COVID-19 Vaccine Hesitancy—A Scoping Review of Literature in High-Income Countries

Vaccine hesitancy forms a critical barrier to the uptake of COVID-19 vaccine in high-income countries or regions. This review aims to summarize rates of COVID-19 hesitancy and its determinants in high-income countries or regions. A scoping review was conducted in Medline®, Embase®, CINAHL®, and Scopus® and was reported in accordance with the PRISMA-SCr checklist. The search was current as of March 2021. Studies which evaluated COVID-19 vaccine hesitancy and its determinants in high-income countries (US$12,536 or more GNI per capita in 2019) were included. Studies conducted in low, lower-middle, and upper-middle income countries or regions were excluded. Factors associated with vaccine hesitancy were grouped into four themes (vaccine specific, individual, group, or contextual related factors). Of 2237 articles retrieved, 97 articles were included in this review. Most studies were conducted in U.S. (n = 39) and Italy (n = 9). The rates of vaccine hesitancy across high-income countries or regions ranged from 7–77.9%. 46 studies (47.4%) had rates of 30% and more. Younger age, females, not being of white ethnicity and lower education were common contextual factors associated with increased vaccine hesitancy. Lack of recent history of influenza vaccination, lower self-perceived risk of contracting COVID-19, lesser fear of COVID-19, believing that COVID-19 is not severe and not having chronic medical conditions were most frequently studied individual/group factors associated with increased vaccine hesitancy. Common vaccine-specific factors associated with increased vaccine hesitancy included beliefs that vaccine are not safe/effective and increased concerns about rapid development of COVID-19 vaccines. Given the heterogeneity in vaccine hesitancy definitions used across studies, there is a need for standardization in its assessment. This review has summarized COVID-19 vaccine hesitancy determinants that national policymakers can use when formulating health policies related to COVID-19 vaccination.


Introduction
Since its first reported case in December 2019, the coronavirus-2019 (COVID- 19) pandemic has culminated in nearly 179 million infections and 3.88 million deaths globally as of 24 June 2021 [1]. Lockdowns, social distancing measures, and movement restrictions were implemented as a result to abate the spread of infection worldwide [2]. The aftermath of the pandemic has negatively affected global economies. For example, the Internal Labor

Selection of Sources of Evidence, Data Charting Process, and Data Items
Citations retrieved from the four databases were exported into Endnote Software Online (Clarivate Analytics, Philadelphia, PA, USA). Duplicated citations were removed prior to screening of articles. Two independent reviewers (J Aw and JJB Seng performed the initial pilot exercise for the screening of the first 200 records (based on title and abstract). Thereafter, the titles and abstracts of all retrieved articles from the four databases were screened by the same reviewers independently. The full-text articles of potentially relevant articles were evaluated prior to inclusion in this review. All disagreements in the inclusion phase of the review were discussed to reach a consensus. For discrepancies which could not be resolved between the two reviewers, arbitration was made with a third independent reviewer (SYS SEAH).
To chart data from the included articles, a standardized Microsoft Excel data collection sheet was used. This information included the name of author, title of study, publication year, sample size, study design and methodology, characteristics of patient population, tools used to evaluate COVID-19 vaccine hesitancy, reported hesitancy rates, and determinants associated with increased COVID-19 vaccine hesitancy.

Critical Appraisal of Individual Sources of Evidence
The risk of bias appraisal for included studies was not performed as this was not the objective of this scoping review.

Summary and Synthesis of Results
Descriptive statistics were used to summarize the characteristics of studies included in this review. Vaccine hesitancy rate was reported from individual study according to the definition described in each study. In studies which reported only vaccine acceptance rates, vaccine hesitancy rates were computed using the formula: [100 (%)-vaccine acceptance rates (%)]. In cohort studies which reported longitudinal rates of vaccine hesitancy, the mean vaccine hesitancy rates were extracted. Other variables collected included the study design and methodology, characteristics of participants and determinants of COVID-19 vaccine hesitancy. Graphical charts and tables were used to present the results.
There is no widely accepted definition for cut-off with regards to a high vaccine hesitancy rate. Assuming COVID-19 vaccines can stop transmissibility and that COVID-19 has a R 0 of 2-3.5, a 60-70% vaccination uptake is estimated for herd immunity [16]. We therefore define high vaccine hesitancy as 30 or more percent in this review.
A narrative summary of factors associated with increased COVID-19 vaccine hesitancy was presented. The determinants of vaccine related hesitancy were grouped into three main categories: contextual influences, individual/group influences, and vaccine/vaccination specific issues, as proposed by the Strategic Advisory Group of Experts (SAGE) on immunization [6,17]. These determinants were reported as per described in individual studies.
A framework diagram was used to summarize the most frequently studied determinants of COVID-19 vaccine hesitancy in high-income countries or regions.

