Join us on January 29th to kick off our Spring 2025 CTML Seminar Series! Emilie Højbjerre-Frandsen will start our series with her talk "Prognostic Score Adjustment for Marginal Effect Estimation with GLMs." This talk will take place at 12:00PM at Berkeley Way West, 5th Floor, Room 5401.
The study proposes a new method to improve the efficiency of randomized clinical trials by using historical data to train a prognostic model and incorporating the resulting scores as covariates in the analysis of non-continuous outcomes using generalized linear models (GLMs). This generalizes the prognostic score method for linear models (Schuler, 2022). This approach improves the precision of marginal treatment effect estimates and achieves semi-parametric efficiency under the assumption of an additive treatment effect. I will present a simulation study to illustrate the efficiency gains of this method in different data generating processes.