Causal inferences from complex observational and randomized studies with time-varying exposures or treatments.

I review identification and estimation of direct and indirect effects of time-varying treatments or actions. I describe the relationship between a number of modeling approaches: marginal structural mo