Longitudinal studies in which exposures, confounders and outcomes are measured repeatedly over time have the potential to allow causal inferences about the effects of exposure on outcome. There is particular interest in estimating the causal effects of medical treatments (or other interventions)...