On the identification of time-varying treatment effects and ignorability
Vanessa Didelez, Leibniz Institute for Prevention Research and Epidemiology - BIPS, Bremen
The talk will address and discuss some parallels between the identification of causal effects of time-varying
treatments and the concept of ignorability in longitudinal data. We consider discrete as well as continuous
time situations. A graphical condition, stability, plays a role analogous to that of missingness at random,
but is applicable to general longitudinal data and mirrors the corresponding concept in causal inference.
Dawid, Didelez (2010). Identifying the consequences of dynamic treatment strategies: A decision theoretic overview. Statistics Surveys, 4, 184-231.
Didelez (2015). Causal Reasoning for events in continuous time: a decision–theoretic approach, Proceedings of the 31st Annual Conference on Uncertainty in Artifical Intelligence - Causality Workshop.
Farewell, Huang, Didelez (2017). Ignorability for general longitudinal data. Biometrika, 104(2), 317-326.
Røysland, Didelez, Nygard, Lange, Aalen (2017). Causal reasoning in survival analysis: reweighting and local independence graphs. In preparation.