SARS-CoV-2 transmission and control: from understanding social contact and mobility to formulating effective vaccine policy Restricted; Files Only

Abstract

Background: Human behavior influences the spread of SARS-CoV-2 both individually and across communities. At both levels, vaccines emerged as an instrumental tool for prevention and control. The goal of this dissertation is to understand individual-level social contact and community-level human movement patterns in the context of SARS-CoV-2 transmission and utilize this knowledge to inform effective vaccine policy.

Aim 1: We estimate the effect of receiving a COVID-19 vaccination on change in individual-level contact rates in a longitudinal cohort sampled from U.S. households. We found that in the context of increasing contact rates over survey rounds, individuals who newly completed primary vaccine series had additional increases in contacts compared to individuals who remained unvaccinated. A mathematical framework integrating competing effects of changing vaccine coverage and contact rates showed that vaccine protection against infection was insufficient to fully offset observed patterns of increase in contact rates, but transmission remained below levels expected under pre-pandemic contact rates.

Aim 2: We infer spatial patterns of transmission across waves of COVID-19 in Georgia, USA through a novel mathematical framework with a multilayered transmission process and informed by spatiotemporally resolved data on social contact, human mobility, and vaccination. We find that in counties with smaller populations, lower contact rates and higher vaccination coverage, intercounty mobility contributes to a higher proportion of onward transmission. In addition, we present evidence that in an interconnected spatial network with a patchwork of local uptake in mitigation measures, the net infection flow is still from counties with lower mitigation to counties with higher mitigation.

Aim 3: We assess the utility of guiding the timing of future COVID-19 re-vaccination strategies with serological surveillance for SARS-CoV-2 in Mozambique over a ten-year horizon. We use a mathematical model informed by local contact rates to simulate using population-level seroprevalence thresholds to trigger the timing of re-vaccination campaigns among older adults and compared this approach to re-vaccination at fixed time intervals. We find that, in this context, serology-triggered vaccination strategies are unlikely to minimize both deaths and the number needed to treat to prevent one death (NNT) compared to fixed time interval strategies.

Potential impact: This dissertation generates valuable insights on transmission dynamics and infection control, weaving together the use of novel behaviorally related data collected during a pandemic and assessing the impact of an innovative temporally targeted vaccination strategy. These insights will guide analysis of behavior data and their incorporation into mathematical models to assess the likely impact of intervention strategies for future outbreaks of infectious diseases. 

Table of Contents

TABLE OF CONTENTS

CHAPTER 1      BACKGROUND.. 1

1.1       COVID-19 history and disease burden. 1

1.2       Natural history of SARS-CoV-2. 1

1.3       Evolving data sources on human behavior relevant to infection transmission. 2

1.4       Individual-level human behavior and infection transmission. 4

1.4.1        Social contact patterns relevant for infection transmission. 4

1.4.2        Behavioral adaptations during outbreaks. 5

1.5       Human movement and spatial patterns of transmission. 6

1.5.1        Human movement and spatial patterns of transmission. 6

1.5.2        Incorporating human movement into mathematical models. 7

1.6       Intervention strategies for SARS-CoV-2 control 7

CHAPTER 2      STUDY RATIONALE AND SPECIFIC AIMS. 9

2.1       Overarching goal 9

2.2       Aim 1 rationale and overview.. 9

2.3       Aim 2 rationale and overview.. 10

2.4       Aim 3 rationale and overview.. 10

CHAPTER 3      THE EFFECT OF COVID-19 VACCINATION ON CHANGE IN CONTACT RATES. 12

3.1       Abstract 12

3.2       Introduction. 14

3.3       Methods. 15

3.3.1        Sampling. 15

3.3.2        Survey data. 16

3.3.3        Latent class model to classify risk tolerance at baseline. 16

3.3.4        Modeling effect of vaccination on contact rate. 17

3.3.5        Estimating the impact of contact change on transmission potential 18

3.4       Results. 19

3.4.1        Participant description. 19

3.4.2        Change in contact rates over time. 20

3.4.3        Vaccination and contact rates. 20

3.4.4        Variation in contact rates by key covariates. 21

3.4.5        Change in individual-level vaccination status over time. 22

3.4.6        Effect of vaccination on change in contact rates. 27

3.4.7        Impact of differential changes in contact rates among vaccinated and unvaccinated on transmission. 30

3.5       Discussion. 32

3.6       Conclusion. 34

3.7       Supplementary File. 35

3.7.1        Comparing distribution of covariates among initially enrolled study population and those completing follow-up. 35

