Vaccine Communication Online: Counteracting Misinformation, Rumors and Lies 3031244893, 9783031244896

Citation preview

Vaccine Communication Online Counteracting Misinformation, Rumors and Lies Edited by  Tamar Ginossar · Sayyed Fawad Ali Shah David Weiss

Vaccine Communication Online

Tamar Ginossar Sayyed Fawad Ali Shah  •  David Weiss Editors

Vaccine Communication Online Counteracting Misinformation, Rumors and Lies

Editors Tamar Ginossar Department of Communication and Journalism University of New Mexico Albuquerque, NM, USA

Sayyed Fawad Ali Shah School of Communication and Journalism Auburn University Auburn, AL, USA

David Weiss Department of Communication and Journalism University of New Mexico Albuquerque, NM, USA

ISBN 978-3-031-24489-6    ISBN 978-3-031-24490-2 (eBook) https://doi.org/10.1007/978-3-031-24490-2 © The Editor(s) (if applicable) and The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover illustration: Joshua Windsor / Alamy Stock Photo This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Contents

1 Introduction  1 Tamar Ginossar, Sayyed Fawad Ali Shah, and David Weiss 2 Vaccine  Misinformation on Social Media: Historical Contexts, Lessons Learned, and Paths Forward 11 Beth L. Hoffman, Jaime E. Sidani, Jessica G. Burke, Kar-Hai Chu, and Elizabeth M. Felter 3 HPV  Vaccine Misinformation Online: A Narrative Scoping Review 35 Yuan Wang, Kathryn Thier, and Xiaoli Nan 4 Analyzing  Social-Cyber Maneuvers for Spreading COVID-19 Pro- and Anti- Vaccine Information 57 Janice T. Blane, Lynnette Hui Xian Ng, and Kathleen M. Carley 5 Vaccine  Support and Hesitancy on Twitter: Opposing Views, Similar Strategies, and the Mixed Impact of Conspiracy Theories 81 Itai Himelboim, Jeonghyun Janice Lee, Michael A. Cacciatore, Sungsu Kim, Diane Krause, Kate Miller-Bains, Kristin Mattson, and Jennifer Reynolds

v

vi 

Contents

6 Online  Foreign Propaganda Campaigns and Vaccine Misinformation: A Comparative Analysis103 Dror Walter and Yotam Ophir 7 Online  Public Outreach to Promote Public Health: Insights from Israeli Non-­Governmental Organizations125 Erez S. Garty, Noa Hirsch-Choritz, Keren Landsman, Uri Lerner, Itamar M. Netzer, and Aviv J. Sharon 8 From  Polio to Covid-19: Anti-Vaccine Misinformation and Rumors in Pakistan147 Zia Ullah, Shah Sawar Khan, and Sayyed Fawad Ali Shah 9 Promoting  Dialogue by Thinking Differently about Framing and Correcting Misinformation163 Anat Gesser-Edelsburg 10 Conclusion191 Tamar Ginossar, Sayyed Fawad Ali Shah, and David Weiss Index197

Notes on Contributors

Janice T. Blane  received her PhD from the School of Computer Science at Carnegie Mellon University. She is a member of the Center for Computational Analysis of Social and Organizational Systems (CASOS). She holds a BS from the United States Military Academy in Electrical Engineering and an MS from the University of Illinois at Urbana— Champaign in Electrical and Computer Engineering. Janice served as an assistant professor at West Point and is an operations research and systems analyst for the US Army. Her current research focuses on social cybersecurity and the spread of disinformation on social networks. Jessica  G.  Burke,  PhD, MHS, is the vice dean and a professor in the Department of Behavioral and Community Health Sciences at the University of Pittsburgh School of Public Health. She is passionate about using engaged, creative, and systems-oriented methods to address determinants of health and methods that include visual components that facilitate dialogue and action. Dr. Burke’s scholarship uses innovative quantitative and qualitative methods to explore systems and the mechanisms linking context influences and health, engages communities in the process of research and its translation, and develops tailored interventions. The book Methods for Community Public Health Research: Integrated and Engaged Approaches that she co-edited provides details about this agenda. Michael A. Cacciatore,  PhD, is a co-director, Center for Health & Risk Communication. He is also Associate Professor of Public Relations at Grady College of Journalism and Mass Communication, University of

vii

viii 

NOTES ON CONTRIBUTORS

Goergia. Dr. Cacciatore teaches research methodology and introduction to public relations in the Department of Advertising & Public Relations. His research focuses on science and risk communication with an emphasis on media coverage of and opinion formation for such topics. Dr. Cacciatore’s research has examined the communication of science and risk topics ranging from nanotechnology to food safety to global climate change. A significant portion of this research has tracked media depictions of science and risk issues, paying particular attention to the role of social media in the communication process. His other research has focused most directly on the interplay between media, values, and risk in public opinion formation. Dr. Cacciatore’s work has been published in Public Understanding of Science, Science Communication, Risk Analysis, New Media & Society, and Health Affairs, among others. Kathleen  M.  Carley,  Ph.D. is Professor of Computers and Society in the School of Computer Science; director of the Center for Computational Analysis of Social and Organizational Systems (CASOS); director of the center for Informed DEmocracy And Social-cybersecurity (IDeaS) at Carnegie Mellon University; IEEE Fellow; and CEO of Netanomics. She holds two SB degrees from the Massachusetts Institute of Technology, a PhD from Harvard University, and an HD from the University of Zurich. She is the recipient of the USGA Academic Award at GEOINT 2018 for her work on geo-spatially enabled dynamic network analytics, the Allen Newell award for research excellence, the Lifetime Achievement Award from the Sociology and Computers Section of the ASA (2001), and the Simmel Award for advances in social networks from INSNA (2011). Her research combines social science and computer science to address complex social and organizational issues. Her pioneering research led to the areas of computational social science, dynamic network analysis, and social cybersecurity, computational linguistics tools (AutoMap & NetMapper), agent-based dynamic-network simulation frameworks (Construct), and a network analysis and visualization technology for spatial, dynamic, and high-dimensional network data (ORA-PRO,ORA-WEB). She has over 400 publications and has served on multiple National Academies panels. Kar-Hai Chu,  PhD, MS, is Associate Professor of Public Health at the University of Pittsburgh. His interdisciplinary training focused on applying technology and social network analysis to different aspects of interactive online education. He has studied various public health issues, including tobacco control, community health coalitions, and social-media-based

  NOTES ON CONTRIBUTORS 

ix

health surveillance. His current research is focused on innovative methods of using online technologies to support health science, including studies on school-based-e-cigarette prevention programs, and using Twitter to monitor tobacco and vaccine-related cancer communication. Elizabeth  M.  Felter,  DrPH, MCHES, is an assistant professor in the Department of Behavioral and Community Health Sciences in the School of Public Health at the University of Pittsburgh. Dr. Felter is also an affiliated faculty in the Evaluation Institute for Public Health and the Center for Health Equity. Her teaching and research foci are in the areas of health and risk communication and evaluation. She has been a Certified Health Education Specialist (CHES) since 2001 and a Master Certified Health Education Specialist (MCHES) since 2011. Erez S. Garty,  PhD, is head of Science Communications at the Davidson Institute of Science Education, Israel. Anat Gesser-Edelsburg,  PhD, is an associate professor (tenured); the head of Health Promotion Program, School of Public Health, Faculty of Social Welfare and Health Sciences; and the founding director of the Health and Risk Communication Research Center at University of Haifa. During 2020,  Anat was a visiting scholar at the School of Public Health, University of Illinois at Chicago. She is associate editor of Disaster Medicine and Public Health Preparedness, BMC Public Health, and academic editor of PLOS ONE. Her most recent book is Risk Communication and Infectious Diseases in an Age of Digital Media. Anat is also a researcher at The Center for Evaluation of Health Promotion Interventions and at The Emili Sagol Creative Arts Therapies Research Center, University of Haifa. Anat has won or collaborated in many research grants, among them TellMe  and  Asset, two European Commission funded research grants, and she has published extensively in peer-reviewed journals. Her areas of research include health  and  risk communication, positive deviance, social marketing, persuasive communication, health-promotion programs, entertainment-education and qualitative research. Anat investigates a variety of health-related issues, including emerging infectious disease (EID) communication, vaccination compliance, drugs  and  alcohol abuse, drinking-and-driving, sex education, nutrition, and hospital-­acquired infections prevention. Tamar  Ginossar  is a professor in the Department of Communication and Journalism and the BA/MD Program, and the director of the Institute

x 

NOTES ON CONTRIBUTORS

for Social Research at the University of New Mexico (UNM). As a health, science, and environmental communication scholar, her research focuses on health information behavior and advocacy in diverse online, cultural, geographical, and health contexts including infectious diseases. Her work has been published in interdisciplinary journals and books, including AIDS Care, Health Communication, Health Education and Behavior, the International Journal of Environmental Research and Public Health, the Journal of Computer-Mediated Communication, the Journal of Medical Internet Research, Topics in Antiviral Medicine and Vaccine. She is currently working with interdisciplinary teams on examining communication about vaccination over social media using computational methods. This effort is funded by the UNM award for women in science, technology, engineering, and mathematics (STEM) and the Ohio State University Translational Data Analytics Institute. She is teaching health communication and mixed-methods research courses at the undergraduate and graduate levels and is known for designing new and innovative courses, such as Cannabis and Communication and mHealth, as well as public speaking for STEM students. For these efforts, she received the UNM Presidential Teaching Award in recognition of her teaching excellence. Itai  Himelboim,  PhD, is the Thomas C.  Dowden Professor of Media Analytics and director of the SEE Suite, at Grady College of Journalism and Mass Communication, University of Georgia. Dr. Himelboim studies the role social media plays in news, politics, and international communication. Through applying network analysis, he examines political talk and information flow. Dr. Himelboim’s research involves computer-mediated social networks and their implications for political communication, international communication, and the news.  He examines political discussions on online forums and the micro-blogging social media Twitter, as well as international networks, based on foreign news reporting and the flow of information technologies.  The role of news media, traditional and online, in political communication is examined via their websites and their presence in social media spaces. His research has been published in top journals in the field, including Journalism and Mass Communication Quarterly, Journal of Broadcasting and Electronic Media, Communication Research, Journal of ComputerMediated Communication (JCMC) and Journal of Public Relations Research.  These and other studies have been presented in major national and international conferences in the field.

  NOTES ON CONTRIBUTORS 

xi

Noa Hirsch-Choritz,  PhD, MBA, is an emergency and OBGYN nurse and the regional director of Terem Emergency Medical clinics in Israel. Beth  L.  Hoffman,  PhD, MPH, is a postdoctoral associate in the Department of Behavioral and Community Health Sciences at the University of Pittsburgh School of Public Health. Her research primarily focuses on the public health implications of the portrayal of health topics on popular television programs and the spread of health information and misinformation on social media. Her research examining the anti-vaccine movement on social media has been featured by multiple news outlets, including CNN, The New York Times, and a February 2020 documentary on the Hulu streaming service. Shah Sawar Khan,  MS, is an independent researcher. His areas of interest include health communication, journalism, and strategic communication. Sungsu  Kim,  PhD, is an assistant professor in the School of Communication, Kookmin University, South Korea. Diane Krause,  MS-MPH, CPH, RDN, is Health Education Specialist, Oak Ridge Associated Universities. Keren Landsman,  MD, MPH, is a medical doctor specializing in epidemiology and head of the non-profit public health organization Mida’at— For Informed Health, RA. Jeonghyun  Janice  Lee,  PhD, is an assistant professor in the Manship School of Mass Communication, Louisiana State University. Uri Lerner,  PhD, is a project manager at Maccabi Healthcare Services. Kristin Mattson,  MPH, MCHES, is Health Education Specialist Project Manager, Oak Ridge Associated Universities. Kate Miller-Bains,  PhD, is currently working as a Research Scientist at University of Virginia. She is a quantitative methodologist focused on using rigorous research design to improve equity and access in education. She is interested in how to best form, sustain, and leverage partnerships that bring together the expertise of researchers, educators, and policymakers and allow stakeholders to answer policy-­relevant questions.

xii 

NOTES ON CONTRIBUTORS

Xiaoli  Nan,  PhD, is Distinguished Scholar-Teacher and Professor of Communication Science at the University of Maryland-College Park, where she is the director of the Center for Health and Risk Communication. Dr. Nan is an affiliate professor in the Department of Behavioral and Community Health, a faculty associate of the Joint Institute for Food Safety and Applied Nutrition, and a full member of the Marlene and Stewart Greenebaum Comprehensive Cancer Center’s Population Science Program. Dr. Nan’s research is broadly concerned with health and risk communication, focusing on (a) the design of persuasive messages to influence health risk perceptions and behaviors and (b) the role of traditional and emerging media (e.g., social media, mobile media, virtual reality) in promoting (and hindering) public health. Dr. Nan’s research addresses the basic processes of human judgment and decision making and the implications of these processes for effective health and risk communication. Dr. Nan’s interdisciplinary work tackles pressing public health challenges including cancer prevention, vaccination, food safety and nutrition, and climate change. At Maryland, Dr. Nan regularly teaches courses on health communication, persuasion and attitude change, media effects, and quantitative research methods. Dr. Nan has published extensively in her areas of specialization with over 70 peer-reviewed journal articles. Itamar M. Netzer,  MD, MBA, is currently working as director of the Netanyah-­Taybe region in Sharon Shomron district, Clalit Health Services, and directs innovation in the district. Lynnette Hui Xian Ng  is currently working toward the PhD degree in societal computing with Carnegie Mellon University, Pittsburgh, PA, USA. As a Graduate Researcher with the Center for Informed Democracy, her research examines social cybersecurity and digital disinformation. She received the bachelor’s (Hons.) degree in computer science from National University of Singapore, Singapore. This work was done while she was an AI Engineer with Defence Science and Technology Agency, Singapore. Yotam  Ophir,  PhD (2018, University of Pennsylvania), is Assistant Professor of Communication at the University at Buffalo, State University of New York. Ophir studies political, science, and health communication, misinformation, persuasion, and media effects. Ophir authored and coauthored over 20 articles which were published in many journals including Journal of Communication, Political Communication, Communication Methods & Measures, Journalism, Nicotine & Tobacco Research, and

  NOTES ON CONTRIBUTORS 

xiii

American Journal of Public Health, as well as multiple book chapters for edited volumes. His research combines novel computational tools for automated content analysis, such as machine learning, topic modeling, and network analysis, with experimental and survey designs, used together to study media content and its effects on audiences. Jennifer  Reynolds,  MPH, CHES, is associate manager of Health Communications and Marketing, Oak Ridge Associated Universities. Sayyed  Fawad  Ali  Shah is an assistant professor in the School of Communication and Journalism at Auburn University. He is also chair of the Health Communication Working Group of the American Public Health Association. He received an MA in Journalism and Mass Communication from the University of Peshawar, Pakistan, and a PhD in Communication from the University of New Mexico. His primary research focuses on journalistic practices across cultural contexts, the role of health journalism in reducing health disparities, and health communication. He is a leading journalism and health researcher. He is currently working with health journalists in Pakistan to explore the factors that influence their role conception and role performance. Additional projects include leading research that aims to identify factors associated with vaccine acceptance/ hesitancy among Pakistan’s Pakhtun ethnic population—marked as a high-risk group in the polio eradication campaigns—and to improve vaccine uptake through communication interventions among this population. The project is funded by the National Communication Association’s Research Cultivation Grant. Moreover, he is also spearheading a study that aims at improving the mental wellbeing of local journalists in the US Southeast. His work has been published in multi-disciplinary journals such as Vaccine, Journalism, Health Education & Behavior and the Journal of Medical Internet Research. Aviv  J.  Sharon,  PhD, is on the Faculty of Education in Science and Technology, Technion—Israel Institute of Technology, Haifa, Israel. Jaime  E.  Sidani,  PhD, MPH, CHES, is an assistant professor in the Department of Behavioral and Community Health Sciences at the University of Pittsburgh School of Public Health. A trained health educator, her research has focused on the epidemiology and prevention of emerging tobacco products, associations between media exposure and health, and the presence and spread of health-related information and

xiv 

NOTES ON CONTRIBUTORS

misinformation on social media. She has presented extensively on COVID-19 misinformation and its potential impacts on vaccine uptake. Kathryn Thier  obtained her BA in History from Brown University and her MS in Journalism from Columbia University. She is currently pursuing a PhD in Communication Science and Social Cognition. Before coming to the University of Maryland, she taught journalism and public relations for five years at the University of Oregon, where she co-founded The Catalyst Journalism Project, a research and teaching collaborative focused on the intersection of solutions and investigative journalism. Kathryn’s research interests include media effects of solutions journalism and health communication. Zia Ullah,  MPhil, is Lecturer in Journalism and Mass Communication at Abdul Wali Khan University Mardan, Pakistan. He is currently pursuing his PhD in Communication, Media and Cultural Studies at Solent University, UK. Dror  Walter,  PhD, is an assistant professor in the Department of Communication, Georgia State University. Dr. Walter’s research is centered on the intersection between traditional media effects theories and novel computational social science methods. His research addresses the ways computational methods such as network analysis, unsupervised machine learning, and supervised machine learning can aid in the analysis of online politically relevant content. Specifically, his past and current work is often situated within the field of political communication with projects such as the conceptualization, measurement and impact of thematic diversity, strategies of political candidates on social media, impact of news framing on candidates’ electoral success, and inductive approaches to nation branding. Additionally, he also studies extremist forms of political discussion focusing and political/health misinformation. Yuan Wang  is a PhD candidate in the Department of Communication with a specialization in Communication Science and Social Cognition. Her research areas include health and risk communication. Yuan is particularly interested in designing interventions to mitigate the negative impact of misinformation and exploring how publics seek, process, and share information about contentious issues in the emerging media environment. Prior to attending the University of Maryland, she received a BA in Journalism and a BA in Law from Renmin University of China, and she received an MPhil in Communication from The Chinese University of Hong Kong.

  Notes on Contributors 

xv

David Weiss  is the director of Liberal Arts and Integrative Studies at the University of New Mexico and an associate professor and the associate chair of UNM’s Department of Communication and Journalism. His teaching and research interests include media studies and mediatization, political communication, popular culture, strategic communication, and their increasing intersections with health and science communication. As a critical scholar of the discourse, structures, and societal impact of the media, he critically investigates the roles played by media in society and the complex ways that those pursuits play themselves out. These content areas might best be described as “culture war” issues: mediated and other publicly communicated messages or texts that are located at the points where media, language, and popular culture intersect with the most powerful issues and institutions of our time: religion, politics, law, public health, sex and sexuality, gender, and/or race and ethnicity. His work has been published in venues including Vaccine, the Journal of Popular Television, and the Journal of Magazine and New Media Research.

List of Figures

Fig. 4.1 Fig. 4.2 Fig. 4.3 Fig. 4.4 Fig. 4.5 Fig. 4.6 Fig. 4.7 Fig. 4.8 Fig. 4.9 Fig. 4.10 Fig. 5.1 Fig. 5.2 Fig. 5.3 Fig. 5.4 Fig. 6.1

Social-Cyber Maneuver Analysis Pipeline (Blane et al., 2022) 62 Actor Types in Top 100 Superspreaders 66 Pro-Vaccine Hashtag Co-occurrence Network (June 2020) 67 Anti-Vaccine Hashtag Co-occurrence Network (June 2020) 68 Pro-Vaccine Hashtag Co-occurrence Network (June 2021) 68 Anti-Vaccine Hashtag Co-occurrence Network (June 2021) 69 Agents Conducting Explain Maneuvers Over Time by Stance 71 Agents Conducting Dismay Maneuvers Over Time by Stance 72 Agents Conducting Back Maneuvers Over Time by Stance 73 Agents Conducting Neutralize Maneuvers Over Time by Stance 73 Comparisons of thematic messaging strategy based on vaccine stance (supportive vs. hesitant) 89 Comparisons between thematic messaging strategies based on vaccine stance and vaccine type 90 The interaction effect of vaccine-supportive vs. vaccinehesitant messages expressing a conspiracy theory thematic messaging strategy 92 The interaction effect of vaccine type messages expressing a conspiracy theories strategy 94 Overall volume of activity on Twitter for each of the ten countries’ operations (top pane); volume of vaccine-related activity on Twitter for each of the ten countries’ operations (middle pane); and the number of vaccine related tweets as a percentage of total tweets for each of the ten countries’ operations (bottom pane)111

xvii

xviii 

List of Figures

Fig. 6.2

Fig. 6.3 Fig. 6.4

Fig. 6.5

The results of the manual content analysis for the 9 campaigns that tweeted about vaccines, including total tweets coded (up to 50 tweets when available; top left); valence (top right); share of posts referring to vaccine harms (bottom left); and share of posts referring to vaccine conspiracies (bottom right)112 The topic distribution for the 6 top countries in terms of vaccine communication volume (removed countries with n = 2,3 and 10) 113 The topic network modeled from the corpus of all vaccine related English tweets by accounts from 9 countries which posted about the subject. Nodes represent the 16 topics in the model; Edges represent co-occurrence of topics in documents (calculated using cosine similarity); Color represents community membership (using the Louvain algorithm). Network is weighted, and undirected. Edges filtered using the ‘backbone extraction’ method 115 The theme distribution for the 6 top countries in terms of vaccine communication volume (removed countries with n = 2, 3, and 10). Themes calculated using the network communities in Fig. 6.4 117

List of Tables

Table 4.1 Table 4.2 Table 5.1 Table 5.2 Table 6.1 Table 7.1

BEND Maneuvers 60 COVID-19 Vaccine Twitter Data Collection Keywords 63 Multiple Regression Analysis for Vaccine Strategies Predicting Engagement91 Multiple Regression Analysis for Vaccine Strategies by Vaccine Types Predicting Engagement 93 Topic labels, and top unique words (FREX) for each topic 114 Metcalfe’s Science Communication Model as Exemplified in NGOs’ Activities 141

xix

CHAPTER 1

Introduction Tamar Ginossar, Sayyed Fawad Ali Shah, and David Weiss

Communicating about Vaccines Online: Understanding and Counteracting Misinformation, Rumors, and Lies focuses on vaccine communication on social media in diverse contexts. It provides empirical reports from a range of theoretical and methodological perspectives by leading international scholars. Our contributors explore a wide range of issues raised by communication spread via online platforms and the reception, resistance, and reproduction of such communication by multiple stakeholders in various contexts to better understand the content, creators and spread of misinformation on such online platforms—and to suggest ways for effectively countering that misinformation. The idea for this book was formed in 2019, representing our long-­ standing interest in communication about vaccination, and the realization

T. Ginossar (*) • D. Weiss Department of Communication and Journalism, University of New Mexico, Albuquerque, NM, USA e-mail: [email protected]; [email protected] S. F. A. Shah School of Communication and Journalism, Auburn University, Auburn, AL, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 T. Ginossar et al. (eds.), Vaccine Communication Online, https://doi.org/10.1007/978-3-031-24490-2_1

1

2 

T. GINOSSAR ET AL.

that public health leaders had finally began to turn their attention to this previously under-researched and largely unanswered public health risk. In spring 2020, it became painfully clear that this topic had finally received the world’s attention, but no clear pathways were charted on how to address this misinformation. While not providing a full 360-degree review of this complex and evolving topic, in this book we aim to highlight different contexts and processes of misinformation about vaccination, and begin to reflect on lessons learned. The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that causes coronavirus disease 2019 (COVID-19), is an elusive virus that seemed to hit the world out of nowhere, quickly spreading from China to the rest of the world in early 2020. For most global news audiences, it rapidly turned from a distant international story about a virus that originated in an exotic animal market to a long, stressful, uncertain, and often deadly reality. Although the public health sector had been warning about, and preparing for, a vast, global pandemic for decades, it seems that most of the public, including public health officials, was ill-prepared when the pandemic descended. In addition to shortages in trained medical teams and of basic protection and medical equipment, communication to the public was inconsistent in most countries. This communication even included some false information delivered by prominent public health organizations. For example, prior to April 3, 2020, U.S. public health authorities did not recommend face mask use in public. “The initially limited evidence on asymptomatic transmission and concern about mask shortages for the health care workforce and people caring for patients contributed to that initial decision,” reported Feng and their colleagues (2020, p.  434). Some of this misinformation was caused by the ever-­ changing and uncertain situation, and represented the limited knowledge in the face of this new crisis. However, some of this misinformation, unfortunately, was known to be untrue at the time. The notable example above, in which the public was called to avoid using masks by asymptomatic individuals early in the pandemic arguably demonstrates such failure to communicate evidence-based information, and, when this call was reversed on April 3, 2020 by the Centers for Disease Control, may have further reduced trust in the government. While the public often felt that they, and their public health leaders, were in uncharted territory, one group was well-prepared for the pandemic, and was able to immediately draw on their well-honed skills in communication. Unfortunately, this group was the very loosely

1 INTRODUCTION 

3

connected, largely virtual anti-vaccine movement. Our own research (Cruickshank et al., 2021; Ginossar et al., 2022) shows that within a week from the time the pandemic became known to people outside of China, anti-vaccination advocates were already applying their well-designed frameworks to the new virus. After another week, they had largely dropped the falsehood that vaccines cause autism, and focused more on claims about threats to civil rights and to privacy. As the virus continued to spread, so did this influential misinformation from these groups, including disguised efforts by certain governments and organizations, celebrities of different sorts, and individual social media users. With access to social media and available time to members of populations that were now sheltering in place and experiencing social isolation, they were able to spread messages that in the past were largely disseminated interpersonally (Shah et al., 2019; Shah et al., 2021).

