Every practicing physician using an electronic health record (EHR) has seen them. Past diagnosis zombies that stumble endlessly through every encounter record. "Coronary heart disease" that the patient never really had, "diabetes" that was only one possibility among many and a "fracture" that never appeared on any x-ray. These undead conditions clutter the technology-enabled health system basically because of two EHR value propositions:
1) saving physician time by using automatic templates that vacuum up every past diagnosis entry and importing them into the encounter note, and
2) enabling a higher payment "complexity" level by documenting every past diagnostic consideration.
So, if someone somewhere anywhere in the past enters "Ehlers Danlos Syndrome" as an unlikely, but possible, diagnosis or a reason to perform additional testing, the EHR will spawn the same immortal entry of "Ehlers Danlos" forever. That diagnosis will emerge from the HIT server-cloud crypts each and every time the patient sees a physician.
But you don't have to take the DMCB's word for it.
Dr. Faith Fitzgerald chose a far less gruesome allegory. Using Hans Christian Anderson's "The Emperor's New Clothes," this University of California Davis professor wrote a biting essay on the topic of lingering electronic untruths in a recent Annals of Internal Medicine. She describes a patient who, despite having none of the features of Ehlers Danlos, is aggressively treated for it because the condition has been repeatedly listed in the EHR as a diagnosis. Yet, unlike the boy who pointed out the Emperor was naked, Dr. Fitzgerald was unable to overcome her colleagues' unconditional willingness to assume the EHR's information was correct.
"Zombie" and "Emperor's New Clothes" are only two ways to characterize this particular shortcoming of the EHR. db's Medical Rants has a third, describing it as documentation inertia.
Whatever the allegory, fable or jargon, the phenomenon is real. The DMCB did a quick literature search to see if the frequency of untrue electronic diagnoses has been measured, and it can't find any studies.
If readers know of any please share. In the meantime, the DMCB will continue to periodically revisit the issue.
