The tasks seem almost absurd, like a game bored siblings might play to while away hours in the backseat of a car. A doctor might ask you to name as many animals as possible over the course of a minute, or to count backwards by sevens, or to hold random words in your mind while your attention is drawn somewhere else. Your answers could bring about the worrisome news that you need further testing for dementia.
It’s widely known that these quick screening tests — some of them technically copyrighted but often easily downloadable for free — aren’t always right. Some patients might pass the animal-naming test with flying colors but still have dementia; others might be sent for hours of in-depth evaluation that they don’t need. But what predisposes a patient to get a misleading result turns out to be very different form one screening method to another, according to a paper published Wednesday in Neurology Clinical Practice.