Mirror, Mirror, On the Wall, Which EMR is Least Traditional Of All?

Short Link: http://j.mp/64geX6

Originally titled “Traditional EMRs are Problematic, but Let’s Not Throw the Baby Out with the Bath Water,” then “Have Your Cake and Eat it Too: Structured EMR Data AND High Productivity,” I eventually decided that “Mirror, Mirror, On the Wall, Which EMR is Least Traditional of All?” was most clickable. (William Safire said “Avoid cliché’s like the plague”, but while I admired his column in the New York Times I took his advice with a grain of salt.)

The “traditional” EMR is taking it on the chin lately.

punch

I’ve thrown a few punches myself…

http://www.google.com/search?q=%22traditional+EMR%22+site%3Achuckwebster.com

…but I’m not the only one. Check it out yourself…

http://www.google.com/search?hl=en&q=%22traditional+EMR%22

Or take my word for it. I grabbed quotes from the first three pages Google returned and summarized them below by way of paraphrase. I changed vendor and product names to protect the….(innocent? guilty?). Of course, by the time you click the list will have evolved and lengthened (and this post will be on it, how is that for self-reference?).

  1. Our EMR implements in much less time than a traditional EMR
  2. Our EMR costs much less than an a traditional EMR
  3. Our EMR makes physicians more money than traditional EMRs
  4. Our EMR requires less change to existing workflows than traditional EMRs
  5. Our EMR has fewer security vulnerabilities than traditional EMRs
  6. Our EMR helps physicians see more patients than traditional EMRs
  7. Our EMR has the 20% of traditional EMR features, 80% of their value, and 200% of their usability
  8. Studies have refuted the claims of traditional EMR vendors
  9. 30-70 percent of traditional EMR implementations fail
  10. Traditional EMRs cause physicians to see fewer patients than pre-EMR implementation
  11. Traditional EMRs cause physicians lose revenue relative to pre-EMR implementation
  12. Traditional EMRs do not eventually return to pre-EMR implementation productivity levels
  13. Traditional EMRs require pointing and clicking that wastes physician time
  14. Traditional EMRs distract physicians from focusing on their patients
  15. Traditional EMRs implement features mandated by bureaucrats, not physician needs
  16. Traditional EMR features are so cumbersome that some are often not even ever used
  17. Traditional EMRs are sold by salespeople know their software does not help physicians (harsh!)
  18. Traditional EMRs create documents full of cookie-cutter language in which relevant information is hard to find
  19. Web-based EMRs are the alternative to traditional EMRs
  20. Document imaging-based EMRs are the alternative to traditional EMRs
  21. Advertising-supported EMRs are the alternative to traditional EMRs
  22. Less-is-more EMRs are the alternative to traditional EMRs
  23. Speech recognition-based EMRs are the alternative to traditional EMRs
  24. Natural language processing-based EMRs are the alternative to traditional EMRs
  25. EMRs that visualize patient data in a different way are the alternative to traditional EMRs
  26. Open source EMRs are the alternative to traditional EMRs
  27. Mind-controlled EMRs are the alternative to traditional EMRs (just kidding, but cool video)

Pile-On!

If you believe EMR marketing departments these days, most EMRs are not traditional at all, perhaps an example of the Lake Wobegon Effect (where “all the children are above average”). As my mom used to say, “What in the world is going on here?” (“Nothing Mom!”) What is it about traditional EMRs that have made them so radioactive even traditional EMR vendors are beginning to attack them?

EMR vendors with a-, un-, non- or anti-traditional EMRs (not sure of the correct prefix here) invoke and attack a “traditional” EMR straw man with some or all the following characteristics. A traditional EMR has structured template-driven forms with complicated, inflexible, time-consuming, distracting (from the patient) drop-down menus, check boxes, radio buttons that require massive change to physician workflow. Traditional EMRs have so many features, mandated by so many bureaucrats, that access to the right feature at the right time is effectively impossible, because it is buried. Generated patient notes and letters contain repetitive superfluous cookie-cutter language in which relevant information is hard to find and therefore often ignored.

I think that’s the heart of the traditional EMR stereotype. I refer to them as traditional, hunt-and-peck, clickity-clickity-click-click-click EMRs. 

Traditional “Hunt and Peck” EMRs are not usable for high-productivity data and order entry, nor can they be flexibly adapted to anticipate individual user requirements and preferences. These EMRs are not designed on the process-aware foundations that are required for systematic optimization of clinical performance, workflow efficiency, and user and patient satisfaction.

However, some traditional EMR critics go too far and advocate getting rid of the extraordinarily valuable structured data. I’ve been criticizing “traditional” EMRs for years, publically since September 15, 1999, to be exact. I know the culprit–its flaws and its virtues–and structured data is not the problem.

