Real World Health Care Data Analysis

Chapter 7: Stratification for Estimating Causal Treatment Effects

7.1 Introduction

7.2 Propensity Score Stratification

7.2.1 Forming Propensity Score Strata

7.2.2 Estimation of Treatment Effects

7.3 Local Control

7.3.1 Choice of Clustering Method and Optimal Number of Clusters

7.3.2 Confirming that the Estimated Local Effect-Size Distribution Is Not Ignorable

7.4 Stratified Analysis of the PCI15K Data

7.4.1 Propensity Score Stratified Analysis

7.4.2 Local Control Analysis

7.5 Summary

References

7.1 Introduction

Stratification is an intuitive and commonly used approach to adjusting for baseline confounding. In brief, one can remove bias by comparing treatment outcomes within subgroups (strata) of “like” patients and then averaging the results ...