Data Availability Statement
Data analyzed in the study is included in the published article and Supplementary File Tables S1 and S2.

Results
A flowchart for inclusion of articles in this review is illustrated and of 2237 citations retrieved, a total of 97 articles were included based on the inclusion criteria ( Figure 1). The percentage of agreement between the reviewers during the inclusion was 90.7%.

Results
A flowchart for inclusion of articles in this review is illustrated and of 2237 citations retrieved, a total of 97 articles were included based on the inclusion criteria ( Figure 1). The percentage of agreement between the reviewers during the inclusion was 90.7%. Figure 1. Flowchart for retrieval of articles. a Records excluded: Did not evaluate COVID-19 related vaccine hesitancy (n = 1425); Evaluated vaccine hesitancy in non-high-income countries (n = 34); non-English articles (n = 23); irrelevant systematic reviews/meta-analyses (n = 3); b Records excluded: Studies included countries other than high-income countries (n = 14); studies are editorials, commentaries, news article and/or opinions without original data (n = 132); Study looked at willingness of guardians enrolling children in vaccine trials (n = 1); studies retracted (n = 3)

Characteristics of Included Studies
In the summary table of the characteristics of included studies, approximately half of the included studies were conducted in Year 2021 (50.5%) while the other half were performed in 2020 (49.5%) ( Table 1). Most of the studies were conducted in North America (43.3%) and Europe (34.0%). Four studies involved cross-continent collaborations. Of note, Records after removal of duplicates (n = 1730) Figure 1. Flowchart for retrieval of articles. a Records excluded: Did not evaluate COVID-19 related vaccine hesitancy (n = 1425); Evaluated vaccine hesitancy in non-high-income countries (n = 34); non-English articles (n = 23); irrelevant systematic reviews/meta-analyses (n = 3); b Records excluded: Studies included countries other than high-income countries (n = 14); studies are editorials, commentaries, news article and/or opinions without original data (n = 132); Study looked at willingness of guardians enrolling children in vaccine trials (n = 1); studies retracted (n = 3).

Characteristics of Included Studies
In the summary table of the characteristics of included studies, approximately half of the included studies were conducted in Year 2021 (50.5%) while the other half were performed in 2020 (49.5%) ( Table 1). Most of the studies were conducted in North America (43.3%) and Europe (34.0%). Four studies involved cross-continent collaborations. Of note, the two countries with the highest number of studies were U.S. (n = 39, 40.2%) and Italy (n = 9, 9.3%). Cross-sectional study design (n = 75, 77.3%) and online survey methodology (n = 87, 89.7%) were most frequently described in studies included. Further details of included studies are available in Supplementary File Table S2. a Finland (n = 1), Greece (n = 1), Israel (n = 1), Japan (n = 1), Kuwait (n = 1), New Zealand (n = 1), Qatar (n = 1), Saudi Arabia (n = 1). b Parents/caregivers (n = 1), firefighters / first responders (n = 1), Blacks living with HIV (n = 1). c Studies using a combination of either of online, face to face, paper questionnaire, telephone or mail methods included. d Studies involving participants from multi-continents (n = 4) are omitted in the table. e Studies reporting guardians' vaccine hesitancy for their wards included in the analysis (if a study reports both guardians' hesitancy for wards and for themselves, the higher of the 2 is taken into consideration for count here). # Each category has been reported as per described in individual study and the net total will be more than 97 due to some studies having combination methods.

