From a procedural standpoint, the reasoning is easy, the patient are tinier and little mistakes can lead to bigger problems. This certainly bares out in the literature regarding resident procedures. Now one could argue it’s because residents don’t do enough, which is true, but procedures in general are on a downtrend, so you give procedures to trainees mostly likely to need that skill in the future, ie not a general resident who just wants to get that intubation that they will never do again.
From a knowledge base, there should be increased supervision initially that graduates to less supervision overtime. I think most places follow that rule. Now, if by less autonomy, you mean calling an attending at telling them about a patient you just admitted even though you are an upper level, I think that is person depend. Personally, I prefer to be woken up and hear and see all admissions (we have in house call mind you). And I do that for every level of fellowship training. And it’s not because I don’t trust a third year fellow, it’s mostly so A) I know how much clinical and paperwork I have ahead of me for the admission and B) so I don’t walk into any surprises. At the end of the day, it’s my name on the chart for the person most responsible for the patient, not the trainee.
If instead by lack of autonomy one means, calling a consultant and having a consultant for everything, I think that’s just academic medicine. It’s probably over the top, but it is what it is.
If there is one think I think that has decreased provider autonomy overall, it’s that everything is protocolized. You don’t need to think through problems and troubleshoot as much as you used to, you just put them on the protocol. I certainly understand the potential benefits, but also have seen associated harm. Not sure that will change though as that is a cultural thing.