Great question. Depending on which schools you are interested in for Doris Duke, you could potentially get hugely different experiences. Bottomline : Makes more sense to do Doris Duke at a program that you can eventually see yourself matching to. NIH program has prestige, and people who complete it eventually match well but no one is interested in going to the NIH for Rad-Onc. People who do Doris Duke match very well too and they have the added advantage of having an "in" at their Doris Duke institution (obviously not guaranteed though).
There are plenty of research opportunities available at NIH CRTP program, and you have a vast research enterprise available to you, but for Rad-Onc this enterprise will be largely basic science. I wanted a mix of basic science and clinical science and so Doris Duke was much better for me. Doris Duke for Rad Onc is also very variable depending on the program you go to...prior to accepting my Doris Duke fellowship, I actually spoke to the Rad Onc chairmen at each program before accepting one (eventually at UTSW - a program that is on a ridiculously upward trajectory) and it was primarily because of these conversations that I made my decision.
I've spoken with some people who did Rad Onc research as part of these fellowship programs who had very little to show for it by the time they finished (1 poster at meeting) vs others who had a large amount in the ways of multiple papers/oral presentations/etc not to mention a whole year of networking in the community. While I'm sure the intrinsic differences in those people accounted for much of it, a critical feature remains the functionality of the department not just the name.
I'd be happy to talk more specifically about this, just PM me.
