May 3, 2016
May 6: A General Framework for Evaluating Bias in Two-stage Instrumental Variable Models
VISITING FACULTY SEMINAR
May 6, 2016
2:30p.m. – 3:30p.m.
Fei Wan, Ph.D.
Assistant Investigator, Biostatistics Unit
Group Health Research Institute, Seattle WA
A General Framework for Evaluating Bias in Two-stage Instrumental Variable Models
Unmeasured confounding is a common concern when researchers attempt to estimate a treatment effect using observational data or randomized studies with non-perfect compliance. To address this concern, instrumental variable (IV) methods, such as two-stage predictor substitution (2SPS) and two-stage residual inclusion (2SRI), have been widely adopted. In many clinical studies of binary and survival outcomes, 2SRI has been accepted as the method of choice over 2SPS but a compelling theoretical rationale has not been postulated. We propose a novel two stage modeling framework to understanding the bias and consistency in estimating the conditional treatment effect for 2SPS and 2SRI when the outcome is binary, count or time to event. Under this framework, we demonstrate that the bias in 2SRI estimators can be reframed to mirror the problem of omitted variables in non-linear models and that there is a direct relationship with the collapsibility of effect measures. We demonstrate that only if the influence of the unmeasured covariates on the treatment is proportional to their effect on the outcome, then 2SRI estimates are generally unbiased for logit and Cox models. We also propose a novel dissimilarity metric to quantify the difference in these effects and demonstrate that with increasing dissimilarity between the effects of the unmeasured covariates on the treatment versus outcome, the bias of 2SRI increases in magnitude.