Design and Pilot Testing of a Resource to Assess Risk of Bias in Observational Studies Estimating Vaccine Effectiveness

Abstract

This master’s thesis addressed three interrelated research objectives aimed at identifying gaps in the assessment of risk-of-bias (RoB) in observational studies of vaccine effectiveness (VE) and proposing a draft framework for evaluation in the next phases of an ongoing research project. First, I mapped existing RoB tools used in systematic reviews of VE studies (Chapter II). The findings revealed that although RoB assessments are routinely conducted, there is considerable variation in the choice of tools used and their application, many of which were not specifically designed to address VE-related biases. Second, I identified and synthesized key bias concepts relevant to VE estimation and created a framework (Chapter III). I found that most bias concepts fell under three broad categories: confounding bias, selection bias, and information bias, though several cross-cutting biases also emerged. Third, I piloted a consensus-based survey designed to evaluate the framework and its components with the goal to assess its readiness, feasibility and appropriateness for international application (Chapter IV). Collectively, this thesis contributed to the early stages of a larger, ongoing research project focused on developing a new RoB resource for assessing biases in observational studies of VE.