EMR Physician Builder Day 2 - prose out of buttons

/ Dartmouth-Hitchcock Medical Center/
You have the helm - mural at Epic (artist unknown) - photo & title by Rob Schertzer

You have the helm - mural at Epic (artist unknown) - photo & title by Rob Schertzer

"How do I get prose out of buttons?" It's part of the art of creating artificial intelligence...and by that I strongly emphasize ARTIFICIAL. There is nothing about an EMR today that would come close to passing the Turing Test for artificial intelligence. An EMR however, if configured properly, can use rules to adapt to each patient encounter.

A patient presents to the Ophthalmology department for a glaucoma related problem. Should the Ophthalmic technician:

  1. Have to take a complete history regardless the presenting problem?
  2. Open the policy and procedure manual in the back office to figure out what their work up should be? Or,
  3. Should the act of typing in the reason for the visit automatically trigger the EMR to open a form or chart note specific to the chief complaint? 

If you said number 3, then we're on the same wavelength but it is easier said than done. An approach like this could reduce the time it takes for this initial work-up from 30 minutes down to just 10-15 minutes. If you said you already have this in your practice, then please call me! If you don't already have this, it is indeed possible to create programming routines that, when you open the Note section of your chart, a pre-configured history and physical will be generated by looking at the reason for the visit. This triggers a subroutine that calls for a smart link to find the correct smart text that is needed for this patient. Just like it sounds, this calls for a lot of forethought. This needs a physician builder who understands the work flow and a production team to pull it off.  Each unique type or group of visit types each need a different work-up form or note.

Once you've gone as far as ticking off checkboxes or clicking on buttons in the patient's chart - what happens with that information? As mentioned in the Day 1 article, first you have to make sure you have used discreet elements when programming these so that each of these items are actually retrievable data. But how do you report these findings in the chart note or letter to the referring doctor? You have to convert buttons to prose. Guess what, again you need a physician builder working in conjunction with the production team to make sure the data gets converted into words that make sense. This can be a well formatted list, sentences, or even a table.

It seems that I am getting even more convinced of a couple of things through this physician builder course. The amount of work needed to customize enterprise wide EMR systems is mind-boggling. It requires physicians who know what the EMR is capable of achieving to be actively involved in the process. Customizing the EMR can result in big time savings for each patient as well as a higher quality of care thanks to Best Practice Advisories (upcoming article...stay tuned!) or similar event triggers specific to your practice getting built into the system. The time invested up front should more than pay for itself in the long run but few people are likely ready, willing or able to make that commitment.