Vaccine Hesitancy Rates across Studies
Pertaining to the definition of vaccine hesitancy used to derive its proportion, just slightly over half of the studies (51.5%) conformed to SAGE working group definition of vaccine hesitancy a . a Detailed data available in Supplementary Materials S2

Contextual Related Factors
A total of 25 themes were identified and grouped under the eight sub-categories in "Contextual determinants of vaccine hesitancy" ( Table 2). "Sociodemographic related variables", "policies / politics related factors", and "communications and media environment related factors" were most frequently studied themes.  Reasons for vaccine hesitancy/acceptance were explored in 21 studies through open ended questions but only 7 (7.2%) attempted to describe methods on thematic analysis. Of these, only four are qualitative studies with well described methodologies [23,[25][26][27]. Cross-sectional studies (n = 75, 77.3%) were the most frequent study designs while online surveys (83,85.6%) were the most frequently used methods a .
Pertaining to the definition of vaccine hesitancy used to derive its proportion, just slightly over half of the studies (51.5%) conformed to SAGE working group definition of vaccine hesitancy a . a Detailed data available in Supplementary Materials Table S2.
Two studies found an increased vaccine hesitancy mainly in nursing staff among healthcare workers with clinical fronting roles [67,72].

Group/Individual Related Factors
A total of seven sub-categories of factors with 22 themes were identified for the "Individual/group determinants of vaccine hesitancy" (Table 3). "Beliefs, attitudes about health and prevention", "past experiences with vaccinations", and "health-system and providers-trust and personal experience" were most well-studied.

