Other than not giving it a name, this NYT article gives a great overview of something called diagnostic bias. If we start looking harder for something, we'll see the prevalence increase. If we look harder for diabetes in men, we're going to find more diabetes in men. If we find the same amount of diabetes in men as in women, but we tried to find it in men about twice as hard, chances are there is more diabetes in women than in men.
Just think about the math: if men have a 20% true prevalence and we find 75% of cases in a population of 100 men, we identify 15 cases for an apparent prevalence of 15%. If women have a 30% true prevalence and we find 50% of cases in a population of 100 women, we identify 15 cases for an apparent prevalence of 15%.
Real world example? Ever heard of the diagnosis disparity between men and women with heart disease?
Subscribe to:
Post Comments (Atom)
1 comment:
Which is why - in my field - people look at the increase in diagnoses of ADHD, autism, etc. and think that people are just making up these symptoms and the conditions don't really exist or we must somehow be causing them with video games, too much sugar, etc.
Post a Comment