EMRs won’t prevent the spread of Ebola (or the next scary outbreak)

about-ebolaAccording to IBM, there are 2.5 exabytes of data created every day, and most of it is unstructured. Imagine receiving all the words ever spoken by human beings on your doorstep each and every day. Now, imagine consuming that, making sense of it and trying to keep up with the ever-accelerating pace of data creation each day.

As a physician, I experienced firsthand the angst that comes with trying to keep up with even a very specialized scope of expertise. Thanks to the overwhelming quantity of peer-reviewed publications and practice guideline updates that only increase each year, we are long past the time when a clinician could possibly keep up with all the advancements in their own practice area, let alone those of adjacent areas of medicine or the latest public health concerns on a global scale.

When a patient with an Ebola infection recently presented to a Texas emergency department (ED) and the diagnosis not made on the initial visit, I was bewildered that an electronic medical record (EMR) workflow failure was implicated – even if that mention was later retracted. EMRs as they are designed today cannot prevent the spread of Ebola – or any other public health scare; they do not create the situational awareness needed to do so.

I am sure many health systems are now creating hard stops in the EMR workflow to ensure that patients always have a travel history entered, similar to how an allergy history is taken. This seems logical. Many are probably creating cascading questions that require more in-depth questioning based on responses and expanding the visibility of answers to make sure all staff can see them.

Will this work for Ebola risk? Quite possibly, given that media attention has created a heightened sense of awareness. However, unless the person taking the history at any of the thousands of EDs in the U.S. alone understands the significance of someone saying they were recently in “Monrovia” or “Freetown,” the utility of this questioning is limited. Having grown up near Freetown, Massachusetts, my reply of ‘Freetown’ wouldn’t necessarily raise red flags in certain areas of the country.

Now expand that out to the thousands of known endemic illnesses with long understood geographic connections, even in the U.S.: Lyme disease in the Northeast, Coccidiodomycosis in the San Joaquin desert. How many of these illnesses – which cause far more social harm than Ebola – are missed every day? No amount of point-of-care technology engineering can ever solve this. There are an infinite number of ways that someone can say they were spending time near Lyme, Connecticut three months ago without ever arousing a moment of concern by the diligent person taking the history in Los Angeles because the EMR workflow requires them to. Those unstructured answers will be meaningless on their own. More data does not create awareness.

The best approach we can take is to connect a reasonable amount of data collected at the moment of care with a real-time cognitive computing engine capable of providing decision support in the form of situational awareness. There are many valuable entities like the CDC who track illnesses related to geography, industry mishaps, outbreaks and other public health events. This database changes every day. No human workflow is capable of keeping up with this one component of an ill patient’s presentation in the way that powerful, real-time and cognitive analytics can. By connecting those at the moment of care with these technologies, we will actually reduce the amount of questions, answers, and data needed and instead focus more on creating insights about our patients that every clinician can use.

Education will not address this challenge because it is impossible to know what the next Ebola will be. To address these types of issues, caregivers do not need knowledge. They need instantaneous access to situational awareness based on the unique context of the patient in front of them.