Generating useful structured data requires the imposition of precise structure and meaning on collected data. Such data are easier to aggregate (important for clinical research such as outcome studies), easier to transport (in the sense of transferring specific data between applications), and more appropriate for driving the automated performance of patient-specific tasks such as alerts and workflows. I published my solution to the need for both structured data and high productivity in 2001, when I wrote:

“In our opinion, the combination of structured data entry, workflow automation, and screens designed for touch screen [today including stylus] interaction optimally reduces inherent tradeoffs between information utility and system usability on one hand, and speed and accuracy of data entry on the other. Successful application of touch screen technology requires that only a few, but necessary, selectable items be presented to the user in each screen. Moreover, workflow, by reducing cognitive work of navigating a complex system, makes such structured data entry more usable.” (also discussed in “The Cognitive Psychology of EMR Usability and Workflow”)

Back then “workflow automation” was accomplished by workflow management systems. Today it is associated with business process management and adaptive case management systems (there’s a interesting and relevant debate between those to “camps”, but I’ll leave that to another post).

[CW: The following pertains a EMR workflow system that won the first three HIMSS Davies Awards. Many of the technologies it pioneered, such as workflow engine and process definitions, are beginning to appear in EMRs today.]

Let’s go through the list of criticisms of “traditional” EMRs one by one.

  1. Our EMR implements in much less time than a traditional EMR
    • You’ll be “Live in Five” (back to pre-EMR implementation patients-per-day in five days). Since automated workflows do so much without requiring human intervention, and since tasks are pushed to the right users without them having to perform laborious navigation, learning curves are ultra short.
  2. Our EMR costs much less than an a traditional EMR
    • Our total cost of ownership is substantially less than traditional EMRs because the largest single expense to owning an EMR is the time it takes for you to learn and use it. And if you never do, well, that is the biggest expense of all.
  3. Our EMR makes physicians more money than traditional EMRs
  4. Our EMR requires less change to existing workflows than traditional EMRs
  5. Our EMR has fewer security vulnerabilities than traditional EMRs
    • Workflow management systems have an additional layer of security due to role-based task execution permissions and detailed workflow audit trails. Workflow management systems represent individual tasks at a higher degree of granularity than non-workflow management systems. Workflow engines execute process definitions that contain these tasks. The workflow engine consults a role-based permission system to decide whether or not a user has the right to initiate or complete each task. As a result, a workflow management system can enforce controls over what can and can’t be done before hand, and provide a more detailed audit trail afterwards.
  6. Our EMR helps physicians see more patients than traditional EMRs
    • See response to claim 3
  7. Our EMR has the 20% of traditional EMR features, 80% of their value, and 200% of their usability
    • See post about “EMR Featuritis”
  8. Studies have refuted the claims of traditional EMR vendors
    • Three primary care physicians–two pediatricians and one obstetrics/gynecology and family medicine physician–submitted detailed documentation, hosted site visits, and won the first three HIMSS Davies Awards for ambulatory care excellence.
  9. 30-70 percent of traditional EMR implementations fail
    • “The basic problem is that implementing *most* EHRs is an act of reengineering, and reengineering is a high risk endeavor.” (Source) Implementing an EMR Workflow System is not an act of reengineering, resulting in a virtual 100% rate of successful implementation.
  10. Traditional EMRs cause physicians to see fewer patients than pre-EMR implementation
    • See 3
  11. Traditional EMRs cause physicians lose revenue relative to pre-EMR implementation
    • See 3
  12. Traditional EMRs do not eventually return to pre-EMR implementation productivity levels
    • See 3
  13. Traditional EMRs require pointing and clicking that wastes physician time
  14. Traditional EMRs distract physicians from focusing on their patients
    • One Georgia pediatrician, a winner of the HIMSS Davies Ambulatory Care Award of Excellence for his use of an EHR workflow system, shows his attentiveness using one hand to steady an energetic child and the other hand to enter data and orders out of the corner of his eye, facilitated by large colorful buttons. A workflow engine pushes screens in preprogrammed sequences so he is not distracted by screen-to-screen navigation. His focus remains uninterrupted and attentive to the concerned parent. (Article: “What Makes a Great Pediatric EHR?”)
    • A Chicago pediatrician, who also won the HIMSS Davies Ambulatory Care Award using an EHR workflow system, notes that its customizable workflow has made his office so much more efficient that he can see more patients and spend more time with each patient. Efficiency allows more time to be available to parents and patients.  (Article: “What Makes a Great Pediatric EHR?”)
  15. Traditional EMRs implement features mandated by bureaucrats, not physician needs
    • Short anecdote: I presented to an audience of health professionals almost all of whom worked for various government agencies. An animated question-and-answer period ensued. Many of the questions where “Have you thought of…?”, “Why don’t you add…?” To which I answered “Yes”, “Yes”, and “Yes” and “Our sales people aren’t being asked for it,” and  ”It hasn’t lost us any sales.” There was a moment of befuddled silence, and finally someone from the audience commented “That’s a really stupid reason for not doing something!”
  16. Traditional EMR features are so cumbersome that some are often not even ever used
    • “Since 1994 we’ve been adding pediatric-specific features to the first Windows-based pediatric EMR (and workflow system), however its completely customizable workflows “anticipate” what you need and where you want to go, unlike “hunt-and-peck” EMRs that force you to click your way through cluttered screens and lengthy picklists. Our philosophy is to optimize, not maximize, pediatric EMR features. With an EMR workflow system you can have your features and use them too.”
  17. Traditional EMRs are sold by salespeople know their software does not help physicians (harsh!)
    • Our sales folk sleep very well at night! I do too. I believe EMR workflow systems will be transformational at healthcare industry, organizational, professional and personal levels.
  18. Traditional EMRs create documents full of cookie-cutter language in which relevant information is hard to find
    • I took two courses on natural language generation in graduate school and even presented a paper at a workshop on the topic. Automating the writing of fluent, appropriate, and natural language is difficult. However, getting rid of structured medical data and databases is akin to throwing the baby out with the bath water. The issues are complicated, interesting, and reflect well on EMR workflow systems.
  19. Web-based EMRs are the alternative to traditional EMRs
    • Many workflow management systems and business process management suites are web-based. It’s only a matter of time before EMR workflow systems are delivered in this manner too.
  20. Document imaging-based EMRs are the alternative to traditional EMRs
    • Ironically, many of the first commercial workflow systems (such as FileNet’s WorkFlo system back in the 80s) were document scanning/imaging solutions. Why ironic? I am arguing that EMR Workflow Systems are the next step in EMR/EHR evolution (see my recent white paper on this topic), but workflow systems (and the underlying workflow management systems that create and manage them) are decades old. Healthcare is that far behind other industries in adopting workflow technology.
  21. Advertising-supported EMRs are the alternative to traditional EMRs
  22. Less-is-more EMRs are the alternative to traditional EMRs
    • See 7
  23. Speech recognition-based EMRs are the alternative to traditional EMRs
    • I worked on a speech recognition project in graduate school (writing natural language syntax grammars for the Pilot’s Associate project) so I know a bit about the topic. The Holy Grail is large-vocabulary, speaker-independent, continuous speech recognition. I’m a fan. But it is not an “alternative” to structured data EMRs, any more than scanning/imaging. [3/3/12 update: This is changing!]
  24. Natural language processing-based EMRs are the alternative to traditional EMRs
    • However words get digitized, whether through speech recognition or optical character recognition (or physician fingers), they need to be combined into phrases and sentences and meaning derived. There’s been some interesting work in this area. But it’s not ready to replace structured data entry. [3/3/12 update: This is changing!]
  25. EMRs that visualize patient data in a different way are the alternative to traditional EMRs
    • New ways to view data result in more decisions to make about which view is appropriate at each workflow step. However, the way in which data is presented at a particular step can be determined automatically by the workflow engine. New ways to view data increase the need for EMR workflow automation, not diminish it.
    • EMR workflow systems also generate a lot of new workflow-related data that will need to be visualized. Increasingly, as patients begin to interact with EMR workflow systems via Web portals, their activities (infotherapy, supplying information, documenting compliance) will become part of the EMR record proper, and therefore need to be visualized too.
    • Both are examples of a complimentary relationship between new ways to look at patient data and the larger EMR system within which this data is viewed.
  26. Open source EMRs are the alternative to traditional EMRs
    • There are some excellent open source workflow management systems out there. I’d like to see some pressed into service within health care. In fact, if you’ve read any of my posts about how to solve the pediatric medical home care coordination problem, the more EMR workflow systems the better!
  27. Mind-controlled are the alternative to traditional EMRs (just kidding, but cool video)
    • I look forward to mind-controlled EMR workflow systems! I suspect that even in the “mind’s eye” Fitts and Hicks Laws will hold true.

