Speaker: Miguel Hernan, Harvard School of Public Health
Title: “Causal inference for Dynamic Treatment Strategies: The renaissance of the g-formula”
Abstract: Causal questions about the comparative effectiveness and safety of health-related interventions are becoming increasingly complex. Decision makers are now often interested in the comparison of interventions that are sustained over time and that may be personalized according to the individuals’ time-evolving characteristics. These dynamic treatment strategies cannot be adequately studied by using conventional analytic methods that were designed to compare “treatment” vs. “no treatment”. The parametric g-formula was developed by Robins in 1986 with the explicit goal of comparing generalized treatment strategies sustained over time. However, despite its theoretical superiority over conventional methods, the parametric g-formula was rarely used for the next 25 years. Rather, the development of causal inference methods for longitudinal data with time-varying treatments focused on semiparametric approaches. In recent years, interest in the parametric g-formula is growing and the number of its applications increasing. This talk will review the parametric g-formula, the conditions for its applicability, its practical advantages and disadvantages compared with semiparametric methods, and several real world implementations for comparative effectiveness research.