The Nature of the Problem Vaccine hesitancy, defined as the disinclination or refusal to vaccinate when vaccines are available (“Ten Threats to Global Health,” 2020, para. 28), is as old as the practice of vaccination itself (See also the chapter by Hoffman et al. in this volume.) However, whereas in the past vaccine hesitancy was considered a fringe phenomenon, recent global outbreaks of vaccine-­ preventable diseases such as measles and diphtheria indicate both newly imminent risks and, at the same time, heightened awareness about the need for a paradigm shift in order to achieve herd immunity worldwide. These concerns increased in the wake of the COVID-19 pandemic and the potential of vaccinations to reduce the enormous global morbidity and mortality of the virus that causes the disease. Exposure to online misinformation—which includes falsehoods, rumors, lies, conspiracy theories, and other types of distortions aimed at increasing doubts about the safety and efficacy of vaccination—is related to reduced likelihood to vaccinate (Larson, 2020). Perhaps equally troubling, the propagation of such vaccine-related misinformation is not only related to but actually accelerates concurrent anti-social processes including the erosion of trust in traditional media, state, and health institutions and the rise of political populism (Zucker, 2020). Alongside responses to the COVID-19 pandemic itself, concerns have been raised about an increase in vaccine hesitancy and—in certain locations—access that might contribute to the global decrease in routine childhood vaccination rates.

4 

T. GINOSSAR ET AL.

Public health practitioners and communication scholars are particularly concerned about the impact of misinformation on the success or failure of vaccination campaigns of misinformation spread through social media (Schiavo, 2020). Evidence suggests that COVID-19 related online misinformation, particularly, has resulted in the rejection of other childhood immunizations by hesitant parents (Velasquez-Manoff, 2022). Given the international rollout of highly effective COVID-19 vaccines since 2021, it has become even more important to explore the nature and manifestation of misinformation about vaccines on different social media platforms and its impact on vaccine hesitancy in diverse contexts and cultures. Such examination can also inform interventions and policies in a post-­ COVID-­19 world and in response to potential future pandemics. Misleading, incorrect, and false information online are among the greatest risks to global public health in the twenty-first century (Larson, 2020). The spread of online misinformation about vaccines, in particular, has emerged as a threat to one of the most celebrated accomplishments in the history of global public health—the eradication of infectious diseases through vaccination (Jamison et al., 2020; Wardle & Singerman, 2021). Vaccine misinformation spread via social media platforms has been shown to have negative impacts on vaccination-related attitudes and behaviors (Featherstone & Zhang, 2020). The COVID-19 pandemic brought these impacts to the center of both the political and the public health agendas. Vaccine-related misinformation is negatively impacting COVID-19 vaccination intention (Lu & Xiao, 2023); worse, exposure to recent misinformation has induced a decline in intent to get the COVID-19 vaccine, even among people who had previously said that they would accept the vaccine (Loomba et al., 2021). As this is the case, social media companies have faced mounting criticism from public health and advocacy groups for failing to counter vaccine-related misinformation. In response, the corporate parents of Facebook, Pinterest, Twitter, and YouTube have taken active measures to flag and even remove such misinformation about vaccination. However, these measures have had limited effectiveness (Wardle & Singerman, 2021), have been criticized as unethical (Piccolo et al., 2021), and have even unintentionally led to information vacuums (Guidry et al., 2020). Overall, these efforts have often been unsuccessful and at times have even backfired (Guidry et al., 2020; Tang et al., 2021; van der Linden et al., 2021). Still, it is undeniably important to better understand the nature of online misinformation about vaccination and approaches to its mitigation.

1 INTRODUCTION 

5

Such understanding can inform efforts to promote trust in vaccination worldwide, contribute to policies that would decrease vaccine hesitancy, and effect the level of vaccine uptake necessary for the eradication of preventable infectious diseases.

Preview of the Book This book’s chapters were selected based on the expertise of their authors, the methodological rigor of their scholarship, the diversity of topics they cover, and their range of geographic and cultural perspectives. They represent a wide range of research methods, including historical literature and scoping reviews (Chaps. 2 and 3); computational analysis, including machine learning (Chaps. 4, 5, and 6); and reviews that incorporate the authors’ personal, professional, and practice-based experiences (Chaps. 7, 8, and 9). The book is divided into three sections. The chapters in the first section, titled “Historical Context and Current Status of Vaccine Misinformation,” provide the historical context necessary for an understanding of contemporary vaccine misinformation. In Chap. 2, Beth Hoffman, Jaime Sidani, and Elizabeth Felter document perhaps surprising findings that anti-vaccine messages have—for more than two centuries— consistently used similar strategies, including fear-evoking images, distortion of data, and personal narratives. These authors show that an understanding of well-established rhetorical strategies can be leveraged in order to develop the remedies needed to counter the impact and spread of contemporary misinformation. Next, in Chap. 3, Yuan Wang, Kathryn Thier, and Xiaoli Nan provide a scoping review of the current research concerning the prevalence, characteristics, and impacts of online Human Papilloma Virus (HPV) vaccine misinformation, as well as the efforts to mitigate such impacts. As they show, this online vaccine misinformation is often characterized by negative sentiments, emotional language, anecdotal evidence, and conspiracy theories. The book’s second section, “Infodemic: Dissecting Vaccine Misinformation on Social Media,” comprises chapters that examine the drivers of misinformation sharing and endorsement on social media platforms, especially Twitter. Janice Blane, Lynnette Hui Xian Ng, and Kathleen Carley, the authors of Chap. 4, utilize computational methods to analyze 30 terabytes of COVID-19 tweets in order to explore how

6 

T. GINOSSAR ET AL.

pro- and anti-vaccine groups use varying online manipulation techniques to influence members of their networks. Their findings reveal that both pro- and anti-vaccine advocates use what the authors call “explain” and “dismay” maneuvers to convey their narratives. However, as they show, anti-vaccination posters explain their positions through anecdotes and therefore appeal more to human emotions, whereas the pro-vaccination tweets mostly focus on statistics and information, using facts and reasoning about the safety and efficacy of the vaccine, and thus achieving quite different, if not always encouraging, results. Blane and her colleagues end their chapter by offering practical suggestions on countering vaccine misinformation online. In Chap. 5, Itai Himelboim, Jeonghyun Lee, Michael Cacciatore, Sungsu Kim, Diane Krause, Kate Miller-Bains, Kristin Mattson, and Jennifer Reynolds identify a range of narratives and rhetorical strategies used by anti-vaccine groups in the Twitter discourses surrounding vaccinations against HPV and various childhood diseases. Analogous to the findings of Blane and her colleagues in Chap. 4, Himelboim and his collaborators, too, show a contrast in messaging strategies across vaccine views and by vaccine type; for example, vaccine-hesitant tweets about HPV rely on personal narratives or stories far more than posts that focus on early childhood vaccination. Moreover, conspiracy-related content about childhood vaccines receive more engagement, although such engagement mainly aims to debunk or correct conspiracy theories. In the final Chap. 6 of this section, Dror Walter and Yotam Ophir use computational methods for the analysis of big data to advance our understanding of the nature of vaccine misinformation spread intentionally by trolls from ten different countries. In doing so, they shed light on the political aspects and some of the drivers of vaccine misinformation around the globe. The third section of the book, “Online Vaccine Misinformation Across Cultures: Barriers and Facilitators,” helps readers understand the nature of online misinformation about different vaccine, and especially COVID-19 vaccine, across different cultures. The chapters in this section also discuss strategies used by public health authorities and non-governmental organizations to counter the online misinformation and increase vaccination uptake. In Chap. 7, Zia Ullah, Shahsawar Khan, and Sayyed Fawad Ali Shah provide a review of polio-related misinformation in Pakistan, showing how it paved the way for the spread of misinformation about COVID-19 in that nation. In addition to analyzing the similarities—and the crucial

1 INTRODUCTION 

7

differences—between these two vaccine misinformation waves, the authors also analyze the various strategies used by the Pakistani government to counter them. The co-authors of Chap. 8—Aviv Sharon, Itamar Netzer, Erez Garty, Noa Hirsch-Choritz, Keren Landsman, and Uri Lerner—each lead a governmental or non-governmental organization in Israel. This allows them to bring a unique perspective to their chapter, which describes the evolution of their work to reduce parental vaccine hesitancy and increase childhood vaccination rates. To our knowledge, the lessons they have learned from various public outreach programs that their organizations developed and supported are the first to be reported from a practice-based perspective rather than that of communication experts per se. Their efforts demonstrate the void in governmental response to anti-vaccination propaganda prior to the COVID-19 pandemic, their weak response during this outbreak, and the efforts of NGOs to step in, albeit without professional/ official communication background. These processes can be applied to other contexts as well. The book’s final Chap. 9, by Anat Gesser-Edelsburg, offers a distinctive argument for dealing with online misinformation. To do so, she synthesizes data from multiple research projects in a variety of countries, suggesting that so-called official sources may not always have a monopoly on what can be considered “truth” in regard to vaccine (mis)information online. While they represent a range of foci, methodologies, and ideological positions, the chapters in this book share the goals of advancing our understandings of the prevalence and nature of online misinformation and offering emerging evidence-based practices and strategies to counter that misinformation.

References Cruickshank, I., Ginossar, T., Sulskis, J., Zheleva, E., & Berger-Wolf, T. (2021). Content and dynamics of websites shared over vaccine-related tweets in COVID-19 conversations: Computational analysis. Journal of Medical Internet Research, 23(12), e29127. Featherstone, J.  D., & Zhang, J. (2020). Feeling angry: The effects of vaccine misinformation and refutational messages on negative emotions and vaccination attitude. Journal of Health Communication, 25(9), 692–702.

8 

T. GINOSSAR ET AL.

Feng, S., Shen, C., Xia, N., Song, W., Fan, M., & Cowling, B. J. (2020). Rational use of face masks in the COVID-19 pandemic. The Lancet Respiratory Medicine, 8(5), 434–436. Ginossar, T., Cruickshank, I. J., Zheleva, E., Sulskis, J., & Berger-Wolf, T. (2022). Cross-platform spread: Vaccine-related content, sources, and conspiracy theories in YouTube videos shared in early twitter COVID-19 conversations. Human Vaccines & Immunotherapeutics, 1–13. Guidry, J.  P., Vraga, E.  K., Laestadius, L.  I., Miller, C.  A., Occa, A., Nan, X., Ming, H. M., Qin, Y., Fuemmeler, B. F., & Carlyle, K. E. (2020). HPV vaccine searches on Pinterest: Before and after Pinterest’s actions to moderate content. American Journal of Public Health, 110(S3), S305–S311. Jamison, A., Broniatowski, D. A., Smith, M. C., Parikh, K. S., Malik, A., Dredze, M., & Quinn, S. C. (2020). Adapting and extending a typology to identify vaccine misinformation on twitter. American Journal of Public Health, 110(S3), S331–S339. Larson, H.  J. (2020). Stuck: How vaccine rumors start—And why they don’t go away. Oxford University Press. Lu, J., & Xiao, Y. (2023). Do Socioeconomic disparities matter? Unraveling the impacts of online vaccine misinformation on vaccination intention during the COVID-19 pandemic in China. Journal of Health Communication, 28(2), 91–101. Loomba, S., de Figueiredo, A., Piatek, S. J., de Graaf, K., & Larson, H. J. (2021). Measuring the impact of COVID-19 vaccine misinformation on vaccination intent in the UK and USA. Nature Human Behaviour, 5(3), 337–348. Piccolo, L.  S., Bertel, D., Farrell, T., & Troullinou, P. (2021, May). Opinions, intentions, freedom of expression...and other human aspects of misinformation online. In Extended abstracts of the 2021 CHI conference on human factors in computing systems (pp. 1–5). Schiavo, R. (2020). Vaccine communication in the age of COVID-19: Getting ready for an information war. Journal of Communication in Healthcare, 13(2), 73–75. Shah, S.  F. A., Ginossar, T., & Weiss, D. (2019). “This is a Pakhtun disease”: Pakhtun health journalists’ perceptions of the barriers and facilitators to polio vaccine acceptance among the high-risk Pakhtun community in Pakistan. Vaccine, 37(28), 3694–3703. Shah, S. F. A., Jan, F., & Ittefaq, M. (2021). Health and safety risks to journalists during pandemics. In S.  Jamil, B.  Coban, B.  Ataman, & G.  Appiah-Adjei (Eds.), Handbook of research on discrimination, gender disparity, and safety risks in journalism (pp. 90–103). IGI Global. Tang, L., Fujimoto, K., Amith, M. T., Cunningham, R., Costantini, R. A., York, F., et  al. (2021). “Down the rabbit hole” of vaccine misinformation on YouTube: Network exposure study. Journal of Medical Internet Research, 23(1), e23262.

1 INTRODUCTION 

9

Ten threats to global health in 2019. (2020, December 18). World Health Organization. https://www.who.int/news-­room/spotlight/ten-­threats-­toglobal-­health-­in-­2019 van der Linden, S., Dixon, G., Clarke, C., & Cook, J. (2021). Inoculating against COVID-19 vaccine misinformation. EClinicalMedicine, 33(100772). https:// doi.org/10.1016/j.eclinm.2021.100772 Velasquez-manoff, M. (2022, May 25). The anti-vaccine movement’s new frontier. The New  York Times. https://www.nytimes.com/2022/05/25/magazine/anti-­vaccine-­movement.html. Wardle, C., & Singerman, E. (2021). Too little, too late: Social media companies’ failure to tackle vaccine misinformation poses a real threat. BMJ, 327n26. https://doi.org/10.1136/bmj.n26 Zucker, H. A. (2020). Tackling online misinformation: A critical component of effective public health response in the 21st century. American Journal of Public Health, 110(S3), S269.

CHAPTER 2

Vaccine Misinformation on Social Media: Historical Contexts, Lessons Learned, and Paths Forward Beth L. Hoffman, Jaime E. Sidani, Jessica G. Burke, Kar-Hai Chu, and Elizabeth M. Felter

In 2019 the World Health Organization named vaccine hesitancy a top-10 threat to global health (World Health Organization, 2019). To understand the historical roots of this phenomenon and its contemporary implications, this chapter will begin with the history of vaccine development

B. L. Hoffman (*) • J. E. Sidani • J. G. Burke • K.-H. Chu Department of Behavioral and Community Health Sciences, University of Pittsburgh School of Public Health, Pittsburgh, PA, USA Center for Social Dynamics and Community Health, University of Pittsburgh School of Public Health, Pittsburgh, PA, USA e-mail: [email protected]; [email protected]; [email protected]; [email protected] E. M. Felter Department of Behavioral and Community Health Sciences, University of Pittsburgh School of Public Health, Pittsburgh, PA, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 T. Ginossar et al. (eds.), Vaccine Communication Online, https://doi.org/10.1007/978-3-031-24490-2_2

11

12 

B. L. HOFFMAN ET AL.

and policy for multiple infectious diseases and an overview of the growth of the anti-vaccine movement. Next, we will place findings from research on vaccine misinformation on social media into a broader historical framework. Finally, we will discuss how applying a historical perspective can help counter the impact and spread of vaccine misinformation, thus improving vaccine education, promotion, and policy.

Variolation Evidence suggests that variolation, or the inoculation of non-immune individuals with pustules from a current smallpox victim, may have been practiced in China as early as 1000 C.E. in attempts to prevent smallpox (Kinch, 2018). For centuries, smallpox devastated societies around the globe, and variolation resulted in individuals contracting a mild form of the disease, with a case-fatality rate approximately 10 times lower than that from naturally-contracted smallpox (Riedel, 2005). In 1670, Circassian traders introduced variolation to the Ottoman Empire, where it gained the attention of Lady Montagu, wife of the British ambassador to the Empire (Riedel, 2005). Initially, the British medical community was resistant to this idea, but due to the advocacy of Lady Montagu and others, by the 1740s the use of variolation became widely adopted by wealthy and educated segments of the British population (Kinch, 2018). Nearly simultaneously (in 1721), an ocean away, in Boston, a smallpox epidemic occurred, affecting half of the city’s 12,000 citizens (Beall OT, 1954). As the epidemic began, the Reverend Cotton Mather reached out to the medical community about variolation. Only one physician, Dr. Zabdiel Boylston, was receptive to the idea (Niederhuber, 2014). Opposition to the variolation program was led by Dr. William Douglass and Benjamin Franklin’s brother, James Franklin, who launched The New England Courant, a newspaper devoted to countering variolation (Niederhuber, 2014). Articles published arguing against variolation in the Courant represent some of the first reported instances of tabloid journalism, which set the stage for the use of mass media to amplify anti-vaccine sentiment. As did anti-variolation forces in Britain, opponents in Boston expressed concern about variolation as “Eastern medicine” (Niederhuber, 2014) due to its origins in Asia. Opponents of variolation also expressed safety concerns and claimed that variolation violated religious law (Niederhuber, 2014). However, Mather and Boylston continued with their inoculation

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

13

efforts, determining that the case-fatality rate among variolated individuals was seven times lower than among those who contracted the disease naturally (Beall OT, 1954). These data, published in medical journals and reported on by the press, quieted opposition and influenced the rapid adoption of variolation in both Europe and the British colonies during the latter half of the eighteenth century (Riedel, 2005).

The Smallpox Vaccine Based on the observation that dairymaids who contracted cowpox became immune from smallpox, Edward Jenner hypothesized that deliberately inoculating an individual with cowpox could protect against smallpox (Riedel, 2005). In 1796, he collected matter from the cowpox lesions of a dairymaid and inoculated an 8-year-old boy. Although the boy developed mild symptoms, he did not develop smallpox when Jenner exposed him to the disease three months later (Riedel, 2005). Jenner concluded that his procedure had been effective, and in 1797 he coined the term “vaccination” when describing the procedure in a self-published booklet (Riedel, 2005). Smallpox vaccination proved so effective that in 1809 Massachusetts began requiring proof of smallpox vaccination for children to attend school, and other states soon followed suit (Kinch, 2018). In 1813, the U.S. Congress passed the Act to Encourage Vaccination. As part of this act, Dr. James Smith was appointed national Vaccine Agent and maintained a supply of cowpox scabs to mail to physicians around the country (Lanzarotta & Ramos, 2018). In 1822, Dr. Smith mistakenly mailed smallpox rather than cowpox scabs, resulting in his dismissal as national Vaccine Agent and the repeal of the Act (Lanzarotta & Ramos, 2018). Thus, although smallpox vaccination continued in the U.S. throughout the nineteenth century, it was decentralized, with states and local governments, as opposed to the federal government, overseeing vaccine distribution and community vaccination efforts. Across the Atlantic, in 1840, Great Britain passed its first national law related to vaccination. This law banned variolation and provided free vaccinations for the poor, to be administered by newly created Poor Law Unions (National Archives, 2021). After a report suggesting that vaccination rates remained low, the British government increased its efforts to improve coverage, and in 1853 began mandating smallpox vaccination for infants up to three months of age (Porter & Porter, 1988). In 1857,

14 

B. L. HOFFMAN ET AL.

British Medical Officer Jon Simon gathered extensive data and published his results regarding the benefits of vaccination, followed by additional reports from other scientists with the Medical Department. These reports provided the basis for legislation in 1867 that extended the mandate to children up to 14 years of age, with cumulative penalties for noncompliance (Durbach, 2000; Porter & Porter, 1988). Public opinion was split between those who supported these laws as a way to maintain a robust economy and those who were against mandatory vaccination for various reasons, including association with Poor Law Unions or on religious grounds (National Archives, 2021). In response to the 1853 mandate, activists in London formed the Anti-Vaccination League, and in 1867 founded the Anti-Compulsory Vaccination League of Great Britain; the mission of both leagues focused on countering infringements of personal liberty (Wolfe & Sharp, 2002). One of the leaders of this movement, William Tebb, stoked anti-vaccine sentiment by using false data that purported to show that smallpox vaccinations had killed more than 48,000 people in England and Wales (Kinch, 2018). In 1885, following an anti-vaccine demonstration in Leicester that attracted more than 100,000 people, Great Britain appointed a royal commission to investigate smallpox vaccination safety and efficacy (Wolfe & Sharp, 2002). The commission released a report in 1896 supporting the recommendation for smallpox vaccination; however, the report also recommended the abolition of cumulative penalties. Two years later, Parliament passed a new Vaccination Act that eliminated cumulative penalties and allowed parents to obtain an exemption (Wolfe & Sharp, 2002). In the U.S., vaccination in the early nineteenth century proved so effective at controlling outbreaks that, paradoxically, vaccination levels actually decreased by mid-century due to complacency (Wolfe & Sharp, 2002). Declining vaccination caused a resurgence of smallpox in the 1870s, which prompted increased efforts by state and local governments to enforce existing vaccination laws. These efforts, combined with the 1879 visit of William Tebb from Great Britain, motivated the formation of the Anti-­ Vaccination Society of America as well as that of regional leagues which were successful in repealing compulsory vaccination laws in six states (Wolfe & Sharp, 2002). In the Midwest, Lora Cornelia Little, whose son died in 1896, a year after receiving his smallpox vaccine, began a passionate campaign against vaccines, leveraging her own family’s story as part of her appeal to growing concerns about state overreach (Kinch, 2018). To

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

15

promote her claims, she used the Truth Teller, a periodical focused on homeopathy and alternatives to medical interventions (Kinch, 2018). By the end of the nineteenth century, smallpox in the U.S. afflicted primarily the urban poor (Kinch, 2018). Unfortunately, actions taken to increase vaccination rates among this population sometimes fueled antivaccine beliefs. For example, during a smallpox outbreak among New York City’s homeless population in 1893–1894, individuals trained to give vaccines were paid 30  cents for each person they vaccinated, resulting in instances of vaccinating the same person multiple times in order to increase their own income (Kinch, 2018). Such actions lent support to claims by anti-vaccine advocates that the primary motive behind vaccination was profit as opposed to health. In 1902, an outbreak of smallpox in Cambridge, Massachusetts, prompted the city to enforce the state law requiring vaccination of all residents (Parmet et al., 2005). The courts upheld the state’s right to implement mandatory vaccination to protect community health. When Reverend Henning Jacobson refused to be vaccinated, he was convicted and fined (Parmet et al., 2005). In 1905, the U.S. Supreme Court heard his case, Jacobson v. Massachusetts. The Court upheld Massachusetts’ mandatory vaccination law, writing that states may limit individual freedoms in the name of public health (Parmet et al., 2005). This ruling was followed in 1922 by Zucht v. King, in which the Supreme Court upheld a San Antonio, Texas, ordinance requiring school children to be vaccinated against smallpox. Writing for the majority, Justice Louis Brandeis cited the earlier Jacobson v. Massachusetts ruling, stating that the Court “had settled that it is within the police power of a state to provide for compulsory vaccination” (Brandeis, 1922). In the first half of the 1900s, Britian, France, and Belgium instituted smallpox vaccination programs in colonial Africa, with France applying its 1902 mandatory smallpox vaccination law to its colonies (Schneider, 2009). However, due to under-investment in vaccine production and transport, large segments of the population remained unvaccinated (Schneider, 2009). In 1948, the success of the smallpox vaccine in developed countries prompted the World Health Assembly of the World Health Organization (WHO) to formally issue a proclamation stating its plan to eliminate smallpox worldwide (Kinch, 2018). Throughout the 1960s and 1970s, the WHO launched a coordinated global effort. Consequently, in 1977 the last case of smallpox occurred, in Somalia (Hajj Hussein et al., 2015). In 1980, the World Health Assembly certified the world to be free

16 

B. L. HOFFMAN ET AL.

of naturally-occurring smallpox, marking the first and only instance of global eradication of a human infectious disease through vaccination (Immunization Action Coalition, 2013).1 Following the eradication of smallpox, there was interest in eradicating other diseases that infect humans. The World Health Assembly targeted poliomyelitis (polio) because of its devastating effects on children as well as findings in Europe and North America that eradication was possible through vaccination. The following section will examine the history of the development of several vaccines and the co-development of anti-vaccine messages and movements, starting with the extraordinary story of the polio vaccine.

