Very Little Manual Entry in EHR

In typical doctor’s note: 18% text manually entered;46%,
copied;36% imported. Lots of copy/paste! Welcome to #EHR! https://t.co/xKKjSoKvCi pic.twitter.com/c3iiTSHnYx

— Harlan Krumholz (@hmkyale) May 30, 2017

This is some fascinating data on how much of a doctor’s EHR note is entered manually versus some other automated method. I honestly wouldn’t have guessed that only 18% of the doctor’s EHR note was being entered manually. Although, from the doctor’s perspective, they still see a copied section of note as something they largely entered manually since a good doctor that copies something into the note generally also reviews it to make sure that it’s accurate for the patient they’re seeing.

What’s ironic is that every doctor I know would love for their note to be 100% automated so that they didn’t have to create any clinical note. In fact, that’s kind of what I outline in the perfect EHR workflow – Video EHR. Doctors would love to just see and interact with patients and have the EHR documentation be completely automated so they could just reference it as needed. Sadly, we’re not there yet. Not even close.

Plus, the critics of this type of automation would argue that automatic note creation will take (many aptly argue that it already has taken) the life and soul out of a note. They appropriately suggest that these auto-generated EHR notes are impossible to effectively read and have ruined patient notes. What used to be an elegantly written (although often illegible) note has now become an auto-generated mess of a note which makes it hard to find the relevant findings, issues, and treatment plan.

Except for a few rare exceptions, these critics are spot on in their analysis of the EHR note. The problem with these criticisms is that it’s not the automation which is making these notes useless. It was the automation’s focus on billing which has made these notes useless. In order to satisfy higher levels of billing, the Jabba the Hutt EHR note was created and is still thriving in healthcare today. Now we’re seeing organizations doing machine learning on this ugly billing notes to try and make the notes useful for patient care.

The difference between a note designed around patient care and one designed for billing is shocking.

What we need to realize is that automated notes don’t have to mean lower quality notes. However, improved patient care has to be the goal of the automated notes and not billing if we want to achieve that vision.

It’s not clear to me if many EHR vendors can achieve both visions of a quality billing note and a note designed around patient care or if it will require a new approach to documenting patient visits to achieve both goals. I have no doubt EHR vendors are going to try to do both. The problem is that most of them already tell themselves that they have a great clinical note that improves care. That attitude is preventing changes to the note that would make them more effective clinically.

I’m all for more automation in healthcare and particularly in doctor’s note creation. Every doctor I know wants to stop being a data entry clerk and spend more time being a doctor. However, we need to rethink our approach to automated note creation so it does more than effectively bill for services. Seems obvious, but I assure you that’s a dramatic change in mindset for many EHR organizations.