Earlier quoted context omitted.
Maybe, I haven't read the study, but there is a huge selective pressure here. That is, these days you tend to order autopsies only when there was something mysterious about the death.
> Maybe, I haven't read the study I don't get that. If you haven't read the study, how can you comment on it? I understand that you'd have your opinion canned and ready in the normal case but this study supposedly moves the needle and I'd be very interested on what it does to your position after you read the study, but without that it's as if you are rejecting out of hand any evidence the study might provide.
I can comment based on my observation (lately shared by many) that most of the medical literature is bunk and my experience participating in more than 100 autopsies which tells me that 30% is absurdly high, 15% is too high and the range reported is bizarrely large. In those over 100 autopsies I recall only two surprises. Mostly we learned nothing the clinicians didn't know. Occasionally we found something unknown that wouldn't have altered treatment or course. My experience was that the autopsies were overwhelmingly uninformative.