The Polio Vaccine In the first half of the twentieth century, epidemics of poliomyelitis were regular occurrences, with up to 20,000 cases of paralytic polio reported in the U.S each year. On April 12, 1955, the Salk polio vaccine, which utilized a formalin-inactivated virus (IPV), was declared safe and effective and licensed following a placebo-controlled trial (Offit, 2005). The vaccine reduced the instance of paralytic polio from 13.9 cases/100,000 in 1954 to 0.8 cases/100,000 in 1961 (Baicus, 2012). However, in 1955, two weeks after mass vaccination began, five children became paralyzed after receiving the polio vaccine (Offit, 2005). An investigation found that these children had received a vaccine produced at Cutter Laboratories, a facility in which the virus had not been properly inactivated (Offit, 2005). Although the vaccine was immediately recalled, 380,000 individuals had already been vaccinated from this batch. Ultimately, 94 individuals and 166 of their close contacts developed polio (Baicus, 2012), causing considerable public concern about the safety of the vaccine. Following this 1955 incident, the National Institutes of Health (NIH) expanded its Division of Biologics Control, and later that year, Congress passed the Polio Vaccination Assistance Act, marking the first federal involvement in immunization since the early 1800s (Immunization Action Coalition, 2013). Although the Salk vaccine proved to be very effective at curbing polio, the Cutter incident created mistrust in the pharmaceutical industry that persists today despite extensive regulation put in place to prevent similar events.

1

 Rinderpest, a disease that affects livestock, was declared eradicated by the WHO in 2001.

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

17

In 1963, the Sabin polio vaccine, a live-attenuated vaccine administered orally (OPV), was licensed (Baicus, 2012). The U.S. and countries worldwide quickly adopted the OPV due to lower cost, ease of administration, and development of better intestinal immunity (Bandyopadhyay et  al., 2015). In rare instances, individuals developed vaccine-associated paralytic polio following administration of the OPV.  However, the live virus was capable of mutating, which led to polio outbreaks of altered viruses. Thus, in 1997, the Centers for Disease Control and Prevention (CDC) recommended switching back to the IPV vaccine (Baicus, 2012). Globally, the WHO has also advocated for the use of the IPV, although OPV is still used for routine immunization (Baicus, 2012). In 1988, the World Health Assembly passed a resolution to globally eradicate polio by the year 2000, but this goal has yet to be achieved (Immunization Action Coalition, 2013). In 1994, the WHO declared the Western Hemisphere polio-free, followed by a similar declaration about Europe in 2002 (Immunization Action Coalition, 2013). As of 2021, wild-type polio existed only in Pakistan and Afghanistan, but between January 2019 and June 2021, 32 countries experienced outbreaks of polio from vaccine-derived poliovirus (Bigouette, 2021). Vaccine derived poliovirus was originally used in the oral polio vaccine (OPV) and has changed over time and behaves like wild poliovirus. It also causes paralysis. It spreads easily among individuals and communities that are not vaccinated against polio. Reports regarding vaccine-related polio paralysis and infection from vaccine-derived poliovirus have fueled claims from the anti-vaccine community that polio is actually caused by the vaccine, as opposed to a virus (Dutta, 2008). Moreover, in post-colonial countries, specifically Nigeria, Pakistan, and Afghanistan, mistrust of vaccines is often fueled by conspiracy theories concerning a “Western plot” to sterilize non-White communities (Warraich, 2009), rooted in historical traumas and a history of colonial abuses (Shah et  al., 2019). Health care workers have been assaulted or even killed, as religious leaders spread claims that the vaccines are contaminated with HIV or with chemicals intended to sterilize their populations (Warraich, 2009).

18 

B. L. HOFFMAN ET AL.

The Diphtheria, Tetanus, and Pertussis (DTP) Vaccine In the 1940s, the diphtheria, tetanus, and pertussis (DTP) vaccine, which combined diphtheria and tetanus toxoid with the whole-cell pertussis vaccine, came into widespread clinical use (Immunization Action Coalition, 2013). By the 1970s, concerns about potential adverse reactions, particularly related to the whole-cell pertussis formulation, reached a tipping point in Great Britain and Japan. In Great Britain, the 1974 publication of a series of case studies of more than 30 children who had developed severe neurological impairment following DTP immunization fueled these concerns (Baker, 2003). In 1975, concerned parents in Great Britain joined together to form the Association of Parents of VaccineDamaged Children (Baker, 2003). Overall, confidence in administering the vaccine decreased; a 1977 survey revealed almost half of general practitioners were unsure about giving the vaccine to a child unless a parent specifically requested it (Baker, 2003). As a result, vaccination rates plummeted, and in the late 1970s there were three epidemics of pertussis in Great Britain. Although the Joint Committee on Vaccination and Immunization affirmed the safety of the DTP vaccine in 1974, initially the British government did little to restore public confidence (Baker, 2003). Not until 1982 did public health officials use media channels to promote a major immunization campaign. As a result, newspapers such as The Times of London increased their coverage of the recent pertussis epidemics with headlines such as “more babies die” and “whooping cough cases at new record level” (Baker, 2003). This shift, from emotionally charged headlines about the potential dangers of the DTP vaccine to emotionally charged headlines about the dangers of pertussis, was likely influential in increasing immunization rates (Baker, 2003). However, officials’ efforts were met by strong resistance from the anti-vaccine community, whose members accused the media of fear-mongering and biased coverage (Baker, 2003). Similarly, in 1974, the Japanese media intensified reports of two infants who died after receiving the DTP vaccine (Kinch, 2018). Although mandated pertussis vaccination had been extremely successful in curbing the disease—with no childhood deaths from pertussis in Japan in 1972—the subsequent uproar caused the Japanese government to suspend mandatory pertussis vaccination (Kinch, 2018). By the late 1970s, the anti-DTP

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

19

movement spread to the U.S., the Soviet Union, and Australia (Baker, 2003). In the U.S. in 1982, a Washington, D.C., television station broadcast titled DPT: Vaccine Roulette amplified concern over the vaccine (Baker, 2003; Kinch, 2018). When concerned parents contacted the station following the broadcast, they were provided with the phone numbers of other concerned individuals, effectively igniting a grass-roots movement (Kinch, 2018). Following this broadcast, Dr. Harris Coulter and Barbara Loe Fisher founded the National Vaccine Information Center. Although the name of this organization made it sound unbiased, its purpose was to disseminate information supporting the growing anti-vaccine movement (Kinch, 2018). A few years later, Dr. Coulter and Ms. Fisher published DPT: A Shot in the Dark, which used emotional appeals to feed concerns over the DTP vaccine (Kinch, 2018). Throughout the 1970s and ’80s, the U.S. medical community firmly supported vaccination. Unfortunately, many physicians were caught off guard by the rapidly increasing resistance from parents to the DTP vaccine, and were not sufficiently prepared with the facts and skills needed to address the complexity of the risks and benefits of DTP vaccination (Kinch, 2018). Consequently, shortly after the 1982 broadcast of DPT: Vaccine Roulette, the American Academy of Pediatrics, the American Medical Association, and the CDC launched an aggressive media campaign that helped stabilize immunization rates (Baker, 2003; González, 1982). A major consequence of the anti-DTP movement in the U.S. included a significant rise in litigation. By 1986, there were more than 250 annual lawsuits related to the DTP vaccine, compared to only two lawsuits in 1978 (Freed et  al., 1996). Despite numerous population-based studies that found no link between the vaccine and adverse neurological events, this stream of lawsuits led two of the three manufacturers of the vaccine to cease production, resulting in a vaccine shortage (Freed et al., 1996). In response to concerns about both vaccine shortages and growing parental activism, Congress passed the National Childhood Vaccine Injury Act in 1989, which established the Vaccine Adverse Event Reporting System (VAERS), the National Vaccine Injury Compensation Program (NVICP), and the National Vaccine Program Office (Mariner, 1992). The VAERS allows health professionals and members of the general public to submit reports describing adverse reactions to vaccines, and the NVICP provides compensation for individuals injured by vaccinations on a “no fault” basis (Centers for Disease Control and Prevention, 2015). The National Childhood Vaccine Injury Act also mandated promotion of safer vaccines,

20 

B. L. HOFFMAN ET AL.

which allowed the NIH to sponsor multiple clinical trials of an acellular pertussis vaccine. This accelerated the development of the DTaP2 vaccine, which was licensed in 1991 by the U.S. Food and Drug Administration (Fine, 2003). The National Childhood Vaccine Injury Act also commissioned the Institute of Medicine (IOM) to review adverse vaccination events. In 1991 and 1994, the IOM released two comprehensive reviews on adverse events anecdotally tied to the DTP and MMR vaccines. The reports found insufficient evidence to indicate a causal relationship between vaccines and most adverse events, specifically regarding the pertussis vaccine and neurological damage (Baker, 2003; Freed et  al., 1996). In 2011, the IOM issued another report concluding that vaccines cause few health problems (Immunization Action Coalition, 2013).

The MMR Vaccine Perhaps the best-known case of the impact of anti-vaccination activism on public perceptions and concerns regarding vaccine safety relates to the combined measles, mumps, rubella (MMR) vaccine. The FDA licensed the vaccine in 1971 (Hajj Hussein et al., 2015). Concerns over the MMR vaccine were magnified in 1998, when Mr. Andrew Wakefield3 claimed to have uncovered a link between the vaccine and autism spectrum disorder (Hajj Hussein et al., 2015). In his paper, published in the Lancet, Wakefield described 12 children who developed symptoms of autism following MMR vaccination (Hajj Hussein et al., 2015). Investigative journalism by Brian Deer of the British newspaper The Sunday Times revealed numerous methodological problems with Wakefield’s manuscript, as well as two significant conflicts of interest that Wakefield had failed to disclose (Deer, 2011). In 2010, the Lancet publicly retracted the paper (Deer, 2011). However, by the time of the retraction, Wakefield’s claims had already been amplified by television interviews and publicity encouraged by his employer, the Royal Free Hospital School of Medicine (Kinch, 2018). Furthermore, unlike the DTP controversy, which took years to spread globally, Wakefield’s claims were able to spread rapidly around the world, thanks to the Internet and widespread social media usage (Kinch, 2018).

2 3

 The “a” stands for “acellular.”  Previously a physician, Wakefield was striped of his British medical license in 2010.

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

21

Additionally, in 1999, shortly after the publication of Mr. Wakefield’s paper, the U.S. government released a report suggesting that thimerosal, a preservative used in some vaccines, might expose infants to more mercury than previously thought (Gross, 2009). Thimerosal is not used in any formulation of the MMR vaccine, but this report, combined with the Wakefield article, ignited the theory that autism might be related to vaccine-­induced mercury exposure. One of the most vocal advocates promoting these claims was actress and model Jenny McCarthy. McCarthy spearheaded the “Green our Vaccines” movement, which advocated for the removal of “toxins” from vaccines (Stern & Markel, 2005). Similar to their role in promoting Wakefield’s claims, media outlets again played a prominent role in disseminating this misinformation. For example, in 2008, McCarthy appeared on CNN for a vaccine-autism “debate” stating that her son’s autism was due to mercury in vaccines. No scientists or healthcare workers were present to provide scientific or medical facts (Gross, 2009). Despite being stripped of his medical license, Wakefield has continued to tout his opinions and findings aggressively. He wrote a book, gave dozens of public lectures, and developed a following among wealthy Americans and celebrities including McCarthy and actor Charlie Sheen, both of whom have children with autism (Kinch, 2018). In 2016, Wakefield wrote and directed the documentary Vaxxed about his claims around harms of vaccination. Initially scheduled to be screened at New York City’s prestigious Tribeca Film Festival, outrage from other filmmakers and the medical community led to the withdrawal of the documentary (Kinch, 2018, p.  241). Nonetheless, Wakefield moved to Texas and began a traveling tour of the documentary, and the documentary was shared extensively over social media (Bennato, 2017).

The HPV Vaccine In 2006, the FDA approved Gardasil, the first vaccine specifically designed to prevent cancer by protecting against four of the most prevalent strains of HPV (Brookes, 2016). In 2014, the FDA approved Gardasil 9, which protects against five additional strains of the virus (Brookes, 2016). In the U.S., routine HPV vaccination is recommended for girls and boys ages 11–12 (vaccination may start as young as 9), and females and males ages 13–26 who have not been vaccinated previously (Markowitz et al., 2014). On October 5, 2018, the FDA approved the vaccine for men and women

22 

B. L. HOFFMAN ET AL.

ages 27–45 (Grady & Hoffman, 2018). Currently, the CDC recommends two doses of the HPV vaccine for most young people who begin the series prior to age 15, with the second dose given between six and 12 months after the first (Centers for Disease Control and Prevention, 2021). For individuals over 15 or those who are immunocompromised, the CDC recommends three doses of the vaccine, with the second dose given one to two months after the first, and the third dose given six months after the first (Centers for Disease Control and Prevention, 2021). Despite evidence supporting the safety and efficacy of the vaccine, vaccination rates among adolescents remain low (Walker et al., 2018). As of 2021, 75% of U.S. teens have had at least one dose of the HPV vaccine, with 58.6% of teens being fully vaccinated, in contrast to the >90% vaccination rates among teens for many other vaccines including meningitis, MMR, and chickenpox (Pingali, 2021). Despite its strong safety record (Valentino & Poronsky, 2016), the HPV vaccine is much less frequently adopted than other routine childhood vaccines. One reason may be parental vaccine hesitancy. A 2019 survey found that 23% of U.S. parents reported being hesitant about the HPV vaccine and that adolescents living with vaccine-hesitant parents were significantly less likely to receive the vaccine. In this sample, parents voiced concerns about potential side effects and the “novelty of the vaccine” (Szilagyi et al., 2020). Previous research suggests that other reasons for hesitancy include receiving misinformation from notable public figures, concerns about vaccine safety, lack of access to healthcare, and concern that the vaccine will increase risky sexual behavior (Perkins et  al., 2010). These fears are often stoked by media coverage of adolescents supposedly harmed by the vaccine (Dunn et al., 2015; Keelan et al., 2010). In Japan, media coverage and policy surrounding the HPV vaccine followed an almost identical course to that of the DTP vaccine 40 years prior. In 2013, Japan added the HPV vaccine to the country’s recommended immunization schedule, but unconfirmed reports of side effects by the media led the government to withdraw its recommendation for the vaccine just three months later (Hanley et al., 2015). Following this change, the HPV vaccination rate for 12-year-old girls in Japan fell from 70% in 2012 to just 0.1% in 2014 (Tanaka et  al., 2017). Despite the Vaccine Adverse Reactions Review Committee’s 2014 conclusion that there was no evidence to support a causal relation between the HPV vaccine and reported adverse reactions, and increasing calls for Japan to reinstate the

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

23

HPV vaccine in their schedule of vaccines, the Japanese government has yet to do so (Tanaka, 2020). In contrast to Japan and the U.S., Australia has achieved high rates of HPV vaccination and is on track to virtually eliminate cervical cancer from the country. High rates of vaccination are likely due to both a National HPV Vaccination Program, introduced in 2007, for girls and expanded to boys in 2013 (Hall et al., 2019), as well as school-based vaccination programs combined with adolescent and parental education to increase vaccine confidence and reduce anxiety (Skinner et  al., 2015). Despite the relatively recent introduction of the National Program, in 2016 over 72% of 15-year-old Australian boys and girls were fully vaccinated. A recent modeling study estimates that cervical cancer could be eliminated in Australia within the next two decades if current vaccination and screening practices are maintained (Hall et al., 2019). Crucial to this effort will be addressing rising levels of vaccine hesitancy.

Vaccine Hesitancy in the Twenty-First Century All 50 U.S. states have passed legislation requiring children to be up to date on certain immunizations in order to attend school. However, all but six states allow religious or other personal-belief exemptions (National Conference of State Legislatures, 2021). Research suggests that states with higher rates of nonmedical exemptions have lower MMR vaccination coverage (Olive et al., 2018), which is associated with disease outbreaks (Zipprich et al., 2015). In the U.S., the growing number of communities with vaccination levels below those needed for herd immunity resulted in over 1200 cases of measles in 2019, a fourfold increase over the prior year (Centers for Disease Control and Prevention, 2020). As noted above, the news media have often played a substantial role in propagating anti-vaccine sentiment since vaccines’ inception. The advent of social media has further facilitated the spread of content including anti-­ vaccine material and is a likely contributor to rising levels of vaccine hesitancy and decreased vaccination rates, particularly in more-developed countries (Betsch et  al., 2012; Schmidt et  al., 2018). By increasing the ease of finding like-minded people, social media may magnify anti-vaccine rhetoric and catalyze anti-vaccine behavior (Brewer et  al., 2017; Dunn et al., 2017). Integrating findings from research on anti-vaccine rhetoric and the spread of misinformation on social media into a broader historical

24 

B. L. HOFFMAN ET AL.

framework highlights how social media facilitates the amplification and diffusion of centuries-old anti-vaccine arguments and techniques. For example, in the eighteenth century, many individuals in Europe and the American colonies opposed variolation because they viewed it as unnatural and as “Eastern medicine” (Kinch, 2018); in the nineteenth century, opponents of smallpox vaccination cited concerns about civil liberties and safety. Similarly, anti-vaccine Facebook posts were found to include assertions of perceived dangers of “Western medicine,” claims that vaccines are not natural, and statements that mandatory vaccination is a violation of civil liberties (Hoffman et al., 2019). This study also found that many of these individuals often cite historical events such as the Cutter Laboratories IPV inactivation failure described earlier when claiming that vaccination is not safe (Hoffman et al., 2019). Content of anti-vaccination messages on other social media platforms also demonstrates similarities to anti-vaccine arguments from centuries past. For example, YouTube anti-vaccine videos made appeals to “naturalism” and emphasized the right to refuse vaccines (Venkatraman et al., 2015). These vaccine-opposing messages also related to past incidents. An examination of anti-vaccine Twitter messages (“tweets”) from 2019 found that Twitter users who authored the most messages (i.e., top authors) frequently mentioned the 1989 National Childhood Vaccine Injury Act and shared an article citing deaths reported to the VAERS (Bonnevie et  al., 2020; Mariner, 1992). Another study examining Twitter messages related to the #DoctorsSpeakUp pro-vaccine event found that anti-vaccine tweets were more likely than pro-vaccine tweets to contain personal narratives and references to research or science, although the data were distorted (Hoffman et al., 2021).

COVID-19 In 2020, concerns and uncertainty about the novel coronavirus SARS-­ CoV-­2 and the associated COVID-19 disease pandemic propelled vaccine misinformation and vaccine hesitancy into the spotlight. Increased global social media use further contributed to the spread of misinformation that appeared to be traveling faster than the virus itself, leading the WHO to declare an “infodemic” (Zarocostas, 2020). This changing digital landscape coupled with the long history of the anti-vaccine movement created the perfect storm. For reference, the last WHO-declared pandemic was the 2009 H1N1 outbreak, during which only 36% of US adults used at least one social media site such as Facebook, Instagram, Snapchat, or

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

25

Twitter; today, that percentage has doubled to 72% (Pew Research Center, 2019). Research indicates that anti-vaccine groups were already mobilized during the early months of the pandemic (February–March 2020) and quickly began disseminating conspiracy theories via social media (Cruickshank et al., 2021; Ginossar et al., 2022; Kalichman et al., 2021). These misinformation efforts often targeted vulnerable communities and echoed arguments used in decades passed. For instance, capitalizing on the fear of covert sterilization of non-White communities to fuel hesitancy about polio vaccination, prominent anti-vaccine activists spread unsubstantiated claims that COVID-19 vaccines were designed to lower fertility rates in Black communities (Center for Countering Digital Hate, 2021). Antivaccine activists also spread misinformation on predominantly Black social media networks about the disease being manufactured for racial population control (Ross, 2020; Sanz, 2020). This targeted misinformation exploiting past abuses inflicted by medical, public health, and research entities on these communities, combined with institutional racism and lack of equitable access to healthcare (including vaccines), may partially explain hesitancy to receive a COVID-19 vaccine by the very racial and ethnic minority groups that were hardest hit by the virus (Callaghan et al., 2021). This mis/disinformation has been spread by a rather small number of main sources. A March 2021 report by The Center for Countering Digital Hate found that 73% of anti-vaccine content posted or shared on Facebook from February 1 to March 16, 2021, most of which was about COVID-19, was attributable to 12 individuals deemed “the disinformation dozen” (Center for Countering Digital Hate, 2021). The themes in social media discourse related to COVID-19 vaccine are similar to those of the pre-­ pandemic discourse, including conspiracy theories, lack of trust in pharmaceutical companies, promotion of alternative remedies for treating COVID-19 as opposed to vaccination, and an emphasis on personal freedom (Ginossar et  al., 2022; Hughes et  al., 2021; Islam et  al., 2021). Additionally, as with Jacobson v. Massachusetts, the Supreme Court case that affirmed the constitutionality of state vaccine mandates, challenges to COVID-19 vaccine mandates have cited violations of personal freedom (Canellos & Lau, 2021).

26 

B. L. HOFFMAN ET AL.

Conclusion Although many individuals view the anti-vaccine movement as a contemporary phenomenon fueled by Andrew Wakefield’s 1998 Lancet article decrying the MMR vaccine, vaccine skeptics have been active for hundreds of years, well before Edward Jenner developed the first modern vaccine. Understanding this history can help today’s public health practitioners more effectively recognize and counter such sentiments. For example, while banning certain organizations such as the National Vaccine Information Center from creating social media accounts may help curb the spread of anti-vaccine misinformation, knowing that this organization was founded in the 1980s, well before the rampant use of social media, suggests that such restrictions may not be effective at countering the antivaccine movement more broadly due to its now well-established offline organization. Considering vaccine resistance through an historical lens, as this chapter has done, also demonstrates how anti-vaccine organizations have consistently leveraged fear-inducing images, distortion of data, and personal narratives to reach susceptible audiences; thus, public health agencies operating in the current environment should direct their resources to specifically target and counter these campaigns. Understanding the type of messages and strategies that anti-vaccinators have deployed historically to disseminate misinformation is essential. For instance, COVID-19 misinformation suggesting racial targeting of Black communities should be countered with messaging for the specific population; community groups and leaders can be recruited to help support these efforts. Indeed, community-­informed research with Black and Latinx communities in the U.S. highlights the importance of acknowledging past abuses, disseminating information via trusted messengers, and addressing structural barriers that decrease access to vaccinations (Balasuriya et al., 2021). Anti-vaccine messaging can also be stopped before it diffuses widely through real-time surveillance of anti-vaccine groups and messages on social media platforms, news reports, or other outlets, which would allow responses to be generated and delivered to stem the flow of misinformation quickly. While today messages are spread on social media via “influencers,” the anti-vaccine movement has long used vocal proponents to expand its cause, from Dr. William Douglass and James Franklin in the early eighteenth century to Jenny McCarthy and Charlie Sheen in the 21st. Today, public figures such as Robert F. Kennedy Jr., Sherri Tenpenny, and other

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

27

members of the “disinformation dozen” are linked to most of the vaccine misinformation that proliferates on popular social media platforms such as Facebook and Twitter. In December 2020, the 20 largest anti-vaccine Twitter accounts, including those of the “disinformation dozen,” had over 39 million followers (Center for Countering Digital Hate, 2021). By contrast, at the time of this writing, in early 2022, the CDC had just over 4 million followers. An historical lens also allows scholars and practitioners to see that arguments against vaccine mandates have remained consistent across centuries, and that while vaccine mandates have been upheld by the U.S. Supreme Court and can increase vaccination rates, public support is critical to prevent negative repercussions such as increased anti-vaccine resistance. As political leaders work to develop and implement COVID-19 mandates, history suggests it will be vital for public health professionals to simultaneously develop outreach and communication campaigns to address the rationale and historical precedent for vaccine mandates. Finally, investigating anti-vaccination phenomena through an historical lens, as well as through the current social media-based outreach, demonstrates the need for investment in literacy-based campaigns for lay populations. In particular, media literacy—which provides a “framework to access, analyze, evaluate, create and participate with messages”—can help individuals critically analyze the media messages with which they come into contact (Thoman & Jolls, 2005, p. 21). One strength of utilizing a media literacy framework is that it is not limited to any one form of media. Anti-vaccine activists and those who seek to spread health misinformation can adeptly pivot to current popular media, so arming populations with adaptable skills may be of value. As British politician Edmund Burke once said, “those who don’t know history are doomed to repeat it.” Anti-vaccine beliefs, often dismissed as illogical, are deeply rooted in political, spiritual, and philosophical arguments that have persisted over time. Understanding this perspective is key to ensuring that the arc of vaccine history bends toward acceptance.

