The Pareto distribution (20% of the people doing 80% of the work) is found everywhere. Unfortunately as Nick Z has shown now the work and patients are getting concentrated in not just the top-tier people but the top-tier and top-volume centers. "Admin and research" can never be (much of) a thing in the community. There is a HUGE chasm between the wants/needs/reimbursement of a system like yours and small clinics and small community or regional hospitals.
But again just go back to patient numbers. Take cFTE out of it. MSKCC had 4031 new starts in 2013 and 4173 in 2020. Did MSKCC Manhattan increase MDs between 2013 and 2020? And if it's true that America as a country had less new RT starts in 2020 vs 2003, isn't that worthy for some discussion?* Now factor in virtual supervision, APM (will only affect centers where MDs *don't* do research at first), declining indications (we just lost postop lung N2 and are losing chemoRT in Stage III rapidly as well; I have a new thing I'm harping on it seems almost every week), declining fractionation, and horrible geographic mobility...
Sometimes your denominator is just 1.0 cFTEs (as if a sub-1 sees zero patients, but you're a non-1.0 cFTE and see as much as a 1.0 cFTE) and sometimes not. In general, we have to assume every non-academic RO is a 1.0 cFTE, and I also believe we must assume every RO trainee is "at risk" of being a 1.0 cFTE. As your own example shows, the association between the cFTE metric and actual work may not be strong. For your purposes as chairman, of course you have to get bogged down in FTE minutiae. For our purposes of zooming out and looking at the big picture, we need to look at such things a little less IMHO. Keep in mind the economics where you assign all the work to 2/3 of your docs and zero to the other third ("We treat ~3700 new patients per year and have 15 docs at ~9 cFTE = 411 new starts per 1.0 cFTE"), whether in your head or IRL, doesn't work outside the walls where super high reimbursements let these kinds of economics work. The IQR for Medicare reimbursement per RO physician shrunk from $341K to $245K from 2013 to 2019. Nationwide, *at least* 30% of rad onc's patient load is Medicare.
If nationwide every rad onc was carrying a Dan Spratt workload (330 pts/yr), consider the math:
1) Of the ~1.9m new cancer cases per year,
~31% (and falling) need RT. That's 589K people, and 589K divided by 330 is 1768 (rad oncs per 330 pts/RO). We have 5500+ rad oncs in America.
2) Of the ~1.1m new RT starts per year in the US, that number divided by 330 is 3333. Again, we have 5500+ rad oncs in America. We can't assume that 2000+ rad oncs in America either want or need admin or research work. (Proton centers don't even know how to proforma admin and research work!)
250 to 400 is a wiiiiide gap! We need to have much more precise and accurate data to have really fruitful discussions about the rad onc workforce and trends over time.
Read a CBO report. Or Johnson & Johnson annual report. Or even an IPCC report. They always start with the big picture and the simple math first. Then they get more detailed and complicated from there. Start with easy and simple first. Remember how as a PGY2 we always wanted to make things more complicated than they really were? (And later you learn, yeah, they're complicated... but you have to see the forest before the trees.)
*Patient volume in the non-Manhattan MSKCC Network (has it grown in terms of sites? it must have?) has increased a lot though.