Time Capsule: Put Down That Computer and Listen: Why Filling Out and Reading EMR Data Screens May Cause Doctors to Shortchange Patients

I wrote weekly editorials for a boutique industry newsletter for several years, anxious for both audience and income. I learned a lot about coming up with ideas for the weekly grind, trying to be simultaneously opinionated and entertaining in a few hundred words, and not sleeping much because I was working all the time. They’re fun to read as a look back at what was important then (and often still important now).

I wrote this piece in May 2008.

Put Down That Computer and Listen: Why Filling Out and Reading EMR Data Screens May Cause Doctors to Shortchange Patients
By Mr. HIStalk

I’m a fan of an interesting learning concept called the Illusion of Knowing. Here’s what it says: you’ve read something carefully, sometimes more than once, so you’re confident you’ve mastered whatever it says. Later, however, when hard-pressed to put the information to good use, you blank out. You didn’t know it after all – you just thought you did because you had passively read it.

(Cue sweat-inducing college final exam dream. You couldn’t find the exam room, and when you finally did, you realized you hadn’t attended any of the classes … you know the rest).

Anyway, some Harvard doctors made me think of that with their recent NEJM report on EMRs. They question whether EMRs really improve care given their emphasis on creating reams of bland and predefined information, but with no capability to encourage fresh, individualized thinking to diagnose and treat patients.

(Note: I’m reading between the lines since the actual lines themselves require a NEJM subscription, which I don’t have because I’m cheap and they use a lot of big words when little ones would do fine).

The authors cite a doctor colleague who said that hunting for useful information in an EMR is like the Where’s Waldo? games of a few years ago. The kicker is this: that colleague is so frustrated with all the meaningless junk in EMRs that he makes index cards to track what’s important.

That’s where I thought of the Illusion of Knowing. A doctor could read all the EMR screens and figure, “Everything I need to know is right there, so if I study it long enough, I’ll figure out how to improve this patient’s life.” That’s EMR Nintendo: recognize and react to some event, which may seem like practicing medicine to a programmer since that’s how logically programming works.

Here’s a problem: doctors don’t have the time to conduct scavenger hunts for vital facts in the handful of minutes per encounter that the benevolent insurance companies and practice managers allow them.

Second problem: EMRs aren’t set up to allow automatic or manual grading of individual factoids, so everything looks potentially important.

Third problem: EMRs try to turn freeform and sometimes tentative thoughts into dropdowns and template-driven generic verbiage that may destroy their original context (that’s what programmers do: impose order and create retrievable database information, so it’s not really their fault).

Another article that was published at about the same time extols the virtues of speech recognition systems. Those create more voluminous and anecdotal information, but the context is perfectly preserved. Unlike discrete data, doctors could re-read a narrative and glean new information after the fact. Programmers hate bunches of text that don’t lend themselves to convenient database structures (although natural language processing can reverse engineer some of it back into data fields).

We in the industry could debate the merits of templates vs. narrative, but that discussion is moot. The real problem is medicine itself. A table of dry patient facts can help support diagnosis and treatment decisions, but even fresh-faced doctors know that patient care isn’t a video game of spotting a symptom and blasting it with drugs or surgery. The first thing they learn in medical school is not how to read charts or write orders, but to go into the patient’s room and look and listen. Sometimes the least-obvious information is the most useful.

Perhaps a redesign of EMRs is in order that takes semantics and metadata into account to better reflect the physician’s thought process and judgment rather than just trying to force those thoughts into a convenient data structure that looks good in a table and uses classification tools that say in black and white what might be better expressed in not just shades of gray, but in rainbows of colors. Or, maybe a well-designed study (not financed by EMR vendors, most likely) would find that chatty paper records lead to better outcomes than terse and categorized electronic ones.

The bottom line is this. EMRs have affected patient outcomes only modestly, if at all. If doctors still have to make index cards, maybe legacy EMR design should be revisited.