References Baicus, A. (2012). History of polio vaccination. World Journal of Virology, 1(4), 108–114. https://doi.org/10.5501/wjv.v1.i4.108 Baker, J. P. (2003). The pertussis vaccine controversy in Great Britain, 1974–1986. Vaccine, 21(25–26), 4003–4010. https://doi.org/10.1016/S0264410X(03)00302-­5

28 

B. L. HOFFMAN ET AL.

Balasuriya, L., Santilli, A., Morone, J., Ainooson, J., Roy, B., Njoku, A., et  al. (2021). COVID-19 vaccine acceptance and access among Black and Latinx communities. Journal of the American Medical Association Network Open, 4(10), e2128575. https://doi.org/10.1001/JAMANETWORKOPEN.2021.28575 Bandyopadhyay, A. S., Garon, J., Seib, K., & Orenstein, W. A. (2015). Polio vaccination: Past, present and future. Future Microbiology, 10(5), 791–808. https://doi.org/10.2217/fmb.15.19 Beall OT, S. R. (1954). Cotton Mather: First significant figure in American medicine. Johns Hopkins University Press. Bennato, D. (2017). The shift from public science communication to public relations. The Vaxxed case. Journal of Science Communication, 16(2), C02. https://doi.org/10.22323/2.16020302 Betsch, C., Brewer, N.  T., Brocard, P., Davies, P., Gaissmaier, W., Haase, N., et  al. (2012). Opportunities and challenges of Web 2.0 for vaccination decisions. Vaccine, 30(25), 3727–3733. https://doi.org/10.1016/j. vaccine.2012.02.025 Bigouette, J. P. (2021). Progress toward polio eradication—Worldwide, January 2019–June 2021. Morbidity and Mortality Weekly Report, 70(34), 1129–1135. https://doi.org/10.15585/MMWR.MM7034A1 Bonnevie, E., Goldbarg, J., Gallegos-Jeffrey, A., Rosenberg, S., Wartella, E., & Smyser, J. (2020). Content themes and influential voices within vaccine ppposition on Twitter, 2019. American Journal of Public Health, 110(S3), S326– S330. https://doi.org/10.2105/AJPH.2020.305901 Brandeis, L.  D. and S.  C. of the U.  S. (1922). U.S.  Reports: Zucht v. King, 260 U.S. 174. Brewer, N. T., Chapman, G. B., Rothman, A. J., Leask, J., & Kempe, A. (2017). Increasing vaccination: Putting psychological science into action. Psychological Science in the Public Interest, 18(3), 149–207. https://doi.org/10.1177/ 1529100618760521 Brookes, L. (2016). The HPV vaccine: Then and now. https://www.medscape. com/viewarticle/866591 Callaghan, T., Moghtaderi, A., Lueck, J. A., Hotez, P., Strych, U., Dor, A., et al. (2021). Correlates and disparities of intention to vaccinate against COVID-19. Social Science & Medicine, 272, 113638. https://doi.org/10.1016/J. SOCSCIMED.2020.113638 Canellos, P. S., & Lau, J. (2021, September 8). The surprisingly strong Supreme Court precedent supporting vaccine mandates. Politico. https://www.politico. com/news/magazine/2021/09/08/vaccine-­mandate-­strong-­supreme-­courtprecedent-­510280 Center for Countering Digital Hate. (2021, March 21). The disinformation dozen. https://www.counterhate.com/disinformationdozen Centers for Disease Control and Prevention. (2015). History of vaccine safety. https://www.cdc.gov/vaccinesafety/ensuringsafety/history/index.html#four

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

29

Centers for Disease Control and Prevention. (2020). Measles cases and outbreaks. https://www.cdc.gov/measles/cases-­outbreaks.html Centers for Disease Control and Prevention. (2021). HPV vaccination recommendations. https://www.cdc.gov/vaccines/vpd/hpv/hcp/recommendations.html Cruickshank, I., Ginossar, T., Sulskis, J., Zheleva, E., & Berger-Wolf, T. (2021). Content and dynamics of websites shared over vaccine-related tweets in COVID-19 conversations: Computational analysis. Journal of Medical Internet Research, 23(12). https://doi.org/10.2196/29127 Deer, B. (2011). How the case against the MMR vaccine was fixed. British Medical Journal Clinical Research Edition, 342, c5347. https://doi. org/10.1136/BMJ.C5347 Dunn, A. G., Leask, J., Zhou, X., Mandl, K. D., & Coiera, E. (2015). Associations between exposure to and expression of negative opinions about human papillomavirus vaccines on social media: An observational study. Journal of Medical Internet Research, 17(6), e144. https://doi.org/10.2196/jmir.4343 Dunn, A. G., Surian, D., Leask, J., Dey, A., Mandl, K. D., & Coiera, E. (2017). Mapping information exposure on social media to explain differences in HPV vaccine coverage in the United States. Vaccine, 35(23), 3033–3040. https:// doi.org/10.1016/j.vaccine.2017.04.060 Durbach, N. (2000). They might as well brand us: Working class resistance to compulsory vaccination in Victorian England. The Society for the Social History of Medicine, 13, 45–62. Dutta, A. (2008). Epidemiology of poliomyelitis—Options and update. Vaccine, 26(45), 5767–5773. https://doi.org/10.1016/j.vaccine.2008.07.101 Fine, A. (2003). Diphtheria, tetanus and acellular pertussis vaccine (DTaP): A case study. http://www.nationalacademies.org/hmd/~/media/Files/Activity Files/Disease/VaccineFinancing/FineBackgroundPaper.pdf Freed, G. L., Katz, S. L., & Clark, S. J. (1996). Safety of vaccinations. Journal of the American Medical Association, 276(23), 1869. https://doi.org/10.1001/ jama.1996.03540230019013 Ginossar, T., Cruickshank, I. J., Zheleva, E., Sulskis, J., & Berger-Wolf, T. (2022). Cross-platform spread: Vaccine-related content, sources, and conspiracy theories in YouTube videos shared in early Twitter COVID-19 conversations. Human Vaccines & Immunotherapeutics, 18(1). https://doi.org/10.108 0/21645515.2021.2003647 González, E. R. (1982). TV report on DTP galvanizes US pediatricians. Journal of the American Medical Association, 248(1), 12–22. https://doi.org/10.1001/ JAMA.1982.03330010004002 Grady, D., & Hoffman, J. (2018). HPV vaccine expanded for people ages 27 to 45. The New  York Times. https://www.nytimes.com/2018/10/05/health/ hpv-­virus-­vaccine-­cancer.html

30 

B. L. HOFFMAN ET AL.

Gross, L. (2009). A broken trust: Lessons from the vaccine–autism wars. PLoS Biology, 7(5), e1000114. https://doi.org/10.1371/journal.pbio.1000114 Hajj Hussein, I., Chams, N., Chams, S., El Sayegh, S., Badran, R., Raad, M., et al. (2015). Vaccines through centuries: Major cornerstones of global health. Frontiers in Public Health, 3, 269. https://doi.org/10.3389/ fpubh.2015.00269 Hall, M. T., Simms, K. T., Lew, J. B., Smith, M. A., Brotherton, J. M., Saville, M., et  al. (2019). The projected timeframe until cervical cancer elimination in Australia: A modelling study. The Lancet Public Health, 4(1), e19–e27. https:// doi.org/10.1016/S2468-­2667(18)30183-­X/ Hanley, S. J. B., Yoshioka, E., Ito, Y., & Kishi, R. (2015). HPV vaccination crisis in Japan. The Lancet, 385(9987), 2571. https://doi.org/10.1016/ S0140-­6736(15)61152-­7 Hoffman, B. L., Colditz, J. B., Shensa, A., Wolynn, R., Taneja, S. B., Felter, E. M., et al. (2021). #DoctorsSpeakUp: Lessons learned from a pro-vaccine Twitter event. Vaccine, 39(19). https://doi.org/10.1016/j.vaccine.2021.03.061 Hoffman, B. L., Felter, E. M., Chu, K.-H., Shensa, A., Hermann, C., Wolynn, T., et  al. (2019). It’s not all about autism: The emerging landscape of anti-­ vaccination sentiment on Facebook. Vaccine, 37(16), 2216–2223. https://doi. org/10.1016/j.vaccine.2019.03.003 Hughes, B., Miller-Idriss, C., Piltch-Loeb, R., Goldberg, B., White, K., Criezis, M., & Savoia, E. (2021). Development of a codebook of online anti-­vaccination rhetoric to manage COVID-19 vaccine misinformation. International Journal of Environmental Research and Public Health, 18(14), 7556. https://doi. org/10.3390/IJERPH18147556 Immunization Action Coalition. (2013). Vaccine timeline. Retrieved September 3, 2018, from Immunization Action Coalition website: http://www.immunize.org/timeline/ Islam, M. S., Kamal, A.-H. M., Kabir, A., Southern, D. L., Khan, S. H., Hasan, S. M. M., et al. (2021). COVID-19 vaccine rumors and conspiracy theories: The need for cognitive inoculation against misinformation to improve vaccine adherence. PLoS ONE, 16(5). https://doi.org/10.1371/JOURNAL. PONE.0251605 Kalichman, S. C., Eaton, L. A., Earnshaw, V. A., & Brousseau, N. (2021). Faster than warp speed: Early attention to COVD-19 by anti-vaccine groups on Facebook. Journal of Public Health, 1–10. https://doi.org/10.1093/ PUBMED/FDAB093 Keelan, J., Pavri, V., Balakrishnan, R., & Wilson, K. (2010). An analysis of the human papilloma virus vaccine debate on MySpace blogs. Vaccine, 28(6), 1535–1540. https://doi.org/10.1016/j.vaccine.2009.11.060 Kinch, M. (2018). Between hope and fear: A history of vaccines and human immunity. Pegasus Books.

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

31

Lanzarotta, T., & Ramos, M. A. (2018). Mistrust in medicine: The rise and fall of America’s first vaccine institute. American Journal of Public Health, 108(6), 741–747. https://doi.org/10.2105/AJPH.2018.304348 Mariner, W. K. (1992). Legislative report: The national vaccine injury compensation program. Health Affairs, 11(1), 255–265. https://doi.org/10.1377/ hlthaff.11.1.255 Markowitz, L. E., Dunne, E. F., Saraiya, M., Chesson, H. W., Curtis, C. R., Gee, J., … Unger, E. R. (2014). Human papillomavirus vaccination: Recommendations of the advisory committee on immunization practices (ACIP). Morbidity and Mortality Weekly Report. Recommendations and Reports, 63(RR-05), 1–30. National Archives. (2021). Victorian health reform. https://www.nationalarchives.gov.uk/education/resources/victorian-­health-­reform/ National Conference of State Legislatures. (2021). States with religious and philosophical exemptions from school immunization requirements. http://www. ncsl.org/research/health/school-­immunization-­exemption-­state-­laws.aspx Niederhuber, M. (2014). The fight over inoculation during the 1721 Boston smallpox epidemic. http://sitn.hms.harvard.edu/flash/special-­edition-­on-­ infectious-­disease/2014/the-­fight-­over-­inoculation-­during-­the-­1721-­bostonsmallpox-­epidemic/ Offit, P. A. (2005). The Cutter incident, 50 years later. New England Journal of Medicine, 352(14), 1411–1412. https://doi.org/10.1056/NEJMp048180 Olive, J. K., Hotez, P. J., Damania, A., & Nolan, M. S. (2018). The state of the antivaccine movement in the United States: A focused examination of nonmedical exemptions in states and counties. PLOS Medicine, 15(6), e1002578. https://doi.org/10.1371/journal.pmed.1002578 Parmet, W. E., Goodman, R. A., & Farber, A. (2005). Individual rights versus the public’s health—100 years after Jacobson v. Massachusetts. New England Journal of Medicine, 352(7), 652–654. https://doi.org/10.1056/ NEJMp048209 Perkins, R. B., Pierre-Joseph, N., Marquez, C., Iloka, S., & Clark, J. A. (2010). Why do low-income minority parents choose human papillomavirus vaccination for their daughters? The Journal of Pediatrics, 157(4), 617–622. https:// doi.org/10.1016/j.jpeds.2010.04.013 Pew Research Center. (2019). Social media fact sheet. https://www.pewinternet. org/fact-­sheet/social-­media/ Pingali, C. (2021). National, regional, state, and selected local area vaccination coverage among adolescents aged 13–17 years—United States, 2020. Morbidity and Mortality Weekly Report, 70(35), 1184–1190. https://doi.org/10.15585/ MMWR.MM7035A1 Porter, D., & Porter, R. (1988). The politics of prevention: Anti-vaccinationism and public health in nineteenth-century England. Medical History, 32(3), 231. https://doi.org/10.1017/S0025727300048225

32 

B. L. HOFFMAN ET AL.

Riedel, S. (2005). Edward Jenner and the history of smallpox and vaccination. Proceedings (Baylor University. Medical Center), 18(1), 21–25. Ross, J. (2020, May 2). Coronavirus misinformation crosses divides to infect black social media. NBC News. https://www.nbcnews.com/news/nbcblk/ coronavirus-­m isinformation-­c rosses-­d ivides-­i nfect-­b lack-­s ocial-­m edia­n1198226 Sanz, C. (2020, June 16). Social media platforms are profiting from COVID-19 misinformation: Nancy Pelosi. ABC News. https://abcnews.go.com/Politics/ social-­m edia-­p latforms-­p rofiting-­c ovid-­1 9-­m isinformation-­n ancy/ story?id=71280523 Schmidt, A.  L., Zollo, F., Scala, A., Betsch, C., & Quattrociocchi, W. (2018). Polarization of the vaccination debate on Facebook. Vaccine, 36(25), 3606–3612. https://doi.org/10.1016/J.VACCINE.2018.05.040 Schneider, W. H. (2009). Smallpox in Africa during colonial rule. Medical History, 53(2), 193. https://doi.org/10.1017/S002572730000363X Shah, S.  F. A., Ginossar, T., & Weiss, D. (2019). “This is a Pakhtun disease”: Pakhtun health journalists’ perceptions of the barriers and facilitators to polio vaccine acceptance among the high-risk Pakhtun community in Pakistan. Vaccine, 37(28), 3694–3703. https://doi.org/10.1016/J. VACCINE.2019.05.029 Skinner, S. R., Davies, C., Cooper, S., Stoney, T., Marshall, H., Jones, J., et al. (2015). HPV.edu study protocol: A cluster randomised controlled evaluation of education, decisional support and logistical strategies in school-based human papillomavirus (HPV) vaccination of adolescents. BMC Public Health, 15(1). https://doi.org/10.1186/S12889-­015-­2168-­5 Stern, A. M., & Markel, H. (2005). The history of vaccines and immunization: Familiar patterns, new challenges. Health Affairs, 24(3), 611–621. https:// doi.org/10.1377/hlthaff.24.3.611 Szilagyi, P.  G., Albertin, C.  S., Gurfinkel, D., Saville, A.  W., Vangala, S., Rice, J.  D., et  al. (2020). Prevalence and characteristics of HPV vaccine hesitancy among parents of adolescents across the US. Vaccine, 38(38), 6027–6037. https://doi.org/10.1016/j.vaccine.2020.06.074 Tanaka, Y. (2020). Time to resume active recommendation of the HPV vaccine in Japan. The Lancet Oncology, 21(12), 1552–1553. https://doi.org/10.1016/ S1470-­2045(20)30608-­2 Tanaka, Y., Ueda, Y., Yoshino, K., & Kimura, T. (2017). History repeats itself in Japan: Failure to learn from rubella epidemic leads to failure to provide the HPV vaccine. Human Vaccines & Immunotherapeutics, 13(8), 1859–1860. https://doi.org/10.1080/21645515.2017.1327929 Thoman, E., & Jolls, T. (2005). Literacy for the 21st century: An overview & orientation guide to media literacy education. https://www.medialit.org/ sites/default/files/01_MLKorientation.pdf

2  VACCINE MISINFORMATION ON SOCIAL MEDIA: HISTORICAL CONTEXTS… 

33

Valentino, K., & Poronsky, C.  B. (2016). Human papillomavirus infection and vaccination. Journal of Pediatric Nursing, 31(2), e155–e166. https://doi. org/10.1016/j.pedn.2015.10.005 Venkatraman, A., Garg, N., & Kumar, N. (2015). Greater freedom of speech on Web 2.0 correlates with dominance of views linking vaccines to autism. Vaccine, 33(12), 1422–1425. https://doi.org/10.1016/j.vaccine.2015.01.078 Walker, T. Y., Elam-Evans, L. D., Yankey, D., Markowitz, L. E., Williams, C. L., Mbaeyi, S. A., et al. (2018). National, regional, state, and selected local area vaccination coverage among adolescents aged 13–17 years—United States, 2014. Morbidity and Mortality Weekly Report, 67(33), 909–917. https://doi. org/10.15585/mmwr.mm6733a1 Warraich, H.  J. (2009). Religious opposition to polio vaccination. Emerging Infectious Diseases, 15(6), 978. https://doi.org/10.3201/eid1506.090087 Wolfe, R. M., & Sharp, L. K. (2002). Anti-vaccinationists past and present. British Medical Journal Clinical Research Edition, 325(7361), 430–432. World Health Organization. (2019). Ten threats to global health in 2019. https:// www.who.int/emergencies/ten-­threats-­to-­global-­health-­in-­2019 Zarocostas, J. (2020). How to fight an infodemic. Lancet, 395(10225), 676. https://doi.org/10.1016/S0140-­6736(20)30461-­X Zipprich, J., Winter, K., Hacker, J., Xia, D., Watt, J., Harriman, K., et al. (2015). Measles outbreak—California, December 2014–February 2015. Morbitity and Morality Weekly Report, 64(6), 153–154.

CHAPTER 3

HPV Vaccine Misinformation Online: A Narrative Scoping Review Yuan Wang, Kathryn Thier, and Xiaoli Nan

Human papillomavirus (HPV) is a common, sexually transmitted virus with several strains responsible for nearly all cervical cancer and many anogenital and oropharyngeal cancers worldwide (de Martel et  al., 2017). Since 2006, several HPV vaccines have shown marked success in decreasing HPV infection and related cancers (Bruni et  al., 2016; Centers for Disease Control and Prevention [CDC], 2021a). The CDC recommends all girls and boys begin the HPV vaccine series at ages 11–12 and catch-up vaccination for everyone through age 26 (CDC, 2021b). Yet HPV vaccine uptake remains low compared with other childhood and adolescent vaccines (Bruni et al., 2016; Pingali, 2021). The consequences of low HPV uptake are dire. The CDC estimates that almost 92% of the cancers in the U.S. attributed to HPV each year from 2013 to 2017 could have been prevented by the 9-valent HPV vaccine (CDC, 2020). Misinformation and misconceptions about the HPV vaccine are a key factor in sub-optimal HPV vaccine uptake in many countries (Zimet et al.,

Y. Wang (*) • K. Thier • X. Nan Department of Communication, University of Maryland, College Park, MD, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 T. Ginossar et al. (eds.), Vaccine Communication Online, https://doi.org/10.1007/978-3-031-24490-2_3

35

36 

Y. WANG ET AL.

2013). The relative newness of the HPV vaccine and its relation to prevention of sexually transmitted diseases has led to confusion and misinformation about its safety, effectiveness, and applicability for both genders (Cheruvu et al., 2017; Dibble et al., 2019; Hofstetter & Rosenthal, 2014; Zimet et al., 2013). Studies suggest that misinformation about the HPV vaccine is prevalent online, especially on social media (Ekram et al., 2019; Massey et al., 2020). In this chapter, we review past and current research on online HPV vaccine misinformation, mapping prevalence of misinformation across various platforms and describing key message and diffusion characteristics. We then review evidence about factors that predict individual susceptibility to online HPV vaccine misinformation. Next, we detail the impact of online HPV vaccine misinformation on individuals’ vaccine-related beliefs, attitudes, and actions, and examine interventions designed to mitigate the impact of online HPV vaccine misinformation. We conclude with a discussion of the current state of knowledge on online HPV vaccine misinformation and offer directions for future research.

Prevalence of Online HPV Vaccine Misinformation In 2021, 86% of U.S. adults said they often or sometimes get information online, whether through websites, search engines, social media, or podcasts (Shearer, 2021). Although online platforms provide convenient access to health information, they also facilitate the proliferation of misinformation (Chou et al., 2020). Over the past decade, a growing body of literature suggests that misinformation about HPV vaccines is prevalent online, via websites, social media, and emerging technologies such as virtual assistants (Alagha & Helbing, 2019; Massey et al., 2020; Shearer, 2021). News websites and search engines have become Americans’ most preferred digital information sources (Shearer, 2021). Studies suggest that most online articles about HPV vaccines have a positive or neutral tone, yet misinformation also abounds. For example, Madden et al. (2012) conducted a content analysis of 89 top search results about HPV vaccines from four leading search engines at the time, including Google, Yahoo, Bing, and Ask.com. They found that although most of the websites were supportive or neutral toward HPV vaccines, 7.9% were opposed to HPV vaccines (e.g., “The truths about Gardasil and cervical cancer are suppressed”; p.  3743). Moreover, 2% of the websites contained conspiracy theories related to HPV vaccines, 6.7% indicated low HPV vaccine effectiveness, and 12.4% indicated high risks associated with HPV vaccines. In

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

37

another study about HPV vaccine-related online news, Mahoney et  al. (2015) found that 30% of Google News articles and 17% of news articles circulated on Twitter discouraged HPV vaccination. Moreover, after Michele Bachmann falsely claimed that HPV vaccines could cause mental retardation during a guest appearance on the Today Show in 2011, online news articles shifted from disseminating scientific facts about HPV vaccines to political debates (Mahoney et al., 2015). More concerns have been expressed regarding the prevalence of health misinformation on social media. Social media allow non-experts to generate content and increase information bubbles through automated algorithms, consequently fueling the creation and diffusion of misinformation (Chou et al., 2020). Numerous studies suggest that health misinformation, including misinformation about HPV vaccines, is common on various social media platforms. On Twitter, Chakraborty et al. (2017) found that 5% of HPV vaccine-related tweets expressed negative opinions toward the vaccins. Moreover, 4.1% of tweets explicitly claimed that HPV vaccines are unsafe (e.g., “Girl dies shortly after receiving HPV vaccine”), and 0.9% said that HPV vaccines could increase risky sexual behaviors (e.g., “HPV vaccines make you promiscuous”; p.  5). In another study, about HPV vaccine-related posts on Instagram, Massey et al. (2020) found that 44.1% of these posts were against HPV vaccines. Among the anti-vaccine posts, about three-quarters (72.3%) provided unsubstantiated claims (e.g., “A few years later we had a patient die after his shots—they called it a SIDS [sudden infant death syndrome] death, at 2 in the afternoon, a few hours after his vaccines”); over half (56.3%) contained conspiracy theories (e.g., “The government knows these risks exist, doctors know these risks exist yet, they still administer vaccines because of money!”); more than one in four (28.1%) falsely claimed HPV vaccines are ineffective; and a majority mentioned the severity (80.1%) and/or likelihood (63.7%) of vaccine-­ related injuries (Massey et  al., 2020, Appendix A). Regarding HPV vaccine-­related videos on YouTube, Ekram et al. (2019) found that most videos were negative toward HPV vaccines (64.5%). Compared to pro-­ vaccine videos, anti-vaccine videos were more likely to omit information or report inaccurate information about HPV vaccines (Ekram et  al., 2019). Health misinformation can also be found in newer online tools. Virtual assistants (also called “voice assistants”), such as Apple’s Siri, are software agents that can respond to human speech with synthesized voices (Hoy, 2018, p. 81). A survey conducted by Pew Research Center (2017) found that 46% of U.S. adults were users of virtual assistants. Although

38 

Y. WANG ET AL.

virtual assistants have become an increasingly popular sources of information, studies suggest that their information quality varies across different platforms. For example, Alagha and Helbing (2019) evaluated virtual assistants’ responses to vaccine-related questions, finding that Google Assistant and Siri accurately answered about 85% of questions about vaccines, whereas Amazon’s Alexa had only a 16% accuracy rate. In another study about virtual assistants’ responses to questions about HPV vaccines, Ferrand et al. (2020) found that 5.6% of responses by Siri and 26.2% of responses by Microsoft Cortana contained misinformation about HPV vaccines. Overall, our review suggests that misinformation about HPV vaccines is often prevalent among various online platforms, including websites, news aggregators, social media, and virtual assistants. Misinformation about HPV vaccines is especially common on social media sites such as YouTube and Instagram. It should be noted, however, that both YouTube and Instagram recently took actions to remove or limit discoverability of posts spreading vaccine misinformation (Milmo, 2021; Strozewski, 2021). Still, emerging evidence suggests that anti-vaccine messages were not taken off social media platforms in a timely manner (e.g., Ginossar et al., 2022). More research is needed to assess whether these interventions effectively reduce HPV vaccine misinformation on these platforms.

