Compare mediation methods and target estimands across longitudinal survival scenarios

Systematically compare existing causal mediation-analysis methods and their associated target estimands for time-dependent mediators, time-dependent confounders, and time-to-event outcomes across distinct longitudinal survival scenarios, in order to identify which method is most appropriate in each scenario.

Background

The paper reviews and evaluates the difference method and the parametric mediational g-formula, while noting that many other methods have been proposed for longitudinal mediation analysis with survival outcomes. The authors state that the relative appropriateness of these methods depends on the scenario and on the target estimand, but that no systematic determination has yet been made.

A neutral comparison study would help establish method-selection guidance, clarify how differing estimand definitions affect conclusions, and support the development of user-friendly software for applied researchers.

References

Although an increasing amount of such methods are being proposed in the literature, it is currently unclear which is the most appropriate in which scenario.

— Causal Mediation Analysis with a Time-Dependent Mediator, Time-Dependent Confounders and a Time-to-Event Outcome: Revisiting the Difference Method  (2608.13094 - Denz et al., 13 Aug 2026) in Discussion, final paragraph of Section 6