I think a big part of it is that long-term survival rates aren't very high. If you need CPR because of an underlying illness, you're probably actively dying. If you're an older doctor, you probably know it.

CPR has a much better success rate when something like an electric shock stops an otherwise-healthy person's heart.

It's usually a good thing that the heart sits inside a boney cage. Not so much when you're trying to coax it into action.

Yes when I was CPR'd I got broken ribs. It was the defib that brought me round though.

The broken ribs were much more painful than the subsequent ICD implant. Although subsequently meeting the people who broke my ribs was actually a very happy and positive experience.