Characteristics of Online HPV Vaccine Misinformation In addition to monitoring the prevalence of HPV vaccine misinformation online, emerging studies have also attempted to identify the characteristics of online HPV vaccine misinformation (e.g., Briones et al., 2012; Chin et al., 2020; Fu et al., 2016). Understanding the typical elements and diffusion patterns of online HPV vaccine misinformation enables the public to spot misinformation (Massey et al., 2020). Below, we review the content characteristics (i.e., message features of misinformation) and diffusion characteristics (i.e., how misinformation is spread) of online HPV vaccine misinformation. With respect to content characteristics, research shows that HPV vaccine misinformation rarely cites scientific facts; instead, it often relies on conspiracy theories and personal anecdotes. For example, focusing on HPV vaccine-related videos on YouTube, Briones et al. (2012) found that

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

39

8.7% of the videos mentioned a conspiracy theory about HPV vaccines, involving the government, pharmaceutical companies, or doctors. In addition, Raghupathi et al. (2020, p. 7) found that “vaccines cause autism” is the most frequently used claim among vaccine-related tweets. In a study about HPV vaccine misinformation on Instagram, Massey et  al. (2020) found that anti-vaccine posts frequently used concealment and distortion as strategies. Concealment posts purport to reveal a lie or unknown facts about HPV vaccines (e.g., “a new study reveals hiding risks of HPV vaccines”), and distortion posts mispresent original information to imply a causal link between receiving HPV vaccines and injuries (Massey et  al., 2020, p. 9). In addition, compared to pro-vaccine messages, misinformation about HPV vaccines was more likely to use personal stories or anecdotal evidence to discourage HPV vaccination (Massey et  al., 2020). On Twitter, the most viral HPV misinformation tweets included a tweet mentioning the existence of toxic ingredients in HPV vaccines, a tweet citing a personal story to claim that HPV vaccines lead to death, and a tweet using a personal story to claim that HPV vaccines cause brain damage (Chin et al., 2020). With regard to sentiments, studies suggest that HPV vaccine misinformation is often associated with negative sentiments (e.g., Ekram et al., 2019; Zhang, Xue, et al., 2021). Xu and Guo (2018), for example, analyzed the discrete emotions associated with anti-vaccine messages, finding that anti-vaccine headlines in online articles were more likely to use words associated with sadness, fear, and anger than pro-vaccine headlines. Moreover, compared to factual information, HPV vaccine misinformation was more likely to originate from individuals who were not health experts (Massey et al., 2020). Compared to pro-vaccine messages, HPV vaccine misinformation was less likely to contain detailed author information, such as location information (Massey et  al., 2020) or an author byline (Fu et al., 2016). What do we know about the diffusion characteristics of online HPV vaccine misinformation? Although it is not clear whether health misinformation in general receives more social media engagement than scientific information (Nan et  al., 2021), evidence in the context of vaccination tends to suggest that vaccine misinformation is more likely to be shared, liked, and commented on than pro-vaccine messages. For example, online articles with anti-vaccine headlines received more shares, likes, and comments on various social media platforms (Xu & Guo, 2018). Briones et al. (2012) found that YouTube videos that disapproved of HPV vaccines

40 

Y. WANG ET AL.

received more likes than videos that advocated for HPV vaccines or ambiguous videos that mentioned both pros and cons of HPV vaccines. Chin et  al. (2020) examined the dissemination patterns of factual claims and misinformation about HPV vaccines on Twitter from 2013 to 2017, finding that misinformation (e.g., “HPV vaccines lead to death”) spreads wider than factual messages (e.g., “HPV vaccines can prevent cancer”) (p. 314). Moreover, the dissemination of misinformation about HPV vaccines followed a long-tail distribution, with fewer than 10% of the misinformation tweets accounting for most of the shares among HPV vaccine-related misinformation (Chin et al., 2020). This finding suggests that although most messages containing HPV vaccine misinformation receive few shares, some of these misleading messages could become extremely viral. Despite the evidence, a recent study (Himelboim et al., In press) found that pro-vaccine tweets that debunked conspiracy theories received more social media engagement than anti-vaccine tweets that promoted conspiracy theories. Still, the authors also suggested that references to conspiracy theories were linked to more engagement only when discussing early childhood vaccines, not in the context of HPV vaccines. Overall, as the above studies suggest, online HPV vaccine misinformation often cites conspiracy theories, relies on personal stories as evidence, is authored by non-experts, lacks detailed author information, and tends to be associated with negative emotions. Compared to pro-vaccine messages, HPV vaccine misinformation receives more shares, likes, and comments, suggesting that misinformation about HPV vaccines might reach more audiences than factual claims.

Susceptibility to Online HPV Vaccine Misinformation Research on individual difference in susceptibility to online health misinformation, in general and for HPV vaccination, is lacking, despite concerns about changes in media consumption and dissemination of health misinformation wrought by the digital revolution (Wang et  al., 2019). Studies about health knowledge, attitudes, and barriers among racial and ethnic minorities note the presence of health misinformation in these populations and often mention historical mistrust of the medical establishment (e.g., Brooks et  al., 2009), but more research is needed on how individual characteristics are associated with susceptibility to misinformation.

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

41

As gatekeepers for their children’s healthcare, parents are exposed to vaccine misinformation as they pursue information from providers and media sources to guide their decision-making. Parents concerned about routine vaccines seek more information than vaccine-confident patents, are less likely to trust their doctors, and are more likely to believe sources questioning vaccine safety (Ames et al., 2017). Generally, individuals with pre-existing negativity toward vaccines perceive medically accurate and misinformation as equally factual and scientific and do not differentially evaluate source trustworthiness or expertise for misinformed and scientifically accurate sources (Shen & Zhou, 2020). Misinformation is linked to vaccine hesitancy. Vaccine misinformation was the most common reason to not vaccinate daughters against HPV given by American parents in a 2009–2012 survey (Cheruvu et al., 2017). Providers appear to play a role in decreasing misinformation susceptibility, as parents who did not receive a provider recommendation for the vaccine series 18% more likely to report misinformation as a factor for not intending to vaccinate their daughter (Cheruvu et  al., 2017). Additionally, increased number of people living in the household was associated with greater odds of listing vaccine misinformation as a reason for not vaccinating one’s daughter (Cheruvu et al., 2017). A possible reason is that people are more likely to believe in online HPV vaccine misinformation when their close networks hold such misperceptions. While lower income and education levels typically predict whether individuals find low-quality health information credible (Benotsch et  al., 2004), the limited research about susceptibility to HPV vaccine online misinformation is mixed. Being unemployed increased beliefs in rumors and misinformation, such as the idea that the HPV vaccine destroys female fertility or promotes premarital sex, while trust in health providers for accurate information decreased belief in misinformation among Malawian caregivers (Adeyanju et al., 2021). However, survey participants reported relying on social media for less than 1% of their health information and reported low trust in that source. In an experiment manipulating the veracity of social media posts, Scherer et al. (2021) found that lower educational achievement and health literacy, distrust in the health care system, and positive attitudes toward alternative health predicted susceptibility to HPV vaccine misinformation among US adults aged 40–80, with those susceptible to such misinformation also likely to believe misinformation about statin medications and cancer treatment. Those who spent more hours per day on social media,

42 

Y. WANG ET AL.

were younger, and had lower household income were more likely to perceive social media misinformation about the HPV vaccine, statins, and cancer treatment as accurate and influential (Scherer et al., 2021). Relevant health experiences (such as cancer), health insurance status, and having an HPV vaccine-age child did not predict susceptibility to social media HPV vaccine misinformation (Scherer et al., 2021). By contrast, Chen et al. (2021) concluded that low knowledge about the HPV vaccine, independent of educational background, increased susceptibility to HPV vaccine online misinformation among Chinese adults aged 18–35. In their experiment, participants with low knowledge about the HPV vaccine exposed to conspiracy theories drawn from social media were more likely to express negative attitudes and lower vaccination intentions than those with higher HPV vaccine knowledge, suggesting low knowledge about the HPV vaccine increases susceptibility to misinformation (Chen et al., 2021). Overall, evidence about the socio-demographic, psychological, and media usage factors that influence susceptibility to online HPV vaccine misinformation is limited. Research shows that some associated factors are lower educational achievement, health literacy, and income; medical mistrust, positive attitudes toward alternative health, low knowledge about the HPV vaccine, youth, time spent on social media, belief in other health misinformation, and, possibly, lack of provider recommendation.

Impact of Online HPV Vaccine Misinformation Despite a significant decrease in the rate of adverse events following receipt of the HPV vaccine between 2015 and 2018, the proportion of parents and caregivers who refused the HPV vaccine for their children due to safety concerns increased by 79.9% during the same time frame (Sonawane et  al., 2021). The increase in parental safety concerns disconcertingly coincided with a rise in negative content related to the HPV vaccine on social media between 2015 and 2017 (Chin et al., 2020; Luo et al., 2019). As public health professionals become increasingly alarmed about the potentially detrimental impact of online HPV vaccine misinformation on public acceptance of the vaccine, evidence showing a causal impact of exposure to online HPV vaccine misinformation on HPV vaccine uptake is still lacking (Ortiz et al., 2019). Nevertheless, emerging evidence suggests that exposure to online HPV vaccine misinformation can have a significant impact on the psychological antecedents (e.g., beliefs and attitudes

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

43

toward HPV vaccines) to vaccine uptake (e.g., Chen et al., 2021; Nan & Madden, 2012). In one of the earliest studies that experimentally tested the impact of HPV vaccine misinformation online on vaccine acceptance (Nan & Madden, 2012), college students were randomly assigned to view an online blog that shared misinformation about the HPV vaccine, linking HPV vaccination to blindness, or a blog that shared accurate, scientific information about the vaccine. Compared to those in the control condition where no blog was viewed, participants who viewed the negative blog containing misinformation reported less confidence in the vaccine’s efficacy and safety, less favorable attitudes toward the vaccine, and lower intentions to get vaccinated. The scientific blog, in contrast, made no difference in beliefs, attitudes, or intentions relative to the control condition. This study demonstrated the powerful impact of misinformation on the psychological antecedents to HPV vaccine uptake. In another study (Calo et al., 2021), parents were invited to view one tweet about the HPV vaccine that either contained misinformation or no misinformation. The tweets also varied in other characteristics such as the source, narrative style, and topic. Parents who saw a tweet with misinformation reported significantly lower motivation to vaccinate their children compared to those who saw a tweet with no misinformation. Tweets with misinformation, compared to those with no misinformation, significantly reduced trust in the message content. Results of this study mirror those from Nan and Madden (2012), showing the clear impact of online misinformation on the psychological antecedents to HPV vaccine uptake. Correlational data also corroborated the links between exposure to online HPV vaccine misinformation and HPV vaccine uptake or its psychological antecedents (e.g., Basch & MacLean, 2019; Hansen & Schmidtblaicher, 2021). For example, one study examined 83,551 tweets posted or reposted by 30,621 users between October 2013 and April 2014 (Dunn et  al., 2015). A key observation from this study was that exposure to a majority of negative opinions (often containing misinformation) online about the HPV vaccine significantly increased a user’s likelihood of posting a negative tweet (37.78%), compared to exposure to a majority of positive and neutral tweets (10.92%). Posting a negative tweet about the HPV vaccine is not equivalent to rejecting the vaccine, of course. However, willingness to post a negative tweet indicates negative positions on key psychological antecedents to vaccine uptake such as attitudes and intentions toward getting the vaccine.

44 

Y. WANG ET AL.

Another study provides compelling evidence for the relationship between exposure to online HPV vaccine misinformation and HPV vaccine uptake, by analyzing 273.8  million exposures to 258,418 tweets posted from 2013 to 2015 (Dunn et al., 2017). Results indicated that HPV vaccine coverage was lower in states where higher proportions of exposures were directed to misleading tweets expressing safety concerns, relaying misinformation, and spreading conspiracy theories. Overall, there is no strong evidence that exposure to online HPV vaccine misinformation leads to vaccine rejection or lower vaccination rates. However, both experimental and correlational studies support the causal links between exposure to online HPV vaccine misinformation and psychological antecedents to vaccine uptake. These findings are similar to current evidence concerning the overall impact of health misinformation (Nan et al., 2021).

Mitigating the Impact of Online HPV Vaccine Misinformation Despite limited evidence that online HPV vaccine misinformation in fact leads to reduction in HPV vaccination rates, public health researchers and professionals alike are rightly concerned about the potential of online misinformation to compromise HPV vaccination efforts. First, there is strong evidence that exposure to online misinformation about the HPV vaccine results in less desirable beliefs, attitudes, and intentions toward HPV vaccination (Calo et  al., 2021; Nan & Madden, 2012). Second, tracking actual behavior has been a challenge in past research; as more studies attempt to examine the link between exposure to online HPV vaccine misinformation and HPV vaccine uptake, new evidence may emerge that justifies broad concerns about online misinformation surrounding the vaccine. As part of the effort to combat the actual and potential negative impact of online HPV vaccine misinformation on public health, emerging studies have examined ways of mitigating the influence of misinformation with one approach involving correcting misinformation through strategic communication (e.g., Borah et  al., 2021; Vraga et  al., 2019) and another approach involving inoculating against the influence of misinformation (e.g., Wong, 2016). Concerned about HPV vaccine misinformation on social media, Vraga et al. (2019) tested the utility of logic-based (i.e., an

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

45

infographic visualizing the logical fallacy) and humor-based (i.e., a humorous cartoon illustrating the flawed logic) messages in correcting the misinformation that the HPV vaccine causes auto-immune disorders on Twitter. Results showed that both the logic-based and humor-based corrective messages were effective in reducing misperceptions compared to the control condition with no correction. Further, the logic-based message was a more effective form of correction than the humor-based message. Another study on correcting HPV vaccine misinformation on Twitter (Kim et al., 2020) corroborated the previous finding that a non-humorous corrective message was more effective than a humorous corrective message in reducing misperceptions. Both studies show that it is possible to mitigate the impact of online HPV vaccine misinformation through the use of strategic corrective messages. Wong (2016) used a prewarning approach to inoculate individuals against HPV vaccine misinformation, finding that both a general inoculation (i.e., all vaccines are safe and effective) and a specific inoculation (i.e., the HPV vaccine is safe and effective) were effective at thwarting future attacks on the HPV vaccine. Other studies explored efforts in reducing misperceptions about HPV vaccination through social media messages, comparing gain- vs. loss-­ framed messages (Borah et al., 2021) and one-sided vs. two-sided messages (Xiao & Su, 2021). Borah et  al. (2021) found that a corrective message associated with a loss frame (e.g., emphasizing the costs of not getting vaccinated) was significantly more effective in reducing misperceptions than a similar corrective message associated with a gain frame (e.g., emphasizing the benefits of getting vaccinated), although such effects might be limited only to individuals low in reflection, or who have the ability to connect new information to existing cognitive frameworks. Xiao and Su (2021) revealed a complex pattern of results where the use of a refutational two-sided message (i.e., presenting arguments both for and against HPV vaccination in addition to an argument refuting the counterargument) outperformed a one-sided message (i.e., presenting arguments for HPV vaccination only) for individuals with low prior misperceptions about the vaccine, whereas the one-sided message appeared to be more effective than the refutational two-sided message for those with high prior misperceptions about the vaccine. A large randomized controlled trial (Zhang, Featherstone, et al., 2021) examined the effectiveness of fact-checking labels associated with different sources in mitigating the impact of misinformation about various types of vaccines including the HPV vaccine. Regardless of the sources associated

46 

Y. WANG ET AL.

with the fact-checking labels, exposure to the labels vs. misinformation significantly increased positive attitudes toward the vaccine. Among the sources that generated the fact-checking, health institutions and research universities were ranked as having higher expertise than other sources, including algorithms, news media, and fact-checking organizations. Overall, a growing number of studies have examined strategies for mitigating the impact of HPV vaccine misinformation online. Some strategies are directed at discrediting misinformation, including the use of corrective messages and fact-checking labels (e.g., Vraga et  al., 2019; Zhang, Featherstone, et al., 2021), and some are focused on the use of persuasive messages in reducing misperceptions (e.g., Borah et al., 2021; Xiao & Su, 2021). Corrective messages are generally effective in reducing the impact of HPV vaccine misinformation on vaccine beliefs, as is consistent with the efficacy of corrective messages in reducing the impact of online health misinformation in general (Walter et al., 2021). The impact of persuasive messages on misperceptions about the HPV vaccine tends to be qualified by message recipient characteristics.

Discussion Summary of Findings In this chapter we provide a narrative scoping review of past and current research on online HPV vaccine misinformation. The broad themes we identified in the literature include research that documents the prevalence of misinformation about HPV vaccines on various online media platforms, analyses that focus on identifying the unique content and diffusion characteristics of HPV vaccine misinformation online, studies that seek to uncover individual characteristics predicting susceptibility to online HPV vaccine misinformation, research that assesses the link between exposure to misinformation about HPV vaccines online and vaccine-related beliefs, attitudes, and behaviors, and, finally, experiments and correlational studies that evaluate interventions designed to mitigate the impact of online HPV vaccine misinformation. Because of the relative newness of the HPV vaccine and the role of online media in propagating misinformation that only recently began to capture researchers’ attention, most studies reviewed in this chapter were conducted after 2010, with accelerated research in this area emerging following the 2016 presidential election, when HPV vaccination was a subject of intense political debate.

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

47

First, in terms of misinformation prevalence, our review suggests that misinformation about HPV vaccines is commonly found on online platforms such as websites, news aggregators, and social media, and newer online tools such as virtual assistants. Misinformation about HPV vaccines seems to be especially concentrated on social media platforms where there is virtually no content gatekeeping. Past studies sounded alarms about the proliferation of vaccine misinformation in general, and HPV vaccine misinformation in particular, on social media sites such as YouTube and Instagram. In response to the public outcry over health misinformation on social media, major social media companies including YouTube, Instagram, Facebook, and Pinterest all took steps to reduce the dissemination of vaccine misinformation on their platforms. It remains to be seen whether their actions can effectively curb vaccine misinformation proliferation online. Second, studies have examined how health misinformation differs from scientific information in terms of its message characteristics and how it is diffused online. Our review suggests that online HPV vaccine misinformation is often characterized by the use of conspiracy theories and personal stories, is authored by non-experts or lacks detailed author information, and tends to induce negative emotions. Moreover, compared with accurate information, misinformation circulates more widely and potentially exerts more impact through receiving more shares, likes, and comments. Understanding the unique content and diffusion characteristics of online health misinformation is important for enabling the public to spot such information and devising effective interventions to reduce its impact. Third, research on the sociodemographic, psychological, and media usage factors that drive susceptibility to online HPV vaccine misinformation is beginning to emerge. Available evidence suggests that believing in HPV vaccine misinformation is predicted by socio-demographic factors such as lower educational achievement, lower income, and younger age, and health-related beliefs such as medical mistrust, positive attitudes toward alternative health, and belief in other health misinformation. Low knowledge about the HPV vaccine and health literacy also plays a role in increasing susceptibility to online HPV vaccine misinformation. More time spent on social media and a lack of provider recommendation are also risk factors. Understanding psychological and behavioral drivers of susceptibility to misinformation is important for developing interventions to counter misinformation online.

48 

Y. WANG ET AL.

Despite concerns about online vaccine misinformation, there is no strong evidence that exposure to online vaccine misinformation leads to vaccine rejection or lower vaccination rates. This is also the case for HPV vaccines. Our review finds no study that directly links exposure to online HPV vaccine misinformation to vaccine resistance. However, both experimental and correlational studies have supported a strong relationship between exposure to online HPV vaccine misinformation and psychological antecedents to vaccine uptake. To the extent that psychological antecedents predict vaccine uptake, it is possible for online HPV vaccine misinformation to have an indirect effect on vaccination behavior. Finally, a growing number of studies have examined strategies for mitigating the impact of HPV vaccine misinformation online. A popular strategy is the use of corrective messages or fact-checking labels after individuals have been exposed to misinformation (or while they are processing misinformation). Research suggests that corrective messages are generally effective in reducing false beliefs about HPV vaccines, compared to misinformation with no accompanying correction. Another approach uses persuasive messages to directly target false beliefs in HPV vaccines, where misinformation either is not presented or is peripheral to the message, and its effectiveness tends to be qualified by message recipient characteristics. Directions for Future Research Future research should continue to monitor the prevalence and characteristics of online HPV vaccine misinformation. As new social networking platforms such as TikTok and Clubhouse become increasingly popular, more research is needed to investigate whether HPV vaccine misinformation is prevalent there. Another fruitful direction for future research is to examine the linkage between online HPV vaccine misinformation’s content characteristics and its virality. Most current studies focus on comparing HPV vaccine misinformation with scientific information; little is known about whether misinformation with certain features is more viral than others. For example, does narrative-based HPV vaccine misinformation spread more widely than that based on pseudoscientific claims? Is HPV vaccine misinformation from certain sources more likely to go viral than from others? Identifying the characteristics of viral HPV vaccine misinformation will inform future interventions to inoculate publics against such misinformation.

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

49

Further, more research is needed on how individual characteristics predict susceptibility to online HPV vaccine misinformation. Limited studies have examined who is more likely to consume and/or believe in online HPV vaccine misinformation. Emerging evidence suggests that socio-­ demographic factors and psychological factors such as conservativism, conspiracy thinking mindset, and religiosity are associated with susceptibility to misinformation (e.g., Featherstone et al., 2019; Porshnev et al., 2021; Stefanone et al., 2019). Future research may seek to replicate these findings in the context of HPV vaccine misinformation. Understanding who is more vulnerable to online HPV vaccine misinformation is important for designing tailored interventions and addressing disparities in HPV vaccine uptake. Finally, future studies should continue to explore which interventions can effectively mitigate the negative impact of online HPV vaccine misinformation and identify the source, recipient, and message factors that strengthen or weaken intervention effectiveness. In addition, a potential avenue for future research is to examine how online HPV vaccine misinformation and correction interventions impact actual vaccine uptake. Evidence is lacking on the causal link between exposure to online HPV vaccine misinformation and vaccine refusal. Similarly, little is known about whether exposure to interventions that prewarn or debunk HPV misinformation can increase HPV vaccination rates.

