Asymmetrical Information - Megan McArdle

01.25.13

The Unintended Consequences of Electronic Medical Records, Continued

They making it easier to copy and paste rather than write

Mickey Kaus has more from readers on the unintended consequences of electronic medical records.  And what he has is horrifying:  I’m an MD who makes a living reviewing medical records for attorneys. Before I switched to this work I was an internist for a large HMO, entering my notes into an EMR. My wife is an internist who works for a company that keeps records on an EMR.

. . . EMRs encourage doctors and nurses to cheat and lie. EMRs have made medical records inaccurate and unreliable. When I read medical records nowadays, I often can’t tell what the hell happened.

As your Alert Readers note, EMR data entry is very time consuming. One workaround is for EMR software to include pre-programmed template notes. When a doctor sees someone with a cough, we no longer write or dictate what we actually did and saw. Instead we find the Upper Respiratory Infection template, which has a standard note for what would be seen and done in an average URI visit. We may or may not type a few words at the bottom. We click OK. The EMR records the note. The nuances that make this URI different from the average one are lost. In an EMR, every URI is an average URI.

When the medical condition is just a cold, losing the details is not a big deal. But every heart attack is not like ever other heart attack— complex medical problems can’t be captured by pre-programmed templates. Templates are used anyway. Data entry just takes too long.

I once reviewed a hospital record from a large national medical center that I can’t name, but [you've] heard of. The patient had a major operation. The operative note was incredibly good. Page after page it recorded in exquisite detail exactly where the surgeon cut, exactly what he retracted, exactly what he saw, exactly what detailed care he took to avoid injury to this organ and that one. I was impressed. I remember thinking, “Wow. No wonder this place has a national reputation.” This was the best documented operation I had ever seen.

In spite of this operation, the patient got worse. Four days later she went back for a repeat of the same operation. And the second operative note was the exactly the same as the first. Identical. Page after page, word for word, exactly the same. Leave aside the impossibility of having two multi-hour operations go exactly the same way, it is not possible to dictate or write two multi-page op notes that are word for word identical. The op notes were frauds. They were templates, worked out with the hospital risk management department to describe what should happen, and entered in the EMR with one click of a mouse. What actually happened? No one can tell.

Please understand, this is not an exception. This is how things are done these days.

Another reasons EMRs lie, is that they are subject to “template bloat.” When you write a template (I’ve done this myself), you don’t just cover what usually happens, you cover important things that might happen. The diagnosis is a cold. But it might be a stroke—so the URI template records the doctor doing a detailed neurologic exam. It might be a heart attack—better put in a complete cardiac exam. And so on. EMR records for colds record pages of stuff that never happened

I want to believe that this is not true, because it's disastrous, and I can't bring myself to believe that doctors would do this.  Maybe the old-fashioned hand-written notes were illegible and incomplete, but is that really worse than something which is just . . . made up?  At least with bad notes, you know that you're missing stuff.  With lovingly written notes about a fictional, perfect operation that didn't happen, you might be dangerously misled by the mistaken impression that you know what's going on.

Can readers comment?  Could this possibly be true?