North, the answer is in the last phrase. 0.5 patients per centre per month.
I have been using that since many months now, it isn't anything that is published by PPHM (not even the treatment centres because I get them from clinicaltrials.gov like everybody I assume.)
The difference is that I have put them in a list with exact dates (as I posted before) and therefore could tune the simulation rather then looking at the total # and apply some enrolment factor. I am currently fine-tuning it more based on something I realized based on the ASM presentation (while that information was out since the beginning on clinicaltrial.gov but I never stood still on it). More about that later if I manage to include it in the model.
PS: If I post something based on links or other information I found then I always mention the source. I think you know that.
If one day I wouldn't then asking me is pointless because it probably means I don't intend to share it. An example, I never shared the complete detailed list of all treatment centres that I have just to respect PPHMs which that those centres are left alone. Otherwise PPHM could as well themselves published the names in stead of the zip codes.