Discussion
This review has highlighted a few salient points and some research gaps. Firstly, it showed that despite the variable rates of vaccine hesitancy across highincome countries or regions, nearly half of studies reported vaccine hesitancy of 30% or more. Our review discovered that only slightly more than half of the studies conducted on COVID-19 vaccine hesitancy conformed to the SAGE proposed definition. In those studies which did not conform, participants who expressed being "unsure" instead of rejecting the vaccine were excluded in the hesitancy rate, leading to a potential falsely reassuring low hesitancy rate.
Studies conducted among high income regions across four continents revealed a high proportion of studies with high vaccine hesitancy mostly in Asia and North America. Countries with the highest vaccine hesitancy rates included UAE, U.S., Hong Kong, and Italy. Compared to low-income countries or regions, the current vaccine hesitancy rates in high income countries or regions are worrisome.
The varying vaccine hesitancy rates across countries or regions are complex and may partly be attributed to differences in ideological beliefs, demographics, and context specific factors, as seen for other vaccinations. For example, vaccine hesitancy appears to have a lesser impact on general vaccine uptake rates in lower-middle income countries or regions and affects lower socioeconomic status individuals to a greater extent [112]. The reasons have been linked to disparities in access, cost, and awareness of vaccines [113]. In contrast, individuals residing in more affluent countries or regions tend to be more vaccine hesitant due to concerns related to the safety of vaccines [114]. This is especially so in the current choices of vaccines made with newer technology which raised doubts and long term safety concerns [115].
The global vaccination census showed that the share of population fully vaccinated against COVID-19 stood at 18.3% in high income countries or regions as of May 2021 [116]. Of note, the proportions of population fully vaccinated against COVID-19 in U.S., Italy, Hong Kong, and UAE were at 36%, 13.7%, 10%, and 38.8% as of May 2021 respectively, reflecting our review results with only UAE bucking the trend [116]. In spite of this, there were flattening of the epidemic curves from February onwards in the U.S. and Israel after the commencement of vaccination exercises, reinforcing the importance of vaccination [1].
The second point our study highlighted was to summarize determinants of COVID-19 vaccine hesitancy that were the frequently studied ( Figure 3). Females, being younger, having a non-Whites ethnicity and having a lower socioeconomic status (e.g., lower education or income levels) were common demographics identified with COVID-19 vaccine hesitancy ( Table 2).
Literature discussing higher vaccine hesitancy in females suggested the underlying reasons attributable to lower perceived risk of COVID-19, higher beliefs in conspiracy related theories about the pandemic compared to their male counterparts [117] and concerns about safety of vaccination in pregnancy and breastfeeding [118].
The association between younger individuals and COVID-19 vaccine hesitancy may be a result of increased public health focus on vaccinating the elderly (due to their risk for severe COVID-19 outcomes) and the lack of outreach on COVID-19 vaccination in social media platforms which they commonly frequent [119]. Vaccines 2021, 9, x 13 of 21 Literature discussing higher vaccine hesitancy in females suggested the underlying reasons attributable to lower perceived risk of COVID-19, higher beliefs in conspiracy related theories about the pandemic compared to their male counterparts [117] and concerns about safety of vaccination in pregnancy and breastfeeding [118].
The association between younger individuals and COVID-19 vaccine hesitancy may be a result of increased public health focus on vaccinating the elderly (due to their risk for severe COVID-19 outcomes) and the lack of outreach on COVID-19 vaccination in social media platforms which they commonly frequent [119].
With respect to ethnicity, Blacks have been shown to have increased mistrust in COVID-19 vaccination with possible reasons due to racism, discrimination and mistreatment within the healthcare systems [120]. We should extrapolate and observe for similar associations to all at-risk populations so that governments and healthcare professionals alike can assess and direct efforts on improving COVID-19 vaccination uptake rates.
Our review also discovered that users of social media/internet as a primary source of COVID-19 related information were more prone to increased vaccine hesitancy. With the advent of infodemic on these non-traditional media platforms, innovations on ways to deliver accurate and timely health information by traditional and non-traditional platforms have become incredibly important. Employing active strategies such as pre-emptive cognitive inoculation techniques and pre-bunking techniques have also been  With respect to ethnicity, Blacks have been shown to have increased mistrust in COVID-19 vaccination with possible reasons due to racism, discrimination and mistreatment within the healthcare systems [120]. We should extrapolate and observe for similar associations to all at-risk populations so that governments and healthcare professionals alike can assess and direct efforts on improving COVID-19 vaccination uptake rates.
Our review also discovered that users of social media/internet as a primary source of COVID-19 related information were more prone to increased vaccine hesitancy. With the advent of infodemic on these non-traditional media platforms, innovations on ways to deliver accurate and timely health information by traditional and non-traditional platforms have become incredibly important. Employing active strategies such as pre-emptive cognitive inoculation techniques and pre-bunking techniques have also been suggested to tackle misinformation [121]. Clear and honest communications form an important bridge between building public trust and reinforcing positive health behaviors or compliance with COVID-19 vaccination [122].
In addition, the other determinants previously mentioned should also be systematically addressed. While it is not within the scope of this review, the way different themes are being measured such as knowledge about COVID-19 disease and vaccination, is an important area of research impacting on the study of vaccine hesitancy across different populations. Our review noted that most studies used self-designed instruments in the evaluation of COVID-19 vaccination knowledge which limits cross-comparison of knowledge levels across populations. Future research should consider developing a standardized instrument for the assessment of knowledge of COVID-19 vaccine and disease which can potentially be adapted for future pandemics.
Hopefully, the summary of these determinants will allow policymakers at the national level to deep dive into local context and conduct multi-pronged, multi-tiered studies coupled with interventions to overcome vaccine hesitancy in high income countries.
