rdrobust denominator for rates

[Eran Bellin]

I have events and five year age ranges that allow calculation of rates per 10,000. I have a breakpoint at a specific age where the intervention is to be applied. I would like to build a regression discontinuity analysis but when I look at rdrobust, there is no place to enter the denominator for each of the rates. It would seem that the variance should be sensitive to the actual denominator but the program has no place for it. While it has weight, this weight is only to enforce a relative weighting in proximity to the cutpoint (I think). If I were to build a poisson regression with a discontinuity and robust vce, I would get coefficients and confidence intervals for the discontinuity and for the slope change, but the ability to retrieve the actual value of the rate for a proposed line for the comparison of pre and post discontinuity would be dependent on how well the poisson model distribution describes the phenomenon and this is often bad and not really improved with negative binomial models. How do I get