This study aims to evaluate the practicability and practical advantages of: 1) Targeted Minimum Loss based Estimation (TMLE), and 2) a high-dimensional propensity score (hdPS) algorithm to fit dynamic Marginal Structural Models (MSM) in a longitudinal comparative effectiveness research (CER) study of adults with type 2 diabetes.
Feasibility and Performance Assessments of TMLE and hdPS Methodologies to Fit MSM in a Real-World CER Study
Investigator: Neugebauer, Romain
Funder: Agency for Healthcare Research and Quality