With the ongoing vaccination drive globally and evolving landscape for COVID-19, it remains premature to conclude the real-world impact of vaccine hesitancy on the true uptake of COVID-19 vaccination. Uptake can be confounded by logistic and administrative challenges in vaccine deployment, vaccine production capacity issues from manufacturers, affordability of vaccines and global allocation of vaccines in the context of limited supplies [123]. This was observed in U.A.E. which had one of the highest percentages of population fully vaccinated for COVID-19 (39.3%) globally in May 2021 despite a reported high vaccine hesitancy [124]. In contrast, the percentage of fully vaccinated population in Canada, which had the lowest vaccine hesitancy, was only 3.3% in May 2021 [124].
Nonetheless, vaccine hesitancy studies will continue to provide insights into possible future directions to drive vaccination efforts. In planning vaccination programs, two considerations related to COVID-19 vaccination are important moving forward. Firstly, if COVID-19 vaccinations can stop transmissibility of COVID-19, at least 60-70% of population needs to be vaccinated [125]. Secondly, in the scenario where COVID-19 vaccines reduce only disease severity but not transmissibility, identifying targeted groups for priority vaccination will become the de facto strategy. Studying vaccine hesitancy across patient subgroups who have the highest mortality and morbidity related to COVID-19 infection will be of paramount importance.
Several research gaps related to COVID-19 related vaccine hesitancy were identified in this review. Currently, there are limited studies which have evaluated longitudinal changes in COVID-19 related vaccine hesitancy. COVID-19 vaccine hesitancy may fluctuate or even increase due to fatigue with lockdown and preventive measures, or secondary to increased complacency coupled with reduced risk perceptions with a long duration of pandemic [126]. Future studies may want to consider evaluating the variation in vaccine hesitancy at different timepoints in the COVID-19 pandemic, given its continued waves of outbreaks in different countries or regions currently. There is also paucity of data related to COVID-19 vaccine hesitancy among pediatric groups as well as a lack of assessment of parental concerns of COVID-19 vaccinations in children. As data from studies evaluating the safety and efficacy of COVID-19 vaccination among children emerges soon, it is an important research area to explore. In addition, this review had noted a dismal number of qualitative studies on COVID-19 vaccine hesitancy. Qualitative studies often enable new themes to be identified which is important for comprehensiveness [127]. Albeit challenges abound in conducting qualitative research due to current climate of social distancing measures and lockdowns, some recommended ways to overcome them include use of digital text communications, video diaries and photovoice, where physical interaction can be minimized [128].
This review is not without its limitations. Firstly, the determinants of vaccine hesitancy listed in this review were factors identified from most studies which employed online surveys predominantly. While this was inevitable given the lockdowns and travel restrictions imposed during the COVID-19 pandemic, population groups with limited access to the internet such as older adults, may not be comprehensively captured.
The findings from online studies may be influenced by self-selection bias, survey fraud, and inability of respondents to seek clarity on questions [129]. Among the included studies only a small proportion of online survey studies reported their findings according to the CHERRIES checklist of internet E-surveys [130]. Future studies should consider adopting this checklist to enhance the scientific rigor of their findings. Moreover, among the included studies, we had noticed a significant number of studies not reporting the education level of participants recruited (n = 38; 39.2%, data available in Supplementary  Table S2). A higher level of education in the participants is associated with the possession of correct information on COVID-19 and less susceptibility to misinformation [131].
Secondly, the grey literature as well as literature from pre-print servers were not searched in this review. Future systematic reviews which seek to evaluate vaccine hesitancy among specific populations or perform an updated review should consider searching these resources to improve the comprehensiveness of the search.
Thirdly, among the 57 themes of vaccine hesitancy found in the systematic review, 26 (45.6%) themes had fewer than five studies. The percentages in Tables 2-4 with themes having fewer than five studies were n = 9 (36%), 40.9% n = 9 (40.9%), and n = 8 (80%) respectively. A possible insufficient exploration of a theme in the included studies has to be taken into consideration while interpreting and contextualizing the results to individual country.
We would also like to point out a preponderance of studies done in the U.S. exploring the two sub-categories on "policies/politics" and "influential leaders, gate-keepers and anti or pro-vaccination lobbies". Due to geopolitical differences, generalizability of these themes may be limited.
Lastly, assessments of methodological quality of the included studies, presentations of strength of statistical associations with vaccine hesitancy and meta-analyses of the vaccine hesitancy rates were not performed as these were not the primary aims of this scoping review. Moreover, the heterogeneity in the definition and assessment of vaccine hesitancy in different studies would not have allowed a meaningful meta-analysis. Researchers who are planning to investigate COVID-19 vaccine hesitancy may want to consider adopting the standardized definition of vaccine hesitancy from SAGE workgroup in future studies. This will facilitate and enable future systematic reviews and meta-analyses to evaluate the variation in vaccine hesitancy rates across countries or regions, as well as the temporal variation in vaccine related hesitancy.

Conclusions
Overall, COVID-19 vaccine hesitancy remains a highly prevalent problem in high income countries or regions. Individuals who were younger, females, non-Whites, and have a lower education or income levels, were more prone to vaccine hesitancy. Trust at different systems levels seem to play an important role in modifying vaccine hesitancy as well. Other commonly studied factors associated with vaccine hesitancy included a history of not receiving influenza vaccination, a lower self-perceived risk of contracting COVID-19, a lesser fear for health outcomes or COVID-19, not believing in the severity of COVID-19, having concerns about the rapid development of COVID-19 vaccines as well as disbeliefs in the safety and effectiveness of the vaccines. Healthcare administrators need to be cognizant of these determinants of vaccine hesitancy when formulating policies related to COVID-19 vaccination and public health messages.