I recently met with Dr Jeffery Gold, a doctor of pulmonology at OHSU who’s research is with EHR-usage. Aside from caring for his patients, his work is in the ICU where they do practice rounds trying to determine what errors are not cognitively recognized within the context of in situ EHR usage.
The current EHR being researched is Epic, of course, and he’s interdisciplinary with a number of colleagues and other investigators in this realm of work. Dr Gold has been interested in learning how the usage of the EHR, not from the end-user point of view but from the design of the interface, effects what is introduced into the patient’s treatment plan based on what is presented.
For example, there are correlations in the metadata in the blood sugar level trends of a diabetic patient, and what might this metadata tell us regarding the patients meds?
One thing that’s been observed is that reading progress notes provides relatively little information about what meta-treatment aspects there are with a particular patient. There are a lot of factors and inputs that are observed, one of which is eye tracking, of which they’ve done a lot of.
One of the problems that Dr Gold’s research has found is that there are many screens that are viewed by clinicians that don’t have the information they’re looking for. This was a point that he brought out in his presentation at the recent Oregon HIMSS Ignite event (and see below).
Dr Gold’s work is very exciting to me in that it’s interdisciplinary and it’s seeking the best way to use this interface, and how to design the use of this tool to maximize the benefits that this new tool in healthcare provides. We’ve just scratched the surface and most users don’t know how powerful the EHR is.
This work is probably in it’s infancy because for the most part the unknowns are unknown at this point. We simply don’t know what the right questions to ask are.
But we’re getting closer as we find out more about the questions that will be asked when we better understand what the data looks like, and what the data behaves like when the data is about the conditions of a patient during “X” timeframe, as these factors are involved with the care and treatment of the patient.
What factors are effecting the outcome, what aren’t, and what questions need to be asked? These issues are fraught with challenges and excitement, of course. I’m glad to have asked Dr Gold these questions about his work and it’s something I find very interesting. Exciting!
Here’s some of what I took away from Dr Gold presentation at the recent HIMSS Oregon Ignite event titled, “Simulation to Improve Use and Safety of Electronic Health Records.”
One thing we’ve learned about EHR usage is the way the data is fragmented in the EHR. In Dr Gold’s research for instance, for one patient, there were 10 minutes of looking through the record, 20 individual screens, 40 total screens (and an average of between 80 – 90 mouse clicks per patient per 10 minutes).
All those screens mean lots of data. The average ICU patient will have between 1400 and 2000 data points entered per day. For the average ICU patient there are 40 lab alerts per day from labs alone. Data is so abundant that sometimes critical factors are not cognitively recognized. The EHR simulation that Dr Gold uses shows that even the most trained practitioners perform poorly with EHR simulations that have built-in errors.
Watch a video of Dr Gold’s Ignite presentation here.