3.7.2        Additional details for Latent Class Analysis (LCA) 37

3.7.3        Exposure classification. 45

3.7.4        Exploring relationship between key covariates. 46

3.7.5        Mean contact rate by time-varying covariates. 50

3.7.6        Effect of vaccination on changes in location-specific contact 55

3.7.7        Sensitivity analysis. 62

3.7.8        Additional Next Generation Matrix analysis on impact on transmission. 67

CHAPTER 4      METAPOPULATION MODEL TO QUANTIFY TRANSMISSION  69

4.1       Abstract 69

4.2       Introduction. 69

4.3       Methods. 71

4.3.1        Model structure. 71

4.3.2        Infection process. 73

4.3.3        Parameterizing between-county transmission. 74

4.3.4        Key model inputs for transmission process. 74

4.3.5        Model initialization. 79

4.3.6        Model calibration. 80

4.4       Results. 81

4.4.1        Results from model calibration. 81

4.4.2        Proportion of SARS-CoV-2 infections imported through intercounty mobility. 83

4.4.3        Directionality of infection flow.. 86

4.5       Discussion. 88

4.6       Conclusion. 90

4.7       Supplementary File. 91

4.7.1        Detailed model structure and equations. 91

4.7.2        Additional data descriptions. 94

4.7.3        Additional model calibration details. 95

4.7.4        Proportion infections from intercounty mobility. 102

4.7.5        Scatter plots of difference in pairwise county attributes and differences in importations and trips. 108

4.7.6        Exploration of using state-wide seroprevalence data to inform age-specific reporting rate. 111

CHAPTER 5      MODELING THE USE OF SEROPREVALENCE TO GUIDE COVID-19 VACCINATION IN MOZAMBIQUE. 117

5.1       Abstract 117

5.2       Background. 118

5.3       Methods. 120

5.3.1        Model structure. 120

5.3.2        Seroconversion and seroreversion. 122

5.3.3        Tiered susceptibility. 123

5.3.4        Data sources and calibration. 123

5.3.5        Forward simulation epidemiological scenarios. 125

5.3.6        Vaccination triggers and analytical outputs. 126

5.3.7        Code availability. 127

5.4       Results. 127

5.4.1        Model calibration. 127

5.4.2        Description of simulated epidemic and changing immunity. 128

5.4.3        Descriptive results from re-vaccination strategies. 129

5.4.4        Impact of different re-vaccination strategy on vaccine NNV.. 130

5.4.5        Tradeoffs in number-needed-to-vaccinate (NNV) 132

5.4.6        Sensitivity analysis. 134

5.5       Discussion. 135

5.6       Conclusion. 138

5.7       Supplementary File. 139

5.7.1        Additional model methodology. 139

5.7.2        Model parameters. 145

5.7.3        Data sources from Mozambique. 149

5.7.4        Calibration results. 154

5.7.5        Assessing correlations between seroprevalence, susceptibility and cumulative deaths in base scenarios with no vaccination. 156

5.7.6        Vaccination impact results for main analysis. 161

5.7.7        Sensitivity analysis. 165

5.7.8        Summary of literature review of key parameters. 177

CHAPTER 6      CONCLUSIONS AND PUBLIC HEALTH IMPLICATIONS 183

6.1       Overview.. 183

6.2       Contributions and future directions of each specific aim.. 183

6.2.1        Aim 1. 183

6.2.2        Aim 2. 186

6.2.3        Aim 3. 188

6.3       Reflections. 190

CHAPTER 7      REFERENCES. 193

CHAPTER 8      APPENDIX.. 222

8.1       Abbreviations. 222

8.2       Publications, presentations, and funding-related activities. 223

8.2.1        Peer Reviewed Publications. 224

8.2.2        Presentations. 225

8.2.3        Grants. 227

About this Dissertation

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	Laney Graduate School
Department	Epidemiology
Degree	Ph.D.
Submission	Dissertation
Language	English
Research Field	Health Sciences, Epidemiology Health Sciences, Public Health
Palavra-chave	Infectious disease epidemiology Mathematical modeling SARS-CoV-2
Committee Chair / Thesis Advisor	Benjamin A Lopman, Emory University
Committee Members	Kristin Nelson, Emory University Stefan Flasche, London School of Hygiene and Tropical Medicine Max SY Lau, Emory University Samuel Jenness, Emory University

Última modificação

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	File download under embargo until 22 May 2025	2024-04-23 20:56:41 -0400	File download under embargo until 22 May 2025

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions