Application of Probabilistic Bias-Analysis Methods to Historical Cohort Mortality Studies

Abstract

The purpose of this research was to describe the development and application of probabilistic bias-analysis methods to adjust results from historical cohort mortality studies with multi-level exposure variables. The bias-analysis models illustrated within this dissertation focus on three key sources of systematic study error: 1) disease misclassification due to loss to follow-up, 2) disease misclassification due to incorrect death data, and 3) exposure misclassification.