Concluding Remarks HPV vaccines are safe and remarkably effective against HPV infection, a primary risk factor for many types of cancer. However, HPV vaccine uptake rates among adolescents remain much lower than those for other routine vaccines, partly due to vaccine hesitancy fueled by online misinformation. Emerging research has provided important insights on the prevalence and characteristics of online HPV vaccine misinformation, improved our understanding of the individual characteristics that make one more susceptible to such information, informed us about how exposure to online HPV vaccine misinformation might jeopardize vaccination rates, and have shown us what types of interventions might be effective in mitigating the negative impact of such information. With health misinformation becoming a major public health concern amidst the COVID-19 pandemic, we need to more systematically address the problem of misinformation across all health domains, especially contentious areas such as vaccination.

50 

Y. WANG ET AL.

References Adeyanju, G. C., Sprengholz, P., Betsch, C., & Essoh, T.-A. (2021). Caregivers’ willingness to vaccinate their children against childhood diseases and human papillomavirus: A cross-sectional study on vaccine hesitancy in Malawi. Preprint (Version 1), Research Square. https://doi.org/10.21203/rs.3.rs-­618575/v1 Alagha, E. C., & Helbing, R. R. (2019). Evaluating the quality of voice assistants’ responses to consumer health questions about vaccines: An exploratory comparison of Alexa, Google Assistant and Siri. BMJ Health & Care Informatics, 26(1), e100075. https://doi.org/10.1136/bmjhci-­2019-­100075 Ames, H. M., Glenton, C., & Lewin, S. (2017). Parents’ and informal caregivers’ views and experiences of communication about routine childhood vaccination: A synthesis of qualitative evidence. Cochrane Database of Systematic Reviews. https://doi.org/10.1002/14651858.CD011787.pub2 Basch, C. H., & MacLean, S. A. (2019). A content analysis of HPV related posts on Instagram. Human Vaccines & Immunotherapeutics, 15(7–8), 1476–1478. https://doi.org/10.1080/21645515.2018.1560774 Benotsch, E. G., Kalichman, S., & Weinhardt, L. S. (2004). HIV-AIDS patients’ evaluation of health information on the Internet: The digital divide and vulnerability to fraudulent claims. Journal of Consulting and Clinical Psychology, 72(6), 1004–1011. https://doi.org/10.1037/0022-­006X.72.6.1004 Borah, P., Kim, S., Xiao, X., & Lee, D. K. L. (2021). Correcting misinformation using theory-driven messages: HPV vaccine misperceptions, information seeking, and the moderating role of reflection. Atlantic Journal of Communication, 0(0), 1–17. https://doi.org/10.1080/15456870.2021.1912046 Briones, R., Nan, X., Madden, K., & Waks, L. (2012). When vaccines go viral: An analysis of HPV vaccine coverage on YouTube. Health Communication, 27(5), 478–485. https://doi.org/10.1080/10410236.2011.610258 Brooks, M. M., Paschal, A. M., Sly, J. R., & Hsiao, T. (2009). African American women and clinical trials: Perceived barriers to participation and potential solutions. American Journal of Health Studies, 24(2), 298–305. Bruni, L., Diaz, M., Barrionuevo-Rosas, L., Herrero, R., Bray, F., Bosch, F. X., de Sanjosé, S., & Castellsagué, X. (2016). Global estimates of human papillomavirus vaccination coverage by region and income level: A pooled analysis. The Lancet Global Health, 4(7), e453–e463. https://doi.org/10.1016/ S2214-­109X(16)30099-­7 Calo, W.  A., Gilkey, M.  B., Shah, P.  D., Dyer, A.-M., Margolis, M.  A., Dailey, S.  A., & Brewer, N.  T. (2021). Misinformation and other elements in HPV vaccine tweets: An experimental comparison. Journal of Behavioral Medicine, 44(3), 310–319. https://doi.org/10.1007/s10865-­021-­00203-­3 Centers for Disease Control and Prevention. (2020, September 3). Cancers associated with the human papillomavirus, United States—2013–2017. US

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

51

Department of Health and Human Service. https://www.cdc.gov/cancer/ u s c s / a b o u t / d a t a -­b r i e f s / n o 1 8 -­h p v -­a s s o c -­c a n c e r s - U n i t e d S t a t e s 2013-­2017.htm Centers for Disease Control and Prevention. (2021a, July 23). HPV vaccination is safe and effective. https://www.cdc.gov/hpv/parents/vaccinesafety.html Centers for Disease Control and Prevention. (2021b, July 23). HPV vaccine. https://www.cdc.gov/hpv/parents/vaccine-­for-­hpv.html Chakraborty, P., Colditz, J. B., Silvestre, A. J., Friedman, M. R., Bogen, K. W., Primack, B.  A., & Lee, A. (2017). Observation of public sentiment toward human papillomavirus vaccination on Twitter. Cogent Medicine, 4(1), 1390853. https://doi.org/10.1080/2331205X.2017.1390853 Chen, L., Zhang, Y., Young, R., Wu, X., & Zhu, G. (2021). Effects of vaccine-­ related conspiracy theories on Chinese young adults’ perceptions of the HPV vaccine: An experimental study. Health Communication, 36(11), 1343–1353. https://doi.org/10.1080/10410236.2020.1751384 Cheruvu, V. K., Bhatta, M. P., & Drinkard, L. N. (2017). Factors associated with parental reasons for “no-intent” to vaccinate female adolescents with human papillomavirus vaccine: National Immunization Survey—Teen 2008–2012. BMC Pediatrics, 17(1), 52. https://doi.org/10.1186/s12887-­017-­0804-­1 Chin, J., Chin, C.-L., Panday, S., Ghazanfari, A., Jagadeesan, G., Wang, Z., Ontengco, A., Chang, A., Liu, B., Schwartz, A., & Caskey, R. (2020, September). Tracking the human papillomavirus vaccine risk misinformation: An explorative study to examine how the misinformation has spread in user-­ generated content. In Proceedings of the International Symposium on Human Factors and Ergonomics in Health Care (Vol. 9, No. 1, pp. 312–316). SAGE Publications. https://doi.org/10.1177/2327857920091069 Chou, W.-Y., Gaysynsky, A., & Cappella, J. N. (2020). Where we go from here: Health misinformation on social media. American Journal of Public Health, 110(S3), S273–S275. https://doi.org/10.2105/AJPH.2020.305905 de Martel, C., Plummer, M., Vignat, J., & Franceschi, S. (2017). Worldwide burden of cancer attributable to HPV by site, country and HPV type. International Journal of Cancer, 141(4), 664–670. https://doi.org/10.1002/ijc.30716 Dibble, K.  E., Maksut, J.  L., Siembida, E.  J., Hutchison, M., & Bellizzi, K. M. (2019). A systematic literature review of HPV vaccination barriers among adolescent and young adult males. Journal of Adolescent and Young Adult Oncology, 8(5), 495–511. https://doi.org/10.1089/jayao.2019.0004 Dunn, A. G., Leask, J., Zhou, X., Mandl, K. D., & Coiera, E. (2015). Associations between exposure to and expression of negative opinions about human ­papillomavirus vaccines on social media: An observational study. Journal of Medical Internet Research, 17(6). https://doi.org/10.2196/jmir.4343 Dunn, A. G., Surian, D., Leask, J., Dey, A., Mandl, K. D., & Coiera, E. (2017). Mapping information exposure on social media to explain differences in HPV

52 

Y. WANG ET AL.

vaccine coverage in the United States. Vaccine, 35(23), 3033–3040. https:// doi.org/10.1016/j.vaccine.2017.04.060 Ekram, S., Debiec, K. E., Pumper, M. A., & Moreno, M. A. (2019). Content and commentary: HPV vaccine and YouTube. Journal of Pediatric and Adolescent Gynecology, 32(2), 153–157. https://doi.org/10.1016/j.jpag.2018.11.001 Featherstone, J.  D., Bell, R.  A., & Ruiz, J.  B. (2019). Relationship of people’s sources of health information and political ideology with acceptance of conspiratorial beliefs about vaccines. Vaccine, 37(23), 2993–2997. https://doi. org/10.1016/j.vaccine.2019.04.063 Ferrand, J., Hockensmith, R., Houghton, R.  F., & Walsh-Buhi, E.  R. (2020). Evaluating smart assistant responses for accuracy and misinformation regarding human papillomavirus vaccination: Content analysis study. Journal of Medical Internet Research, 22(8), e19018. https://doi.org/10.2196/19018 Fu, L. Y., Zook, K., Spoehr-Labutta, Z., Hu, P., & Joseph, J. G. (2016). Search engine ranking, quality, and content of web pages that are critical versus noncritical of human papillomavirus vaccine. Journal of Adolescent Health, 58(1), 33–39. https://doi.org/10.1016/j.jadohealth.2015.09.016 Ginossar, T., Cruickshank, I. J., Zheleva, E., Sulskis, J., & Berger-Wolf, T. (2022). Cross-platform spread: Vaccine-related content, sources, and conspiracy theories in YouTube videos shared in early Twitter COVID-19 conversations. Human Vaccines & Immunotherapeutics, 1–13. https://doi.org/10.1080/ 21645515.2021.2003647 Hansen, P. R., & Schmidtblaicher, M. (2021). A dynamic model of vaccine compliance: How fake news undermined the Danish HPV vaccine program. Journal of Business & Economic Statistics, 39(1), 259–271. https://doi.org/10.1080/ 07350015.2019.1623045 Himelboim, I., Lee, J. J., Cacciatore, M. A., Kim, S., Krause, D., Miller-Bains, K., Mattson, K., & Reynolds, J. (In press). Vaccine support and hesitancy on Twitter: Opposing views, similar strategies, and the mixed impact of conspiracy theories. In T. Ginossar, S. F. A. Shah, & D. Weiss (Eds.), Vaccine communication online: Counteracting misinformation, rumors and lies. Palgrave Macmillan. Hofstetter, A. M., & Rosenthal, S. L. (2014). Health care professional communication about STI vaccines with adolescents and parents. Vaccine, 32(14), 1616–1623. https://doi.org/10.1016/j.vaccine.2013.06.035 Hoy, M. B. (2018). Alexa, Siri, Cortana, and more: An introduction to voice assistants. Medical Reference Services Quarterly, 37(1), 81–88. https://doi.org/10. 1080/02763869.2018.1404391 Kim, S. C., Vraga, E. K., & Cook, J. (2020). An eye tracking approach to understanding misinformation and correction strategies on social media: The mediating role of attention and credibility to reduce HPV vaccine misperceptions. Health Communication, 1–10. https://doi.org/10.1080/10410236.2020. 1787933

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

53

Luo, X., Zimet, G., & Shah, S. (2019). A natural language processing framework to analyse the opinions on HPV vaccination reflected in Twitter over 10 years (2008–2017). Human Vaccines & Immunotherapeutics, 15(7–8), 1496–1504. https://doi.org/10.1080/21645515.2019.1627821 Madden, K., Nan, X., Briones, R., & Waks, L. (2012). Sorting through search results: A content analysis of HPV vaccine information online. Vaccine, 30(25), 3741–3746. https://doi.org/10.1016/j.vaccine.2011.10.025 Mahoney, L. M., Tang, T., Ji, K., & Ulrich-Schad, J. (2015). The digital distribution of public health news surrounding the human papillomavirus vaccination: A longitudinal infodemiology study. JMIR Public Health and Surveillance, 1(1), e2. https://doi.org/10.2196/publichealth.3310 Massey, P. M., Kearney, M. D., Hauer, M. K., Selvan, P., Koku, E., & Leader, A.  E. (2020). Dimensions of misinformation about the HPV vaccine on Instagram: Content and network analysis of social media characteristics. Journal of Medical Internet Research, 22(12), e21451. https://doi.org/ 10.2196/21451 Milmo, D. (2021, September 29). YouTube to remove misinformation videos about all vaccines. The Guardian. https://www.theguardian.com/technology/2021/sep/29/youtube-­t o-­r emove-­m isinformation-videos-aboutall-­vaccines Nan, X., & Madden, K. (2012). HPV vaccine information in the blogosphere: How positive and negative blogs influence vaccine-related risk perceptions, attitudes, and behavioral intentions. Health Communication, 27(8), 829–836. https://doi.org/10.1080/10410236.2012.661348 Nan, X., Wang, Y., & Thier, K. (2021). Health misinformation. In T. Thompson & N. Harrington (Eds.), The Routledge handbook of health communication (3rd ed.). Routledge. Ortiz, R. R., Smith, A., & Coyne-Beasley, T. (2019). A systematic literature review to examine the potential for social media to impact HPV vaccine uptake and awareness, knowledge, and attitudes about HPV and HPV vaccination. Human Vaccines & Immunotherapeutics, 15(7–8), 1465–1475. https://doi.org/10. 1080/21645515.2019.1581543 Pew Research Center. (2017, December 12). Nearly half of Americans use digital voice assistants, mostly on their smartphones. https://www.pewresearch.org/ fact-­t ank/2017/12/12/nearly-­h alf-­o f-­a mericans-­u se-­d igital-­v oiceassistants-­mostly-­on-­their-­smartphones/ Pingali, C. (2021). National, regional, state, and selected local area vaccination coverage among adolescents aged 13–17 years—United States, 2020. Morbidity and Mortality Weekly Report, 70. https://doi.org/10.15585/mmwr. mm7035a1 Porshnev, A., Miltsov, A., Lokot, T., & Koltsova, O. (2021, July). Effects of conspiracy thinking style, framing and political interest on accuracy of fake news

54 

Y. WANG ET AL.

recognition by social media users: Evidence from Russia, Kazakhstan and Ukraine. In G. Meiselwitz (Ed.), Social computing and social media: Experience design and social network analysis (pp.  341–357). Springer. https://doi. org/10.1007/978-­3-­030-­77626-­8_23 Raghupathi, V., Ren, J., & Raghupathi, W. (2020). Studying public perception about vaccination: A sentiment analysis of tweets. International Journal of Environmental Research and Public Health, 17(10). https://doi.org/10.3390/ ijerph17103464 Scherer, L. D., McPhetres, J., Pennycook, G., Kempe, A., Allen, L. A., Knoepke, C.  E., Tate, C.  E., & Matlock, D.  D. (2021). Who is susceptible to online health misinformation? A test of four psychosocial hypotheses. Health Psychology, 40(4), 274–284. https://doi.org/10.1037/hea0000978 Shearer, E. (2021). 86% of Americans get news online from smartphone, computer or tablet. Pew Research Center. https://www.pewresearch.org/fact-­ tank/2021/01/12/more-­t han-­e ight-­i n-­t en-­a mericans-­g et-news-­f rom-­ digital-­devices/ Shen, L., & Zhou, Y. (2020). Epistemic egocentrism and processing of vaccine misinformation (vis-à-vis scientific evidence): The case of vaccine-autism link. Health Communication, 1–12. https://doi.org/10.1080/10410236. 2020.1761074 Sonawane, K., Lin, Y.-Y., Damgacioglu, H., Zhu, Y., Fernandez, M.  E., Montealegre, J. R., Cazaban, C. G., Li, R., Lairson, D. R., Lin, Y., Giuliano, A. R., & Deshmukh, A. A. (2021). Trends in human papillomavirus vaccine safety concerns and adverse event reporting in the United States. JAMA Network Open, 4(9), e2124502. https://doi.org/10.1001/ jamanetworkopen.2021.24502 Stefanone, M. A., Vollmer, M., & Covert, J. M. (2019, July). In news we trust? Examining credibility and sharing behaviors of fake news. In Proceedings of the 10th International Conference on Social Media and Society (pp.  136–147). https://doi.org/10.1145/3328529.3328554 Strozewski, Z. (2021, August 10). Instagram removes hundreds of accounts connected to COVID vaccine misinformation campaign. Newsweek. https://www. newsweek.com/instagram-­r emoves-­h undreds-­a ccounts-connectedcovid-­vaccine-­misinformation-­campaign-­1618125 Vraga, E. K., Kim, S. C., & Cook, J. (2019). Testing logic-based and humor-based corrections for science, health, and political misinformation on social media. Journal of Broadcasting & Electronic Media, 63(3), 393–414. https://doi.org/ 10.1080/08838151.2019.1653102 Walter, N., Brooks, J. J., Saucier, C. J., & Suresh, S. (2021). Evaluating the impact of attempts to correct health misinformation on social media: A meta-analysis. Health Communication, 36(13), 1776–1784. https://doi.org/10.1080/ 10410236.2020.1794553

3  HPV VACCINE MISINFORMATION ONLINE: A NARRATIVE SCOPING REVIEW 

55

Wang, Y., McKee, M., Torbica, A., & Stuckler, D. (2019). Systematic literature review on the spread of health-related misinformation on social media. Social Science & Medicine, 240, 112552. https://doi.org/10.1016/j. socscimed.2019.112552 Wong, N. C. H. (2016). “Vaccinations are safe and effective”: Inoculating positive HPV vaccine attitudes against antivaccination attack messages. Communication Reports, 29(3), 127–138. https://doi.org/10.1080/08934215.2015.1083599 Xiao, X., & Su, Y. (2021). Integrating reasoned action approach and message sidedness in the era of misinformation: The case of HPV vaccination promotion. Journal of Health Communication, 26(6), 371–380. https://doi.org/ 10.1080/10810730.2021.1950873 Xu, Z., & Guo, H. (2018). Using text mining to compare online pro- and anti-­ vaccine headlines: Word usage, sentiments, and online popularity. Communication Studies, 69(1), 103–122. https://doi.org/10.1080/1051097 4.2017.1414068 Zhang, J., Featherstone, J. D., Calabrese, C., & Wojcieszak, M. (2021). Effects of fact-checking social media vaccine misinformation on attitudes toward vaccines. Preventive Medicine, 145, 106408. https://doi.org/10.1016/j. ypmed.2020.106408 Zhang, J., Xue, H., Calabrese, C., Chen, H., & Dang, J.  H. T. (2021). Understanding human papillomavirus vaccine promotions and hesitancy in northern California through examining public Facebook pages and groups. Frontiers in Digital Health, 3(62). https://doi.org/10.3389/fdgth. 2021.683090 Zimet, G. D., Rosberger, Z., Fisher, W. A., Perez, S., & Stupiansky, N. W. (2013). Beliefs, behaviors and HPV vaccine: Correcting the myths and the misinformation. Preventive Medicine, 57(5), 414–418. https://doi.org/10.1016/j. ypmed.2013.05.013

CHAPTER 4

Analyzing Social-Cyber Maneuvers for Spreading COVID-19 Pro- and AntiVaccine Information Janice T. Blane, Lynnette Hui Xian Ng, and Kathleen M. Carley

Introduction From the inception of the coronavirus pandemic in early 2020, scientists scrambled to develop vaccines to ease the pandemic. Even prior to the vaccines’ development, online anti-vaccination efforts were significant. Following the various vaccines’ rollouts, pro-vaccination opinions camps formed to encourage mass uptake of vaccination (Cruickshank et  al., 2021), and online social media opinion seemed to form two main polarizing camps: pro-vaccine and anti-vaccine. Anti-vaccine sentiment is not unique to the COVID-19 vaccine. As described in the second chapter in this book by Hoffman and her colleagues, the controversy dates back to the turn of the nineteenth century,

J. T. Blane (*) • L. H. X. Ng • K. M. Carley Software and Societal Systems, School of Computer Science, Carnegie Mellon University, Pittsburgh, PA, USA e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 T. Ginossar et al. (eds.), Vaccine Communication Online, https://doi.org/10.1007/978-3-031-24490-2_4

57

58 

J. T. BLANE ET AL.

when Edward Jenner invented the first smallpox vaccine and governments instituted vaccination mandates (Poland & Jacobson, 2001). The antivaccine movement maintains skepticism about modern medicine generally and publicly offers religious and ethical arguments against vaccines (Hussain et al., 2018). In December 2020, the United States approved the Pfizer and Moderna vaccines for Emergency Use Authorization (EUA). However, the COVID-19 vaccines had already become central discussion points on social media. Pro-vaccine and anti-vaccine communities led discussions to support their respective vaccine stances. Social media networks such as Twitter and Facebook are the perfect platforms for governments and health organizations to disseminate vaccine messages. They can send relevant information to large numbers of people almost instantaneously. Unfortunately, the same qualities that facilitate pro-vaccine conversation helped move the anti-vaccine movement from the fringe to the mainstream, thus exposing ever-increasing social media users to anti-vaccine rhetoric regardless of the integrity of the information. Conspiracy theories about myths and hoaxes, such as that of the coronavirus being a bioweapon, have appeared and gained traction in the social media space (Ng & Carley, 2021a). By the end of 2020, false information had become so prevalent on Twitter that the platform developed policies to remove false or misleading COVID-19-related information and accounts that violated the platform’s misinformation policies (Twitter, 2021). Both pro-vaccine and anti-vaccine communities compete to convince users to take their desired action; that is, either to vaccinate or not to vaccinate, respectively (Ng & Carley, 2021b). In the field of social cybersecurity, online methods for manipulating beliefs and ideas or persuading others are known as social-cyber maneuvers. They are the foundational elements of influence campaigns, including the BEND framework (2020) described in this chapter. For the present chapter, we conducted a year-long study to examine the pro- and anti-vaccination campaign movements during the coronavirus pandemic in the context of online social cybersecurity maneuvers. Additionally, we applied computational analysis methods to conceptualize the use of online manipulation between bots and non-bots in vaccine-­ related communities and then examined the difference between the bots and non-bot groups of users by performing temporal analysis on their participation over time.

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

59

Related Work Vaccine Discussion on Social Media Polarized communities engaging in vaccination discourse can easily form on social networks. On Facebook, this ease is amplified by the explicit creation of groups catering to specific communities (Orr et al., 2016). On Twitter, these groups are more nebulous because they are inferred through the communications between Twitter users or actors in the social media space (Johnson et al., 2020). In the last decade, studies have found trends of increasing anti-vaccine users on social media, specifically Twitter (Gunaratne et al., 2019), and of increasing geographical and linguistic diversity of such groups (Becker et  al., 2016). Common arguments of the anti-vaccination communities include those which question the safety of vaccines and their ingredients (Hussain et  al., 2018; Orr et  al., 2016), as well as ethical and religious concerns, including civil rights and liberty. However, pro-vaccination messages are more sparse in the Twittersphere, indicating a more vocal anti-­ vaccination movement (Deiner et  al., 2019). Nevertheless, there is minimal inter-communication between the pro- and anti-vaccine communities, resulting in ideological isolation (Orr et al., 2016). Social Cybersecurity As online manipulation has gained increased attention, several new analytical frameworks have been developed to characterize online manipulation techniques. The BEND framework examines how actors use narrative and structural maneuvers such as creating excitement or bridging groups in targeting and manipulating actors and groups (Carley, 2020), which will be discussed in the next section; the ABC(D) framework describes the actors-behavior-content-distribution of the manipulation (Alaphilippe, 2020); and the SCOTCH (source, channel, objective, target, composition, hook) framework presents a campaign overview to summarize the actions (Blazek, 2021). In examining the online discourse surrounding the coronavirus pandemic, one must also consider the role of bots. Bots are inauthentic accounts on digital platforms that have been shown to partake in efforts to sway public opinion. Active and aggressive use of bots was documented in

60 

J. T. BLANE ET AL.

large-scale events such as elections (Uyheng et al., 2021) and social cybersecurity events such as during the Kashmir protests and lockdown (Ng & Carley, 2021c). Bend Framework The BEND framework characterizes the fundamental actions users implement to influence campaigns online (Beskow & Carley, 2019). In the BEND framework there are 16 maneuver,s as shown in Table 4.1: the 4 B’s for augmenting the social network—back, build, bridge, and boost; the 4 E’s for augmenting the narrative—excite, enhance, explain, and engage; the 4 N’s for decrementing the social network—neutralize, nuke, narrow, and neglect; and the 4 D’s for decrementing narrative—dismay, distract, distort, and dismiss. While the BEND framework consists of more than these 16 maneuvers, the framework gets its name from the maneuvers. Prior work applied the BEND framework to conceptualize the tactics used in vaccine-related influence campaigns. In particular, it described the narrative manipulation techniques used on both sides of the vaccine discussion (Blane et al., 2022). Maneuvers are deliberate actions intended to achieve a desired end state and are separated into narrative and network, or community, types. Narrative maneuvers are social-cyber maneuvers in which the content of the message or the topic of discussion is the means for influencing a target audience. Blane et al. (2022) found that pro-vaccine communities primarily used explain messages to convey their narrative to their target audiences, while anti-vaccine communities used dismay maneuvers. Explain maneuvers are messages that elaborate on a topic either through scientific and logical arguments or with relevant facts and supported data. Dismay Table 4.1  BEND Maneuvers

Positive

Negative

Narrative maneuvers

Social network maneuvers

Engage Explain Excite Enhance Dismiss Distort Dismay Distract

Back Build Bridge Boost Neutralize Nuke Narrow Neglect

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

61

maneuvers draw on negative emotions such as anger and sadness to convey a message. Many of the pro-vaccine messages used scientific data and logical explanations for the safety and effectiveness of the vaccine and the reasons why people should get vaccinated. In contrast, anti-vaccine messages appealed to users’ emotions, using sad, angry, or frightening topics as a way to convince others not to get the vaccine. Network maneuvers are actions users take to alter the structure of the social network. Some of these methods can be implemented by manipulating the content of the message and referencing or mentioning others in tweets. The results of these types of social-cyber maneuvers affect who is talking to whom, the relative prominence of individual actors in this network, and the growing or shrinking of groups. Back and neutralize maneuvers manipulate the interactions with opinion leaders. Back attempts to increase the importance of these leaders or create new ones, whereas neutralize decreases their importance. In a study analyzing online community response to the initial (December 2020) Pfizer-BioNTech vaccine rollout, the pro-vaccine community had many high-profile individuals such as world leaders and health organizations advocating receiving the vaccine (Blane et al., 2022). In turn, pro-­ vaccine users engaged in supportive messaging where they backed their community leaders. This maneuver entails positively supporting specific users within the dataset. In contrast, anti-vaccine advocates engaged in many neutralize maneuvers where they attempted to discredit those pro-­ vaccine community leaders. The present study performs a longitudinal analysis of the social cyber-­ security maneuvers in the pro- and anti-vaccination communities, expanding the study of maneuvers to a year. It focuses on the comparison of four of the BEND maneuvers (Explain/Dismay, Back/Neutralize), identifying key influencers and narratives. These concepts thus provide ideas on the social media discourse surrounding the vaccination rollout. In application, the BEND framework characterizes the actions of the sender of the message. Thus, we examine key actors, those senders who have a more significant relative impact, and ask which maneuvers (explain, dismay, back, and neutralize) they are using and toward whom they are directing these maneuvers. In social media, key actors need not be human. Further, in order to learn about the role of different key actors, we analyzed whether they are a bot, news agency, or government official. Two metrics are used to identify key actors—superspreaders and super friends. Superspreaders have

62 

J. T. BLANE ET AL.

information content that is spread more often than content provided by the typical actor, and super friends partake in larger amounts of two-way communication than does the typical actor.

Methods The methods used to analyze the social cybersecurity maneuvers regarding the pro- and anti-vaccine stances are depicted in Fig.  4.1. We used the same interoperable pipeline as in previous studies (Blane et al., 2022). In the data preparation phase, tweets were collected and filtered to create the datasets. This study filtered the data by date and by whether the posts discussed COVID-19 vaccines. Then we used a bot-detection algorithm to create bot attributes for the actors (Beskow & Carley, 2018), utilized a computational linguistic program to extract cues from each tweet, and then applied a stance detection algorithm (Kumar, 2020) to separate the datasets into the pro-vaccine and anti-vaccine communities. We identified and explored the behaviors and characteristics of key actors and evaluated the actors, messages, and the entire network within the scope of the four social-cyber BEND maneuvers. These four maneuvers helped to frame the overall influence campaigns within the dataset. Data Collection We used the Twitter V1 streaming Application Programming Interface (API) to collect tweets related to the coronavirus from June 15, 2020,

Social-Cyber Maneuver Analysis Pipeline Data Preparation Data Filtering Bot Detection Identify Linguistic cues Stance Detection

Key Actors Identification

Social-Cyber Maneuvers Analysis

Key Influencers

BEND Maneuvers Narrative Maneuvers Network Maneuvers

Key Targets Superfriends Superspreaders Bots Groups

Influence Campaigns

Fig. 4.1  Social-Cyber Maneuver Analysis Pipeline (Blane et al., 2022)

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

63

Table 4.2  COVID-19 Vaccine Twitter Data Collection Keywords Filter

Keywords

COVID-­19 tweets Vaccine tweets

Coronaravirus, coronavirus, Wuhan virus, wuhanvirus, 2019nCoV, NCoV, NCoV2019, covid-19, covid19, covid 19 Vaccine, vax, mRNA, vaccination, getvaccinated, shot, jab, dose1, dose2, VAERS, believemothers, mybodymychoice, thisisourshot, immunization, gotmyshot, igottheshot, covidvaccinated, moderna, pfizer, Johnson&Johnson, j&j, Johnson and Johnson, jandj, biontech, Johnson & Johnson

through June 20, 2021. An API is an interface to connect to the Twitter platform that allows researchers to collect a stream of data, filtered for tweets that contain specific keywords. These tweets were collected and filtered twice using the keywords in Table 4.2. First, they were first filtered to identify COVID-19 specific tweets, and then they were filtered again to obtain tweets relevant to the coronavirus vaccine. Finally, we obtained 125 GB worth of tweet,s comprising a total of 17,280,461 tweets from 5,129,926 unique users. To facilitate temporal analysis, we subdivided the dataset into one-week windows separated every four weeks. Each component of this sub-dataset is referred to as a “time period.” Hence, for each month, there is one time period. In each of the datasets, we performed additional data processing, including bot annotation, stance annotation, linguistic cues annotation, and the annotation of BEND maneuvers. These annotations further characterized the users and tweets in each subset which enabled downstream discovery of trends. Bot Annotation We annotated the data by performing bot-probability annotation using the BotHunter algorithm (Beskow & Carley, 2018). This algorithm classifies agents using a random forest method through a multi-tiered approach. The algorithm used user-level features (e.g., screen name length, account age), network-level features (e.g., number of followers), and tweet-level features (e.g., content). Given the probability of the user being a bot, we then classified users as bots if their bot-probability score was above 0.75 and as a non-bot if their bot-probability score was below 0.75, a stable bot threshold value below which users were least likely to

64 

J. T. BLANE ET AL.

flip their classification from bots and non-bots (Ng et  al., 2022). Furthermore, Twitter verified users were classified as non-bots.  inguistic Cues Annotation L We used the NetMapper software (v.1.0.0.65) to extract linguistic cues from the tweet texts (Carley et al., 2018). Many of the cues in NetMapper are similar to those in Linguistic Inquiry and Word Count (LIWC), but in addition, NetMapper calculates localized sentiment and employs findings from affect control theory (Heise, 2007). These cues include the number of pronouns, the frequency of positive and negative terms, the number of abusive terms, the sentiment of the tweet, the sentiment for terms of interest, and many other linguistic indicators. These metrics are useful for capturing the psycholinguistic features of the tweet text and are used as input for determining the BEND maneuvers in social networks. Stance Annotation To divide the datasets for each period into pro-vaccine stance and anti-­ vaccine stance, we first annotated a subset of each dataset’s hashtags as pro- or anti-vaccine. We used a network-based label propagation algorithm that propagates the stance from seed labels throughout the network (Kumar, 2020). We used the instantiation of this algorithm that is available in ORA-PRO software (Altman et  al., 2020). The algorithm uses information from the user-hashtag and user-tweet network to propagate the labels of hashtags from the seed hashtags. It then identifies user stances from the presence of the hashtags corresponding to pro- and anti-vaccine stances in the user-tweet network. The algorithm provides a label and a confidence value for each hashtag and user. We applied this stance propagation algorithm on each of the one-week networks.  nline Maneuvers Annotation Using the BEND Framework O To characterize online manipulation, we used the BEND framework (Carley, 2020). We performed annotation of the BEND maneuvers in each of the one-week networks using the built-in BEND maneuver annotation in the ORA-PRO software v.3.0.9.133 (Altman et al., 2020). The BEND maneuver annotation requires the annotated linguistic cues from the previous step, the user-tweet network, and the user-user network graphs as input. We analyzed the BEND maneuvers in terms of two stanced communities of users and tweets. For each community, we retrieved the key concepts tweeted by the users in the community. We also

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

65

computed the most influential users, a factor measured by the network centrality values for superspreaders as operationalized in ORA-PRO, a social-network analysis software (Altman et al., 2020). (Superspreaders are actors within the network that have an extensive reach either because they have many followers or because their messages and content are shared often.) ORA-PRO labels each actor and tweet with the probability of the user and document conducting and containing, respectively, each of the 16 BEND maneuvers. When analyzing narratives, ORA-PRO creates the ability to observe different aspects of the dataset more closely by counting concepts such as hashtags, URLs, and words. We used this feature to explore the message content of the narrative maneuvers.

Results and Discussion In this section, we describe the results obtained from the analysis of social cybersecurity maneuvers using the BEND framework. Key Influencers In this study, we focused on superspreaders as key influencers within each of the communities. For Twitter data, this includes sharing through replies, retweets, and quotes. Replies are comments that occur within the thread of an original tweet, retweets are reposts of another user’s post without comments, and quotes are retweets with comments. As superspreaders are efficient in propagating their information through these means, they can better reach their target audiences and influence them. For each community over time, we identified the top 100 superspreaders and counted the number of bots, verified accounts, and news agencies. Verified accounts are Twitter-labeled accounts that let other users know the authenticity of a public account. These include public leaders and organizations as well as celebrities and major companies. Figure 4.2 compares the actor types within the top superspreaders for those actors with either pro-stance or anti-stance messaging. For verified users and news agencies, there were generally more pro-­ vaccine than anti-vaccine superspreaders from June 2020 to March 2021. Bots, on the other hand, were greater for anti-vaccine than pro-vaccine superspreaders. However, in April 2021, this trend changed, with more anti-vaccine superspreaders. This trend of a greater number of anti-vaccine superspreaders continued until the end of the data set period.

66 

J. T. BLANE ET AL.

Fig. 4.2  Actor Types in Top 100 Superspreaders

During these last three to four time periods, the anti-vaccine scene changed. Anti-vaccine messages evolved from expressing specific concerns about vaccines to being part of a larger, more general vaccine influence campaign discussing topics beyond whether to vaccinate. In the United States, this larger campaign emerged as more right-wing rhetoric systematically aligned with anti-vaccine messages. In other countries, anti-­vaccine messages became tied to anti-vaccine discussions about vaccine passports (certifications of immunization to allow travel). Thus, the agent types associated with anti-vaccination were no longer fringe users sending obscure anti-vaccine-related propaganda. Rather, they were partisan or issue-oriented leaders, news agencies, and, subsequently, bots that became attached to the anti-vaccine identity. Finally, the number of superspreaders appeared to decrease while the number of verified users increased over time, as shown in Fig. 4.2. Narratives Through Hashtags Hashtags can be used to highlight or classify messages to promote specific topics or people. In our study, we used hashtags to identify narrative themes within the dataset. To compare the narratives at the beginning versus the end of the dataset, we collected and counted the hashtags for the first and last one-week time periods (June 2020 and June 2021) to create networks that connected hashtags used in the same message. We

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

67

then removed neutral COVID-19 and vaccine hashtags, such as #COVID-19, #coronavirus, and #vaccines, as well as hashtags classified as originating from the opposing community. (As an example, pro-vaccine and neutral hashtags were retained for a pro-vaccine hashtag network.) Finally, we combined the top 100 used hashtags for each stance during the first and last time periods to create hashtag co-occurrence networks. Figures 4.3, 4.4, 4.5, and 4.6 show the networks for the pro-vaccine and anti-vaccine hashtag networks from June 2020 to June 2021. The node sizes of the hashtags indicate the relative usage count and the size of the links relate to the number of times that the hashtags were used together. Link values lower than ten are hidden. The June 2020 time period (tweets during the week of June 15, 2020) occurred approximately six months before the initial rollout of the Pfizer and Moderna vaccines, which began to be administered in December 2020. As shown in Fig. 4.3, pro-vaccine messages were mainly associated with general protective measures against the COVID-19 virus. Emphasis was on staying at home, wearing masks, social distancing, and uniting against the coronavirus. Pro-vaccine messages also targeted health organizations and senior government leaders. Vaccine manufacturers, however, were not yet a primary focus of pro-vaccine campaigns. Additionally, many hashtags linked to conversations outside English-speaking regions. Malay

Fig. 4.3  Pro-Vaccine Hashtag Co-occurrence Network (June 2020)

68 

J. T. BLANE ET AL.

Fig. 4.4  Anti-Vaccine Hashtag Co-occurrence Network (June 2020)

Fig. 4.5  Pro-Vaccine Hashtag Co-occurrence Network (June 2021)

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

69

Fig. 4.6  Anti-Vaccine Hashtag Co-occurrence Network (June 2021)

and Spanish hashtags integrated with English hashtags, and hashtags associated with other countries such as India, China, and Pakistan, were also popular. Furthermore, pro-vaccine agents also discussed conspiracy theories and myths, particularly regarding Bill Gates’s plan to use microchips to fight the coronavirus, Plandemic, and QAnon. Some questioned COVID-19 and government intentions in general; for example, “What if we have #lockdown wrong? what if @BorisJohnson @MattHancock had another reason … what if #COVID19 is not real? what if there is no plan #vaccine maybe all kept in our homes for a reason.. #TheGreatAwakening #Qanon.” Others, however, pontificated about the vaccine plan: What if, & I’m even doubting myself, but what if, they knew we’d see through #BillGates demonic #vaccine plan, but they’re pushing it, for now, to distract us, like they distracted us with #Q #Qanon? I mean, are they going to show their hand? #coronavirus #COVID19 #markofthebeast”. Mainstream media sources reported on the issue, “Twitter Alleges Bill Gates Wants to 'Control Population' by 'Poison' of COVID-19 Vaccines, Trends #ExposeBillGates via @indiacom.

70 

J. T. BLANE ET AL.

We compared the pro-vaccine with the anti-vaccine hashtag co-­ occurrence network during this same period, as shown in Fig. 4.4. Many of the hashtags in this dataset referred to anti-vaccine conspiracy theories, making connections to Bill Gates, QAnon, 5G (5th Generation mobile network), and Plandemic. The #billgates hashtag was the most frequently used in this dataset and co-occurred with these conspiracy theories as well as right-wing conservative hashtags such as #trump, #maga (Make America Great Again, a slogan for the Trump 2020 campaign), #kag (Keep America Great, a hashtag used frequently with Trump supporters), and #WWG1WGA (Where We Go one, Where We Go All, a hashtag typically used with QAnon supporters). Other distinct themes within this dataset were hashtags that countered general public health messages. These included #nomask, #hydroxychloroquine, and #hcq. Additionally, among these top hashtags, a set of highly-connected hashtags emerged, tying #Qorona, #billgates, #WHO, and #mRNA to German political conversations about the vaccine. Six months before a vaccine was approved, COVID-19 anti-vaccine messages were already being propagated worldwide, mostly as parts of conspiracy theories. One year after the first time period (June 2020) and six months after the initial rollout of the Pfizer and Moderna vaccines, vaccine conversation evolved with the changing times. As shown in Fig. 4.5, pro-vaccine conversations worldwide centered around getting vaccinated, vaccine equality, vaccination drives, and the new Delta variant of the virus. Hashtags included #vaccineswork and #VaccinesSaveLives that aimed to reassure users, alongside #GetVaccinated and #Impfung (German for vaccination) messages. In contrast to what we observed during the earlier period, more hashtags supported or targeted multiple vaccine manufacturers as countries approved and began administering the many versions of the vaccines, such as Sinovac, Moderna, and Pfizer. Anti-vaccine conversations were similar to pro-vaccine conversations in June 2021 as they contained a more global message and tied discussions to specific vaccine manufacturers. However, their narrative had a different focus than was seen during the previous year, as shown in Fig. 4.6. Many of the differences can be attributed to the Twitter policy changes that removed false or misleading COVID-19 tweets and accounts. This meant that many of the conspiracy theories and much of the fake news surrounding the vaccines, as well as the associated user accounts, were either removed or suspended from Twitter. As a result, anti-vaccine messages on Twitter focused predominantly on negative vaccine-related information. A

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

71

prominent theme was the campaign against vaccine passports. There were also references to ivermectin, a controversial drug not approved for COVID-19 treatment yet which gained popularity within the anti-­vaccine community. Comparing Narrative Maneuvers: Explain and Dismay Using ORA-PRO, we calculated the percentage of explain and dismay messages by pro-vaccine and anti-vaccine users. In Figs. 4.7 and 4.8, we compare these communities over time. As the pro-vaccine explanations were based on scientific evidence, their number began to decrease over time, with fewer actors repeating the same or similar explanations. At the same time, dismay messages among the pro-vaccine community tended to focus on the growing number of cases and deaths, the increased threat from the Delta variant, and delays in or lack of access to vaccines. The two maneuvers combined, creating a desire for vaccinations. In contrast, the anti-vaccine explanations centered on explaining by giving anecdotes and examples—for which there were always new ones. Anecdote explanation and dismay maneuvers were highly synergistic; by coupling them together, it was possible to create a sense of fear and a distrust of science in the anti-vaccination community.

Fig. 4.7  Agents Conducting Explain Maneuvers Over Time by Stance

72 

J. T. BLANE ET AL.

Fig. 4.8  Agents Conducting Dismay Maneuvers Over Time by Stance

The key here is that the pro-vaccine community was getting bombarded with diverse messaging, although supporting vaccination does not create a unified communal emotion or storyline. By contrast, the anti-vaccine maneuvers were synergistic and served to create a unified communal emotion (fear) and a common storyline (don’t trust science, vaccines don’t work, vaccines are not safe). Comparing Network Maneuvers: Back and Neutralize Similar to the narrative maneuver comparison, we calculated the percentage of back and neutralize messages by bots and non-bots using ORA-­ PRO, as shown in Figs. 4.9 and 4.10. Back and neutralize maneuvers performed similarly for the two communities. These maneuvers go hand-in-hand. While trying to support the community opinion leaders, users were subsequently undermining the opposing leaders. Pro-vaccine leaders were prevalent and easy targets for these maneuvers on both sides, but less apparent were the effects of lesser-­ known opinion leaders within each stance base. For each community, the high-profile users were more common targets early in the timeline, but as time progressed, lower-level and non-verified account leaders emerged as more popular targets. For anti-vaccine communities, the result was more apparent, with increased maneuvering in later periods as the pool of leaders evolved with the changing narratives.

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

Fig. 4.9  Agents Conducting Back Maneuvers Over Time by Stance

Fig. 4.10  Agents Conducting Neutralize Maneuvers Over Time by Stance

73

74 

J. T. BLANE ET AL.

Conclusion In this study, we explored a year’s worth of COVID-19 vaccine Twitter data to understand better how pro-vaccine and anti-vaccine communities used social-cyber maneuvers to persuade others to vaccinate or not to vaccinate. For each stance, we identified key influencers, specifically superspreaders, focusing on how bots, verified actors, and news agencies effectively generated and propagated content. We analyzed narratives and conspiracy theories and compared maneuvers over time. Unlike earlier work on vaccines, our study used a computational approach that applied the BEND framework, asking who was targeting whom using which of 16 different maneuvers. Unlike earlier work, the present study did not focus on just bots, but also considered verified actors and new agencies. Our targets were the pro- and anti-vaccination communities. By computationally assessing the BEND maneuvers employed by these actors, we were able to arrive at more fine-grained findings than would have been possible had we taken computational approaches which use topic modeling, making them easier to relate to qualitative research. However, the number of messages that could be tracked and the time frame were larger than can typically be done in a qualitative study. The resulting blend of big data and in-depth assessment better informs interventions in identifying the users who are spreading anti-vaccine messages, the narratives that are propagated, and the way they are propagated, and therefore can provide early warning indications to authorities to prepare the public health messaging required. Our focus was on the vaccine rollout. Hence, we began with data in June 2020 and right from the start observed pro-vaccine and anti-vaccine narratives. This does not mean that the narratives themselves began only in June. Indeed, other researchers identified COVID-19 vaccine conversations occurring as early as February 2020 (Cruickshank et  al., 2021; Ginossar et al., 2022). Rather, our point is that as of six months before the vaccine rollout, not only was there already an ongoing discussion, but the ways in which the pro-vaccine and anti-vaccine narratives were being framed were already distinct. Our analysis showed that earlier conversations for pro-vaccine users did not focus on any specific vaccine. Rather, they tied vaccine narratives to measures intended to keep society safe, such as social distancing, wearing a mask, and uniting against the coronavirus. Explain maneuvers were used, presenting vaccination as another safety measure. In contrast, many

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

75

anti-vaccine messages were associated with COVID-19 conspiracy theories, unapproved drugs for fighting the virus, and right-wing conservative themes. Dismay maneuvers were used in attempts to undercut the potential value of the not- yet-released vaccines. Following the vaccine rollout, pro-vaccine conversations broadened into a larger international discussion about the vaccine. Anti-vaccine discourse also extended globally. On December 16, 2020, Twitter started removing false claims about COVID-19 vaccines. Conspiracy theories were discussed to a lesser degree, most likely as a result of Twitter policies, but as evidenced in tweets about vaccine passports, negative sentiment regarding vaccines in general remained visible. Our findings suggest that Twitter’s censorship, though well intentioned, did not stop the anti-vaccination debate. For example, during roughly the first half of the June 2020–June 2021 time period we considered, the popular (albeit unapproved) COVID-19 remedy was hydroxychloroquine, while during the latter half, ivermectin emerged as the anti-vaccine drug of choice, reflecting that as Twitter moved to censor these quack cures, others begin popping up. Meanwhile, pro-vaccine Twitter users supported vaccination drives, emphasized the safety and efficacy of the vaccine, and stressed efforts to continue to fight the virus. Those promoting vaccination backed supported their arguments with reports from the CDC and the WHO, while anti-vaccine promoters tried to neutralize those actors. Prior research has argued that the approach to persuasion is different for anti-vaccine messages and pro-vaccine messages; e.g., anti-vaccine messages use anecdotal stories, humor, or sarcasm, whereas pro-vaccine messages use information and participation (Scannell et al., 2021). Our findings complement these results and provide further detail. Both the pro- and anti-vaccine communities engaged in both explain and dismay narrative maneuvers to convey their narratives. However, the anti-­ vaccination tweets provided explanations through anecdotes, while the pro-vaccination used statistics and information, attempting to bolster their arguments using facts and reasoning about the safety and efficacy of the vaccine or to prove that the vaccine would cause more harm than good. Pro-vaccine users dismayed their audiences with personal stories of COVID-19 cases and deaths, and anti-vaccine users did the same using vaccine side effects and mortality rates. These dismay maneuvers served two purposes: 1) to confirm to members in your own group that you are justified in your stance, and 2) to try to convince those in the opposing

76 

J. T. BLANE ET AL.

group to change their stance. Though the pro-vaccine community applied used varying message strategies to support their overall narrative in support of vaccination, the anti-vaccine narratives shared a more pointed storyline and an effective message of fear that united the anti-vaccine community. Prior research has discussed the importance of celebrities in spreading stories. We found that both the pro- and anti-vaccine communities also engaged in back and neutralize community maneuvers to support their own celebrity-like opinion leaders and to undercut opinion leaders holding the opposite stance. Pro-vaccinators had more high-profile leaders than those against the vaccine, though; this made them the targets for pro-vaccine backing and anti-vaccine neutralizing. However, these maneuvers were still prevalent among lower-profile opinion leaders. Users supported those who agreed with their position and undermined those who disagreed with them, regardless of their popularity online and in the “real” world. These community maneuvers, when used together in this way, increased the overall level of polarization. Thus, celebrities supported the spread of the stories but only to those already inclined to agree. Another finding that should be further explored is the decrease in the number of superspreaders and the simultaneous increase in the number of verified users in our data. This can be partially attributed to Twitter policies that went into effect in December 2020 and were updated in March 2020 calling for the removal of false or misleading COVID-19 vaccine tweets and offending accounts (Twitter, 2021). It is possible that as the platform’s anti-misinformation policies were enforced, they diminished the anti-vaccine online community. Another possible explanation is that pro-vaccine community leaders in the form of verified accounts gained greater traction for spreading pro-vaccine messages, both because there was less opposition and because they became more engaged and active in this discourse. One of the key findings from our study is that the anti-vaccination conversation is much more focused and consistent among its supporters, while the pro-vaccination conversation is much less focused and presents a more diverse and scattered set of views. Many studies point out that public health messaging intended to manage risks and prevent infectious disease needs to be clear, simple, transparent, and unified (Ghio et al., 2021). Our findings suggest that presenting unified messages on social media is important not only for public health providers but also for members of the public who support health providers’ goals.

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

77

Several limitations nuance our conclusions from this work. The data collected were limited to those available via the Twitter streaming API, which returns 1% of the tweets posted. During our data collection, the Twitter collector experienced issues and had to be restarted, breaking the continuous flow of tweets. Because this was the case, we suggest caution in extrapolating findings. We also note that while the stance propagation algorithm provides an automated way to diffuse stances through the network, it requires manual inspection for the initial seed hashtags to characterize both pro- and anti-vaccination campaigns. Naturally, though, as time passed, the online conversations evolved and their hashtags evolved along with them. This resulted in varying confidence levels for the stance-­ annotated datasets. Additionally, the BEND maneuvers were calculated based on the network and linguistic structure of the tweets for a one-week time period every month. In future work, we hope to incorporate a user’s historical tweets as temporal information to enhance the BEND maneuver calculations. Nonetheless, we hope that our work provides an understanding of some of the social cybersecurity maneuvers that were used during the COVID vaccine discourse. In applying the BEND framework, we explored only who was doing what. Future research should consider the entire framework and thus also explore the targets of these maneuvers as well as their full impact. For example, we found that the frequency of use of neutralize and that of dismay were fairly similar. This is due in part to the fact that in our dataset some of the most-used indicators were the same for the two metrics. However, the long-term implications of the two maneuvers are quite different. Neutralize will eventually result in lowering the number of connections to the opinion leader and reducing trust in them, whereas dismay will have an immediate impact in affecting offline activity on the part of those who feel that dismay. Influence campaigns rooted in offline events continue to take place on social media. They not only manipulate conversations online but have the potential to manipulate people’s beliefs, ideas, and real-world behaviors. This study illustrates the need for reliable, high-quality tools to study current and emerging social-cyber maneuvers, identify them, and evaluate their impacts. Such research can help develop policies that can ultimately facilitate safe, open, and informed discourse online.

78 

J. T. BLANE ET AL.

Acknowledgements  The research for this chapter was supported in part by the Office of Naval Research (ONR) under grants N00014182106 and N000142112229, the Knight Foundation, the United States Army, and the Center for Informed Democracy and Social-cybersecurity (IDeaS). The views and conclusions are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Knight Foundation, the ONR, the United States Army, or the US Government.

References Alaphilippe, A. (2020). Adding a “D” to the ABC disinformation framework. The Brookings Institute. https://www.brookings.edu/techstream/adding-­a-­d-­to-­ the-­abc-­disinformation-­framework/ Altman, N., Carley, K. M., & Reminga, J. (2020). ORA user’s guide 2020. Carnegie Mellon University, School of Computer Science, Institute for Software Research. http://www.casos.cs.cmu.edu/publications/papers/CMU-­ISR20-­110.pdf Becker, B. F. H., Larson, H. J., Bonhoeffer, J., van Mulligen, E. M., Kors, J. A., & Sturkenboom, M. C. J. M. (2016). Evaluation of a multinational, multilingual vaccine debate on twitter. Vaccine, 34(50), 6166–6171. https://doi. org/10.1016/j.vaccine.2016.11.007 Beskow, D. M., & Carley, K. M. (2018). Bot-hunter: A tiered approach to detecting & characterizing ‘automated activity on twitter. SBP-brims: International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction and Behavior Representation in Modeling and Simulation, 3, 3. Beskow, D.  M., & Carley, K.  M. (2019). Social cybersecurity: An emerging national security requirement. Military Review, 99(2), 117–127. Blane, J. T., Bellutta, D., & Carley, K. M. (2022). Social-cyber maneuvers during the COVID-19 vaccine initial rollout: Content analysis of tweets. Journal of Medical Internet Research, 24(3), e34040. Blazek, S. (2021). SCOTCH: A framework for rapidly assessing influence operations. Atlantic Council. https://www.atlanticcouncil.org/blogs/geotech-­ cues/scotch-­a-­framework-­for-­rapidly-­assessing-­influence-­operations/ Carley, K. M. (2020). Social cybersecurity: An emerging science. Computational and Mathematical Organization Theory, 26, 365–381. https://doi. org/10.1007/s10588-­020-­09322-­9 Carley, L. R., Reminga, J., & Carley, K. M. (2018). Ora & netmapper. Proceedings of the International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction and Behavior Representation in Modeling and Simulation. Springer, 3(3.3), 7. Cruickshank, I., Ginossar, T., Sulskis, J., Zheleva, E., & Berger-Wolf, T. (2021). Content and dynamics of websites shared over vaccine-related tweets in

4  ANALYZING SOCIAL-CYBER MANEUVERS FOR SPREADING COVID-19… 

79

COVID-19 conversations: Computational analysis. Journal of Medical Internet Research, 23(12), e29127. https://doi.org/10.2196/29127 Deiner, M. S., Fathy, C., Kim, J., Niemeyer, K., Ramirez, D., Ackley, S. F., Liu, F., Lietman, T. M., & Porco, T. C. (2019). Facebook and twitter vaccine sentiment in response to measles outbreaks. Health Informatics Journal, 25(3), 1116–1132. https://doi.org/10.1177/1460458217740723 Ghio, D., Lawes-Wickwar, S., Tang, M. Y., Epton, T., Howlett, N., Jenkinson, E., Stanescu, S., Westbrook, J., Kassianos, A.  P., Watson, D., Sutherland, L., & Keyworth, C. (2021). What influences people’s responses to public health messages for managing risks and preventing infectious diseases? A rapid systematic review of the evidence and recommendations. BMJ Open, 11(11), e048750. Ginossar, T., Cruickshank, I., Sulskis, J., Zheleva, E., & Berger-Wolf, T. (2022). Cross-platform spread: Vaccine-related content, sources, and conspiracy theories in YouTube videos shared in early twitter COVID-19 conversations. Human Vaccines & Immunotherapeutics, 18(1), 1–13. Gunaratne, K., Coomes, E.  A., & Haghbayan, H. (2019). Temporal trends in anti-vaccine discourse on twitter. Vaccine, 37(35), 4867–4871. https://doi. org/10.1016/j.vaccine.2019.06.086 Heise, D.  R. (2007). Expressive order: Confirming sentiments in social actions. Springer. Hussain, A., Ali, S., Ahmed, M., & Hussain, S. (2018). The anti-vaccination movement: A regression in modern medicine. Cureus, 10(7), e2919. Johnson, N. F., Velásquez, N., Restrepo, N. J., Leahy, R., Gabriel, N., El Oud, S., Zheng, M., Manrique, P., Wuchty, S., & Lupu, Y. (2020). The online competition between pro-and anti-vaccination views. Nature, 582(7811), 230–233. Kumar, S. (2020). Social media analytics for stance mining a multi-modal approach with weak supervision. PhD dissertation. Carnegie Mellon University. Ng, L.  H. X., & Carley, K.  M. (2021a). “The coronavirus is a bioweapon”: Classifying coronavirus stories on fact-checking sites. Computational and Mathematical Organization Theory, 27(2), 179–194. Ng L. H. X., & Carley, K. M. (2021b). Flipping stance: Social influence on bots’ and non bots’ COVID vaccine stance. Proceedings of the Second International MIS2 Workshop: Misinformation and Misbehavior Mining on the Web at KDD 2021. Ng, L. H. X., & Carley, K. M. (2021c). Bot-based emotion behavior differences in images during Kashmir black day event. In R.  Thomson, M.  N. Hussain, C.  Dancy, & A.  Pyke (Eds.), Social, cultural, and behavioral modeling (pp.  184–194). Springer International Publishing. https://doi. org/10.1007/978-­3-­030-­80387-­2_18 Ng, L. H. X., Robertson, D. C., & Carley, K. M. (2022). Stabilizing a supervised bot detection algorithm: How much data is needed for consistent predictions? Online Social Networks and Media, 28, 100198.

80 

J. T. BLANE ET AL.

Orr, D., Baram-Tsabari, A., & Landsman, K. (2016). Social media as a platform for health-related public debates and discussions: The polio vaccine on Facebook. Israel Journal of Health Policy Research, 5(1), 34. https://doi. org/10.1186/s13584-­016-­0093-­4 Poland, G.  A., & Jacobson, R.  M. (2001). Understanding those who do not understand: A brief review of the anti-vaccine movement. Vaccine, 19(17–19), 2440–2445. https://doi.org/10.1016/S0264-­410X(00)00469-­2 Scannell, D., Desens, L., Guadagno, M., Tra, Y., Acker, E., Sheridan, K., Rosner, M., Mathieu, J., & Fulk, M. (2021). COVID-19 vaccine discourse on twitter: A content analysis of persuasion techniques, sentiment and mis/disinformation. Journal of Health Communication, 26(7), 443–459. Twitter. (2021). Updating our approach to misleading information. https://blog. twitter.com/en_us/topics/product/2020/updating-­o ur-­a pproach-­t omisleading-­information Uyheng, J., Ng, L. H. X., & Carley, K. M. (2021). Active, aggressive, but to little avail: Characterizing bot activity during the 2020 Singaporean elections. Computational and Mathematical Organization Theory, 27(3), 324–342.

CHAPTER 5

Vaccine Support and Hesitancy on Twitter: Opposing Views, Similar Strategies, and the Mixed Impact of Conspiracy Theories Itai Himelboim, Jeonghyun Janice Lee, Michael A. Cacciatore, Sungsu Kim, Diane Krause, Kate Miller-Bains, Kristin Mattson, and Jennifer Reynolds

The advent of the vaccine is considered to be one of the greatest medical achievements in the history of public health (Centers for Disease Control and Prevention, 1999; Ehreth, 2003). Nevertheless, there has been concern about growing skepticism toward vaccines. Some of this can be traced back to Andrew Wakefield’s since-retracted research study, which

I. Himelboim (*) • M. A. Cacciatore Grady College of Journalism and Mass Communication, University of Georgia, Athens, GA, USA e-mail: [email protected] J. J. Lee Manship School of Mass Communication, Louisiana State University, Baton Rouge, LA, USA © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 T. Ginossar et al. (eds.), Vaccine Communication Online, https://doi.org/10.1007/978-3-031-24490-2_5

81

82 

I. HIMELBOIM ET AL.

suggested a link between the measles-mumps-rubella (MMR) vaccine and autism (Eggertson, 2010; Howard, 2017). However, skepticism is not limited to the MMR vaccine. Some parents have expressed concerns that vaccinating children against HPV serves as an invitation for them to become sexually active, while other parents describe the vaccine as unnecessary or unsafe (Beavis et al., 2018). In explaining the rise in vaccine hesitancy and skepticism, some have cited the growing role of the Internet, and specifically, user-generated social media content, in providing platforms for small pockets of vaccine-­ hesitant publics to spread misleading and outright false information about the safety and efficacy of vaccines (Larson, 2018). Indeed, social science scholarship has noted an uptick in public apprehension and concern surrounding various types of vaccines (Kempe et al., 2015; Saada et al., 2015). It is perhaps not surprising, then, that we have seen several high-­ profile outbreaks of vaccine-preventable illnesses in countries where such diseases were previously declared eliminated (Centers for Disease Control and Prevention, 2015). Given the importance of high vaccination rates for public health and safety, there is a need to understand the types of vaccine-related information that everyday citizens are potentially encountering and using when forming their perceptions. In this study, we focus specifically on the social media platform Twitter, which has emerged as a major source for news, including health news, in the last 10 years (Shearer & Gottfried, 2017). However, our work takes a different approach than most previous studies in this space. First, we widen our focus to include not only anti-vaccination content, but also vaccine-supportive content. We do so by focusing on comparable categories of content between vaccine-hesitant and vaccine-­ supportive Twitter messages. Second, rather than explore raw incidences of the most common keywords irrespective of time, we focus our approach on the most popular posts—in terms of retweets—for each week of a single calendar year. We believe these two decisions not only paint a more S. Kim School of Communication, Kookmin University, Seoul, South Korea D. Krause • K. Mattson • J. Reynolds Oak Ridge Associated Universities, Oak Ridge, TN, USA K. Miller-Bains University of Virginia, Charlottesville, VA, USA

5  VACCINE SUPPORT AND HESITANCY ON TWITTER: OPPOSING VIEWS… 

83

complete picture of vaccine content on Twitter (by including both vaccine-­ supportive and vaccine-hesitant messages), but also provide a more accurate view of what the average Twitter user is likely to be influenced by (given our focus on an entire year’s worth of content rather than a brief window in time). As parents’ hesitancy grows regarding COVID-19 vaccination for adolescents and children, existing literature regarding communication strategies concerning these age groups becomes more valuable. Thus, while the present study does not focus on the COVID-19 vaccines, the findings we present about vaccines given to children and adolescents should be relevant for informing strategies aimed at increasing COVID-19 vaccination uptake among such age groups.

Literature Review Health Information Online Research indicates that health information has become an especially popular topic of online searches, and that online health information, particularly, has taken on a prominent role in medical decision-making (Kata, 2012). In fact, a 2020 study found that over half (55%) of EU citizens reported that they had sought online health information in the previous 3 months (Eurostat, 2021). In the US, a Pew Research study found that about six-in-ten Americans, or greater than 70% of American Internet users, reported relying on the internet for health information (Pew Research Center, 2013). Further, about 60 percent of Americans who reported searching for health information on the Internet also claimed that the information they encountered in their searches impacted their health decisions (Rainie, 2013). In short, health information searches are both common and seemingly influential to subsequent health behaviors. Of course, the advent and evolution of the Internet in general, and social media in particular, have allowed virtually anyone to publicize their opinions for public consumption. Most important for the present chapter, social media have emerged as a critical source for health information. A recent systematic review of social media use for health purposes identified three primary reasons that users turn to social media for health-related information: to seek and share health information, to provide and receive social support (oftentimes from others who are afflicted with the same illness or disease), and to track progress toward health goals—whether

84 

I. HIMELBOIM ET AL.

one’s own or the goals of friends, family, or peers (Chen & Wang, 2021). The review identified the seeking and sharing of health information as by far the most popular of those three uses, providing further evidence that social media have emerged as potentially powerful and influential sources of health information for the general public. This has certainly been the case for the topic of vaccination. Indeed, Poland and Jacobson (2001) have argued that anti-vaccination beliefs are able to spread more widely than ever before due to the ease of use and wide reach of social media platforms for spreading anti-vaccination messages. Common Vaccination Themes To date, there have been a number of studies exploring the communication strategies employed around the topic of vaccination. This work has covered a variety of online and social media platforms, including web pages and Facebook (e.g., Deiner et al., 2017; Smith & Graham, 2017), YouTube (Basch et al., 2017; Briones et al., 2012), Twitter (Blankenship et  al., 2018; Broniatowski et  al., 2018; Cruickshank et  al., 2021), and even Pinterest (Guidry et al., 2019). A common feature among much of this work is its sole focus on the anti-vaccination movement, with specific studies focused on anti-vaccination webpages or web accounts or the rhetoric or keywords associated with vaccine hesitancy. Less work has focused on how vaccine-supportive messages compare to vaccine-hesitant messages in terms of the persuasive strategies employed. Across the range of studies, some common patterns have been identified concerning the tactics employed within messages expressing vaccine hesitancy. Among the most common tactics are (1) calls to action, which implore people not to vaccinate themselves or their child (Moran et al., 2016); (2) the use of personal narratives to tell an emotional story (Davies et  al., 2002); (3) lectures about the side effects or otherwise harmful effects of a vaccine (Davies et al., 2002); (4) a heavy reliance on outside sources—credible or otherwise—to bolster the validity of claims (Moran et  al., 2016); and (5) the pushing of conspiracy theories that often highlight an “us vs. them” divide (Betsch et al., 2010; Kata, 2012; Moran et al., 2016; Smith & Graham, 2017). The above tactics are not employed without warrant. For instance, calls to action provide audiences with clear instructions about the next steps to take or how to be part of a movement, while personal narratives about a side effect are more likely to engage audience emotions than a set of

5  VACCINE SUPPORT AND HESITANCY ON TWITTER: OPPOSING VIEWS… 

85

statistics about how likely or unlikely that same side effect is to occur (Bruner, 1986; Green, 2006). Even a well-crafted conspiracy theory about the danger of vaccines can provide people with a sense of control (Douglas et al., 2017; Kossowska & Bukowski, 2015; Marchlewska et al., 2018), something parents are likely to desire when making health decisions for their children or adolescents. In the present study, we explore these common tropes in vaccine-­ hesitant messaging. However, contrary to much of the work we have encountered, we also explore vaccine-supportive messages that fit within these same thematic areas. In other words, we code for the common tropes noted above, but also for vaccine-supportive messages that (1) have a call to vaccinate, (2) rely on personal narratives to sell their message, (3) describe the effectiveness (rather than side effects) of a vaccine, (4) use outside sources to add validity to their claims, and/or (5) attempt to correct or otherwise counter a conspiracy theory being pushed. Engagement A key feature of social media is the departure they represent from traditional media in terms of the way information is disseminated and consumed. According to the uses and gratifications theoretical perspective (Katz et al., 1973b), media have always satisfied specific needs within individuals. These include the needs for entertainment, learning, and escaping the stresses of daily life (Katz et al., 1973a). However, with the advent of social media, new motivations have emerged, such as socialization, connecting with others in a community, providing or receiving support, connecting with others, and enhancing one’s own status by gaining the respect of others (Huang & Chang, 2020). Further, while traditional media are characterized by a top-down and one-way approach to communication, social media are interactive. They are built around the premise of a meritocracy where all users have a voice. The participatory nature of social media means that the most popular content is that which garners engagement and spreads (Su et al., 2017; Walther & Jang, 2012). Studies of user engagement on social media have had various focal areas (Ksiazek et al., 2016; Su et al., 2020; Yeo et al., 2020). Some have focused on affective evaluation, which is a broad category focusing on the emotional reactions and attitudinal judgements that users share with other users about posts (Alhabash & McAlister, 2015). Arguably the most

86 

I. HIMELBOIM ET AL.

common measurement of affective evaluation is the use of “like” buttons on platforms like Twitter and Facebook. These affective reactions allow users to quickly show their approval (or disproval) of a given message. Other work focuses on message deliberation—a measure that taps into how much thought users are giving a post or message. The commenting features that populate various social media platforms, including Twitter, are the most common means of addressing deliberation among audience members (Alhabash & McAlister, 2015; Kim, 2018). Finally, and most important for the present work, some studies explore engagement in terms of the viral reach of social media messages. Viral reach is focused on the number of people who have been exposed to the content and is similar to the advertising and public relations concepts word-of-mouth (WOM) and electronic WOM (eWOM), which speak to the relative size of an audience for a given piece of content (Alhabash & McAlister, 2015). The higher the viral reach, the more people who are potentially exposed to a piece of communication, and in turn, the higher the number of people who might possibly be influenced by that communication. On Twitter, the viral reach of a message is best expressed by the metric known as retweets (Chung, 2017; Wang et  al., 2019), a means of showing one’s approval of a tweet. Each of the three areas noted above can be communicated via quantitative metrics on social media platforms like Twitter. For instance, Twitter provides a metric of affective evaluation in the form of the number of “likes” associated with a post, a measure of message deliberation through its count of the number of comments, and a measure of viral reach via the tracking of post retweets (Stavrositu & Kim, 2014). Importantly, these quantitative indicators of engagement provide normative cues to audiences about the importance of a post, or even its credibility (Spartz et al., 2017). Such metrics have even been linked to intentions to adopt preventive-health behaviors (Kim, 2018). As viral reach operates as both an indicator of post popularity and a measure of how many users have been exposed to the content, we focus our analysis on the most popular vaccine-related tweets over a one-year time frame, a process that we will detail in the Methods section of this work. Research Questions With the above information in mind, we propose the following set of research questions:

5  VACCINE SUPPORT AND HESITANCY ON TWITTER: OPPOSING VIEWS… 

87

RQ1: Are vaccine-supportive and vaccine-hesitant messages different in terms of the thematic messaging strategies most commonly employed? RQ2: Do the thematic messaging strategies used by vaccine-supportive and vaccine-hesitant messages differ by vaccine type? RQ3: Overall, which thematic messaging strategies are most successful in attracting engagement? RQ4: Does the engagement triggered by the different thematic messaging strategies differ by vaccine type?

Methods Data The social media analytics and archival solution Brandwatch was used to collect all tweets related to three main types of vaccines—early childhood vaccination (including MMR, MMRV, Hepatitis A and B, Diphtheria, Tetanus, and Polio), and HPV vaccination—from January to December of 2018. For each of these datasets, a Boolean query was created to capture all relevant tweets. For the purpose of this study, content with high-level distribution is of greater importance to understanding the spread of vaccine-related information, so we decided to examine the top shared posts. For each week, the top 10 most retweeted posts were collected, resulting in three datasets (one for each of early childhood vaccination and HPV vaccination) of 520 highly retweeted posts each. Even within the most engaged-with content, it should be noted, engagement levels exhibit wide differences, allowing us to examine the RQs of interest. Information about each tweet included a time stamp, tweet content, author name and handle, and number of retweets. Coding Each post was first coded into one of three groups based on the stance taken toward vaccines or vaccination: hesitant (e.g., “Gardasil is poison and Merck’s pathetic commercials feature teen girls AND boys laying a guilt trip on ‘Mom’ and ‘Dad’ for not giving them the jabs. #Evil”), supportive (e.g., “There is NO cure for #polio, but vaccines can protect a child for life. #VaccinesWork #WorldPolioDay”), or neutral (e.g., “HPV vaccine to be given to boys in England”). Only those messages for which two coders were in agreement were coded further (n  =  946). We next

88 

I. HIMELBOIM ET AL.

employed a grounded-theory approach to identify vaccine-focused thematic categories for coding. This process was informed by the previously established literature (as discussed in the literature review) about common vaccine themes. This iterative coding procedure resulted in the following vaccine-focused thematic messaging strategies: “calls to action” (which include any attempt to either encourage or discourage vaccination), “personal narratives about vaccination” (which include both positive and negative narrative stories about vaccine), “reactions or effectiveness of vaccines” (which include negative vaccine reactions and positive consequences of receiving a vaccine), “conspiracy theories” (whether one is promoting a conspiracy theory or debunking one), and the “use of supporting evidence” (whether supportive of or in opposition to vaccines). Two themes were group-specific. For vaccine-hesitant posts, a category emerged about the process used for testing and approving vaccines; for vaccine-supportive posts, a category emerged about the harm of diseases caused by viruses. These categories were only applicable to the supportive and hesitant posts, and not for neutral posts. As these two categories were not as directly comparable as the others, they were each excluded from analyses involving our research questions. Similarly, only posts with a clear stance toward vaccines (supportive or hesitant) were coded further into persuasion strategies (childhood: n = 368 and HPV: n = 383).

Findings Among the analyzed 893 posts, we coded 518 as being vaccine-supportive, 233 as being vaccine-hesitant, and the remaining 142 as neutral in stance. Next, we explored the five common thematic messaging strategies that were used in vaccine-supportive and vaccine-hesitant messages. RQ1: Are vaccine-supportive and vaccine-hesitant messages different in terms of the thematic messaging strategies most commonly employed? A Chi-square test was performed to determine whether the five thematic messaging strategies on Twitter were different between the vaccine-­ supportive and vaccine-hesitant messages. There were significant differences across each: “calls to action” 𝛘2 (1, 751) = 27.03, p