Estimating Decision-Relevant Comparative Effects Using Instrumental Variables

• Activity Number: 384
• Type: Invited
• Date/Time: Tuesday, August 2, 2011 : 2:00 PM to 3:50 PM
• Sponsor: International Chinese Statistical Association
• Author(s): Anirban Basu*+
• Companies: University of Washington
• Keywords: instrumental variables ; heterogeneity ; local effects ; local IV methods ; prostate cancer

Abstract

Instrumental variables methods (IV) are widely used in the health services and biostatistics literature to adjust for hidden selection biases in observational studies when estimating treatment effects. When treatment effects are heterogeneous in the population and when when individuals' self-selected choices of treatments are correlated with expected idiosyncratic gains or losses from treatments, interpretation of IV results becomes challenging. I present an overview of the challenges that arise with IV estimators in the presence of effect heterogeneity and how we can overcome these challenges. I compare conventional IV analysis with alternative approaches that use IVs to estimate treatment effects in models with response heterogeneity and self-selection. Using SEER-Medicare linked data, I apply the method of local instrumental variables to estimate the Average Treatment Effect (ATE) and the Effect on the Treated (TT) on 5-year direct costs of surgery, radiation therapy and watchful waiting among male Medicare beneficiaries (aged 66 or older) with newly diagnosed prostate cancer. Our results reveal some of the advantages and limitations of conventional and alternative IV methods.