B1396
Title: Sensitivity analysis via the proportion of unmeasured confounding
Authors: Matteo Bonvini - Carnegie Mellon University (United States) [presenting]
Edward Kennedy - Carnegie Mellon University (United States)
Abstract: In observational studies, identification of average treatment effects is generally achieved by assuming no unmeasured confounding, possibly after conditioning on enough covariates. Because this assumption is both strong and untestable, a sensitivity analysis should be performed. Widely used approaches include modeling the bias directly or varying the propensity scores to probe the effects of a potential unmeasured confounder. We take a novel approach whereby the sensitivity parameter is the proportion of unmeasured confounding. We consider several scenarios imposing different assumptions on the probability of a unit being unconfounded. In each case, we derive sharp bounds on the average treatment effect as a function of the sensitivity parameter and propose nonparametric estimators. We introduce a one-number summary of a study's robustness to the number of confounded units. Finally, we explore finite-sample properties via simulation, and apply the methods to an observational database used to estimate the effects of Right Heart Catheterization (RHC) on the care of critically ill patients.