The central issue is whether we can have EMRs with high productivity data entry *and* that can produce structured data.

synthesis

Productivity is the quality and quantify of output divided by cost of input needed to cause the output. The value of data is what you can do with it and you can do many more valuable things with structured, than unstructured, data. The most important input costs are the time and effort of the physician user to master and use an EMR. Reducing these costs is requires improving EMR usability. Improving EMR usability requires improving workflow. The most effective and efficient way to improve EMR workflow is with an EMR workflow system (AKA EMR business process management/BPM system).

EMR workflow systems are compatible with, and can strategically combine all of, the “alternatives” offered by critics of traditional EMRs. In the main, these alternatives are input modalities, delivery platforms, or business models, all of which are needed. But none of them directly addresses the contradictory need for both structured data and high productivity (except perhaps for natural language processing, which is not yet a sufficiently mature technology).

In the long run, it’s not whether you speak into it or write on it, whether you get if for free or have to pay for it (as long as it’s worth it), or whether the code executes in front of you, down the hall, or across the country. It’s whether EMR effectiveness, efficiency, and usability can be systematically improved to achieve well understood, consistently executed, adaptively resilient workflow. Doing so will require both structured data and high productivity. The best, most obvious, most comprehensive, most mature means to achieve these goals are business process management techniques applied to process-aware EMR workflow systems.

So “Mirror, mirror on the wall, which EMR is least traditional of all?”

Good question!

This entry was posted in EHR Workflow. Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

You can add images to your comment by clicking here.