of system design. VistA was created to be rapidly adoptable. All other goals were and have remained secondary. Physicians interning at the VA rotate through frequently and cannot spend a day or more learning the health IT system. Generally, they get a two-hour orientation and the rest they learn on the job. The EHR has to be inherently straightforward to learn and use or efficiency is lost and the VA ends up like every other hospital in America.

VistA’s core functions—orders, notes, notifications and chart review—are easy to manage and enable rapid workflow. In large measure, this is the product of a community development process as opposed to a company competing in the feature function wars that have bloated most EHRs. The effectiveness of VistA is validated by a recent Medscape EHR study in which VistA CPRS was rated number 1 overall for physician satisfaction, number 1 in data entry, and number 1 in usefulness as a clinical tool.

The key component in the effective use of EHRs is time. A well-worn axiom tells us it takes a long time to write a short note. So, perhaps it also takes a long time to design an EHR that generates short notes … that helps get work done efficiently. Ultimately, EHR’s should create time for patients, not take it away. Technology may make patients safer—an arguable point for many in medicine—but if it can’t improve efficiency, it will only contribute to what looks like the ongoing disintegration of American healthcare.

23 replies »

  1. The survey addressed “time wasted” or losing time, which is obviously a subjective measure of perceived waste of time. It is not an objective measure of time spent.

    Respondents who put extra time that they saw as part of their work may not have perceived it as wasted or lost. This is why dissatisfaction may correlate with the physicians perception of EHR’s wasting their time.

    Learning directly from Dr. Macdonald about these nuances would be very helpful.

  2. Thank you. I was beginning to fear for basic scientific interest with this post.

  3. I think the fact that 40% of physicians did not report losing time has been lost in the discussion of this study. I actually emailed my colleague Clem and asked him why more detail was not provided in the publication of this study. He said that JAMA required him to reduce the original paper to a research letter, 600 words instead of 3000. It would have been very helpful to have more detail and do more analysis. I am not sure how much the CPRS statistic from this study means anything.

    I did address this study in more detail and the larger issue of efficacy-effectiveness studies in my blog:

  4. Christine,

    The JAMA article cited said the average was 78 minutes. I previous study had it at 48 minutes. These are surveys and have their limitations but these were the average results for “time wasted”. So if VistA was at ~20 minutes then other ehrs users were probably significantly above the average or >>78 min/day or upwards of 8 hours per week? In this survey perceived time wasted may explain dissatisfaction. Perception is reality.


  5. With the further growth of physician emr and ehr platforms, they are supposed to increase productivity. I would say some do better than others. Wasting 6.5 hours is probably on the slower than average end of the spectrum. I would like to think that some of these developers of ehr and emr platforms consider the efficiency factors more than discussed here.

  6. Well, we’re going to have to agree to disagree on that one. I have spent a lot of time in VA facilities working with physicians and observing. And the defensive documentation issue is a real one.

    Agree on the clinician-developed point and feature wars.

    I’m not going to defend either practice. Fair points.

    The key is going lightweight and usable.

  7. We should survey all physicians in university systems that use both the University system; either EPIC or Cerner and the VIstA system at the VA. That would be a telling survey. Then survey the patients too.

    We would see the difference between the two EHR’s and the difference between the two health systems. it would be very difficult to discriminate the factors…..and that would be exactly the point. The business model a system is built for reflects directly on its design priorities and user experience.

    We would see the tale of two systems.

  8. Physicians did have paper in efficiencies. But, not in filing etc.

    Just remember that the average was 6.5 with the low end at 1:66 with VistA. So, the high end of the range could be over 10 hours. The Medscape survey is helpful in understanding where the current dissatisfaction lies.

    This EHR time is directly related to caring for, processing and getting paid for seeing their patients.

  9. I don’t think that can go very far in explaining VistA taking 1/5 the time of the average physician. I do have a lot of first hand knowledge of VistA use outside the VA and am biased. Just ask a physician who trained at the VA about VistA. 65% of our physicians did and they will tell you their truths.

