Help using Base SAS procedures

Marginal Structural Models in PROC GENMOD

Reply
N/A
Posts: 0

Marginal Structural Models in PROC GENMOD

I have read a number of papers on Marginal Structural Models, but am having a difficult time figuring out how to actually implement them in PROC GENMOD.

The reason I am using MSM is that time-dependent confounding is likely occurring in my analysis... my exposure predicts subsequent levels of time-varying covariates of the relationship between exposure and outcome.

I am modeling a continuous outcome, so my understanding is that I want to first perform a linear regression of the exposure including the prior exposure and all covariates.

It's the next step that baffles me. I know in theory what I'm supposed to do (take inverse of density whose mean is equal to the expected value of the exposure predictor model) but I have no idea how to actually do this!

It is inverse density weighting. SAS can apparently do this, but I am at a loss.

Message was edited by: frederickson Message was edited by: frederickson
Ask a Question
Discussion stats
  • 0 replies
  • 369 views
  • 0 likes
  • 1 in conversation