About the speaker
Jon Michael Gran is associate professor at University of Oslo at the department of Biostatistics. His research focus is on causal inference and survival analysis.
About the presentation
When comparing time varying exposures in a non-randomized setting, one must often correct for time-dependent confounders that influence treatment choice over time and that are themselves influenced by treatment. In a JRSS-C paper from 2018 we suggested a new two-step procedure, based on additive hazard regression and linear increments models, for handling such confounding when estimating average treatment effects on the treated. The method was applied to data from the Swiss HIV Cohort Study, estimating the effect of antiretroviral treatment on time to acquired immune deficiency syndrome or death, and in a simulation study. The method is easy to implement by using available software packages in R and can also be used for mediation analysis.
In this talk I will give a brief introduction to this particular method and, more generally, discuss the definition and estimation of various causal target estimands in situations with time-dependent confounding (that is; the treatment effect on the treated and other alternatives).