Bias in Vaccine Effectiveness Estimation Under Changing Testing Probabilities When Using the Test-Negative Design Open Access

Abstract

Background: In recent years, the test negative design (TND) has emerged as a popular study design in estimating vaccine effectiveness (VE) for COVID-19 vaccines. While a useful and important tool, this design has major considerations that must be made before using it to estimate VE. It is vital to understand how simple changes in test reporting or differences in added protection in disease severity levels can lead to a change in VE estimates.

Objectives: We would like to characterize how sensitive VE estimates from a test-negative design are to changes in test reporting, and how it varies by disease severity and vaccine protection at different levels of disease severity.

Methods: We used a simulation to generate COVID-19 test results and demographic data and study this relationship. In particular, we looked at different patterns of testing probabilities and vaccine protection and how they would bias VE estimates in five different settings.

Results: Vaccine effectiveness for protection against infection has been found to be biased by as much as 8.64% (an estimated VES of 78.6% and a true value of 70%) in settings where testing probabilities were differential by disease severity and there was added vaccine protection against severe cases like hospitalization.

Conclusions: VES is often biased and overestimated when testing probabilities are differential by disease severity, and this is exacerbated by added vaccine protection against severe outcomes. These overestimations were due to a much greater fraction of hospitalized or severe cases represented in the test results compared to those who were mildly symptomatic or asymptomatic.

Table of Contents

Introduction..........1

Methods................5

Results................15

Discussion...........21

References...........24

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Biostatistics
Subfield / Discipline	Biostatistics - MPH & MSPH
Degree	M.S.P.H.
Submission	Master's Thesis
Language	English
Research Field	Biology, Biostatistics
Keyword	Vaccine Effectiveness COVID=19 Test Negative Design
Committee Chair / Thesis Advisor	Natalie Dean, Emory University
Committee Members	Avnika Amin, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Bias in Vaccine Effectiveness Estimation Under Changing Testing Probabilities When Using the Test-Negative Design ()	2023-04-09 21:46:32 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions