Working as Designed

I was listening to a podcast recently, and heard Mike Monteiro being interviewed. Mike is a designer, author, speaker, and apparently a malcontent. In other words, he’s my kinda guy. He’s written a new book, and that’s why he was making the podcast rounds.

Mike’s book, Ruined By Design: How Designers Destroyed the World, and What We Can Do to Fix It, focuses on how we got to where we are, specifically dealing with technology and its effect on our lives. In the introduction, he writes:

The combustion engine which is destroying our planet’s atmosphere and rapidly making it inhospitable is working exactly as we designed it. Guns, which lead to so much death, work exactly as they’re designed to work. And every time we “improve” their design, they get better at killing. Facebook’s privacy settings, which have outed gay teens to their conservative parents, are working exactly as designed . . .

When I read this, I immediately thought of the electronic health record (EHR). Many in the physician community solely blame EHRs (and the software vendors who create them) for all the ails of the practice of modern medicine. In my view, EHRs are working exactly as designed. Or rather, exactly as we designed them to work in the United States. Before I get deluged with tweets and comments from doctors who think that EHRs can work better than they do today, let me clarify: EHRs can and must work better than they do today. We need drastic improvements in usability, user-centered design, and interoperability. All of that is true, yet . . .

In American healthcare, we can’t agree on what goes on the problem list. We’re not even sure what the problem list is. I keep hearing about this past medical history thingamabob. Is that the same as the problem list? If I prescribe a medication, but the patient isn’t taking it, should that med be on the medication list? In the United States, we can’t document by exception (translation: we can’t say the physical exam was normal except for this and that). Physicians are counseled that if you didn’t document it, you didn’t do it. Hence our physical exam has to be long and tedious, else we risk not getting paid, not being judged as practicing quality medicine, and not being able to effectively defend ourselves against a malpractice suit.

Here’s a representative story about American healthcare and its effect on the EHR. In the emergency department, it’s common to check for hidden or occult blood (often abbreviated “OB”) in the stool or rectum as blood doesn’t belong there, and its presence can help point to a diagnosis. In the recent past, this test would occur by the physician doing a quick rectal exam, rubbing his/her glove on a small card with special chemicals on it, and adding a few drops of a solvent. If the doctor sees blue on the card, occult blood is present. The doctor would typically document this in the chart by writing “OB neg” or something like that.

Nowadays, the workflow is more like this: physician goes into the EHR to order the occult blood test. He/she performs and interprets the test as before. The doctor then goes back into the EHR to “result” the test as regulations require most lab tests to be documented in specific ways. The doctor has to enter the result of the OB test, of course. Be he/she also must enter the lot number of the card (for tracking purposes) and the expiration date of the card (to prove that it hadn’t expired, I guess). He/she must show in the EHR that a proper control test was also done on the same card (for quality purposes). The physician must note the result is “final” so that the hospital can properly bill for the test. There might be other steps as well based on which EHR is being used.

There’s a lot to unpack with respect to this occult blood workflow, and we can all lament that a particular EHR or another doesn’t make it as easy as it should be for the physician to order and document this test. But . . . the EHR vendors (at least the ones with which I’ve worked) didn’t decide that physicians must order the occult blood test. The EHR vendors didn’t decide that the lot number and expiration date had to be typed in. The EHR vendors didn’t require that the physician specifically note that the control test was correct. All of these hoops are peculiar to American medicine. And electronic health records built for doctors who practice American medicine have to include these tools.

I’ll repeat my argument: EHRs in the United States are working as designed. Few of us are happy with how they’re working. We need to make the technology better, but if we don’t work to improve the system in which we practice, the technology will never – can never – get where we need it to be.

Am I off target here? Do you have other stories of American medicine gone rogue?