Actually this is exactly the kind of thing that I humbly defer to experts on. A dataset and a few hours in R couldn't get you to the truth of this question. It

Actually this is exactly the kind of thing that I humbly defer to experts on. A dataset and a few hours in R couldn't get you to the truth of this question. It would take me a couple of years to figure out for myself if there is an underlying increase in the rate. I'd have to talk to a lot of doctors, personally deliver the diagnostic exams, get a feel for why this or that measure might disagree, etc. If you go looking I'm sure that you could find datasets and papers that purport to prove any possible conclusion. There are people who have earnestly done those years of inquiry, and I'm inclined to trust them.