Real World Health Care Data Analysis

Chapter 13: Evaluating the Impact of Unmeasured Confounding in Observational Research

13.1 Introduction

13.2 The Toolbox: A Summary of Available Analytical Methods

13.3 The Best Practice Recommendation

13.4 Example Data Analysis Using the REFLECTIONS Study

13.4.1 Array Approach

13.4.2 Propensity Score Calibration

13.4.3 Rosenbaum-Rubin Sensitivity Analysis

13.4.4 Negative Control

13.4.5 Bayesian Twin Regression Modeling

13.5 Summary

References

13.1 Introduction

In Chapter 2, we introduced the two common statistical frameworks for inferring causal effects from non-randomized observational studies and pointed out the key assumptions needed to ensure the validity of each framework. The two frameworks had certain key assumptions in common, including ...