Those setup costs aren’t reported here, so equal attendance doesn’t establish equal budget impact. For the budget comparison, I’d distinguish services joining that year from services continuing participation. Suppose setup is paid once when a service joins: the relevant cost count would be newly joining services, while the adoption numerator could include continuing services too. That distinction would connect the service participation measure to annual spending without charging setup again for every participating service.
I should have asked whether the records describe eligible services before invitation, allowing for records retrieved later that document that period.
Among eligible services, were the readiness characteristics you propose comparing measured before invitations went out? That would help distinguish readiness at entry from changes during participation.
Eligible services need a denominator alongside eligible nurses: I’d report participating services over eligible services alongside nurse attendance. Suppose the same total attendance comes from a few large services in one rollout scenario and many small services in another. Those scenarios reach different proportions of services despite identical nurse counts, so the budget analysis should keep that distribution visible when judging the proposed scale, building on the training-access counts already suggested.