    The key is VistA was designed organically in the field by clinicians. Most systems were designed top down by companies competing on feature functions.

    Design priorities best explain the time difference and dissatisfaction.

  10. Your points are valid questions.

    But, VistA is used outside the VA. Both the Medscape survey and this study do not discriminate where the users are working. But, both are focused on physicians in practice, not working at hospital clinics.

    I urge you to look deeper into VistA, it is available and fully certified with 100% attestation.

  11. Once you are inside the walls, the VA system works well and the VistA EHR is one of the best as noted in this article Due to congressional inactiion, the resource constraints on physician staffing in certain areas of the country are being interpreted as general inefficiency. Many like yourself will throw the baby out with the bathwater, Bubba.

  12. cant speak to the inpatient aspect of epic, do know the outpatient side inside and out, cringe everytime a get hospital records because it will have 30 pages of normal blood sugars, blood pressures, pulses etc and buried within those you will find one small patient note. the doctor generally just puts in a standard template that is usually worthless or he cuts and pastes a previous note that is also full of errors and really does not reflect what going on either, or he uses dragon dictation and if that is the case then the note is usually total jibberish.

  13. bird, your experience is not my nurse wife’s experience with her hospital’s new Epic system. Both nurses and docs input into this bed side system. Patient monitors auto download vitals and she sees one screen not 5 as in the last system.

  14. OK, I reacted badly to the infographic but now that I read the original article I realized that this article was seriously flawed (statistical illiteracy?).

    The original article clearly states that the average lost time was 48 minutes a day for attending physicians and 18 munutes a day for trainees. Of the 59.4% of respondents who lost time they lost 78 minutes.

    Tthe 78 minutes is not a population or sample average of “physicians” as dishonestly summarized above. It is a particular egregious mistake as the article goes on to discuss that “some EHRs took much longer and others took less time” . Clearly 40.6% are not using more time on EHRs and are part of the relevant discussion and deserve not to be ignored.

    I could go on to say the article itself uses some poor analytic constructs to exagerate their case (the “assumption” of a 5 day/ 8 hour work in order to construct an inflated extrapolation of weekly and percentage estimates is a good example).

    This subject deserves proper and honest discussion, it is too bad that was not the case here.

  15. having been on epic since 2000 and i have seen the changes that have occurred and very few of the changes have been to improve the patient experience and care of the the patient. The emr has changed from monitoring the patient to monitoring the doctor. I understand the importance of sharing data, running reports, checking rosters etc but when you can find everything you need in the ehr except what is wrong with the patient then the system needs to change. The epic people are the worst people to deal with. They know they have market captured and dont care about what you as a provider have to say.

  16. That infographic is very poor. It fails at the very basic task of communicating visually. The whole point of a graphic is so visual elements represent data in a clear, concise manner. While the information is clear (through text), the visual is bewildering (people? people-hours? decimal people? black people? grey people?).

    I can only wonder how this relates to the poor design of EHRs.

  17. If docs are spending 6.5 hrs/wk on EMR what were they spending on pen and paper “in the good’ole days”?

    Please include filing, searching, storage and organizing, in your answer.

  18. Another issue worth raising in the discussion here –

    VA docs don’t have the same worries about liability that private sector docs do.

    It IS possible to sue a VA doc, but it rarely happens …

  19. So VA clinicians have to document with an eye towards justifying a 99214 over a 99213? And they have to fulfill meaningful use stage I and II? That would really surprise me and I would have a hard time believing it unless it is proved from a very reliable source.

  20. Recently timed myself to find avg 10 min per chart to finish visit note. ouch! would love paper, dictating, or something

  21. NTM having to “break the glass” 10x to see a pts info, getting logged off every 10s, getting “locked out” for mistyping pw, etc

  22. Bubba’s take:

    I’m not sure I would be pointing to the VA as the height of administrative efficiency, Edmund