Factors Associated with Vaccination Intention against the COVID-19 Pandemic: A Global Population-Based Study

Several vaccines have been developed for COVID-19 since the pandemic began. This study aimed to evaluate the factors associated with COVID-19 vaccination intention. A global survey was conducted across 26 countries from October, 2020 to December, 2021 using an online self-administered questionnaire. Demographic information, socio-economic status, and clinical information were collected. A logistic regression examined the associations between vaccine intention and factors such as perceptions and the presence of chronic physical and mental conditions. The sample included 2459 participants, with 384 participants (15.7%) expressing lower COVID-19 vaccination intent. Individuals who identified as female; belonged to an older age group; had a higher level of education; were students; had full health insurance coverage; or had a previous history of influenza vaccination were more willing to receive vaccination. Conversely, those who were working part-time, were self-employed, or were receiving social welfare were less likely to report an intention to get vaccinated. Participants with mental or physical health conditions were more unwilling to receive vaccination, especially those with sickle cell disease, cancer history within the past five years, or mental illness. Stronger vaccination intent was associated with recommendations from the government or family doctors. The presence of chronic conditions was associated with lower vaccine intention. Individuals with health conditions are especially vulnerable to health complications and may experience an increased severity of COVID-19 symptoms. Future research should evaluate the effectiveness of interventions targeting the vaccine perceptions and behaviours of at-risk groups. As such, public awareness campaigns conducted by the government and proactive endorsement from health physicians may help improve COVID-19 vaccination intention.


Introduction
The coronavirus disease 2019 (COVID- 19) has had a significant global impact, causing more than 480 million confirmed cases and 6.1 million related deaths since the pandemic started in December 2019 [1]. Health services have been greatly disrupted due to a rapid increase in the burden of the disease [2]. To protect citizens from infection, governments have since implemented preventive measures to minimise social gatherings and movements [3]. However, such measures can be extremely costly from an economic and social perspective.
The health and pharmaceutical sectors developed effective vaccines to reduce incidence and disease severity, especially in vulnerable populations with chronic conditions [4]. The COVID-19 vaccines have been found to be effective in minimising the effects of COVID-19 by strengthening the immune system [5]. Despite its benefits, some individuals continue to express an unwillingness to receive the vaccine due to various reasons, such as safety and cost [6,7]. Therefore, it is imperative to examine the perceptions facilitating vaccination intention. Previous studies have been conducted that have evaluated the intention to receive the COVID-19 vaccination. However, these studies are limited to a specific country, with the findings lacking external validity and generalisability [8,9]. Moreover, the studies failed to take into consideration vulnerable subgroups within the population such as those with chronic physical or mental health conditions, who are more vulnerable to illness [8].
The current study aimed to evaluate the associations between COVID-19 vaccination intention and various factors including demographic characteristics, socio-economic background, and the presence of chronic physical and mental health conditions. Various perceptions influencing vaccination intention were also examined. The objective was to identify subgroups that may be less willing to be vaccinated.

Study Setting
A global survey was conducted by the Association of Pacific Rim Universities (APRU) Global Health Program to determine the factors associated with COVID-19 vaccination intention. The Association is made up of 60 leading universities of the Pacific Rim that have received worldwide recognition for their academic excellence and contributions to research. The APRU Global Health program was launched in 2007-2008 with the aim of addressing global and regional health issues through collaborative research efforts. It covers a variety of academic disciplines including non-communicable diseases, such as mental illness. Data were collected from individuals across 26 study sites, including countries and regions from (1) the Asia-Pacific region: Australia, mainland China, Hong Kong, India, Indonesia, Japan, Malaysia, New Zealand, the Philippines, Russia, South Korea, Taiwan, and Thailand; (2) Northern and Southern America: Canada, Central America, Columbia, Ecuador, Mexico, and Peru; (3) Europe: the United Kingdom, France, Germany, and Italy; and (4) the Middle East: Iraq; Saudi Arabia, and Oman.

Participant Recruitment
From October 2020 to December 2021, an international team consisting of more than 20 investigators distributed surveys online to the general population of each respective country or region through different channels, such as social media and website links sent through emails. Eligible study participants were invited through the investigators' networks using snowball sampling. Individuals were eligible to participate if they were (1) aged 18 years old or above; (2) were capable of comprehending the study; and (3) provided informed consent. The data were stored securely on an online platform, with the database being password encrypted and only accessible by research personnel, to ensure the confidentiality of the collected information. The study was approved by the Survey and Behavioural Research Ethics Committee of the Chinese University of Hong Kong (SBRE-20-035), with ethical clearance obtained for all study sites.

Survey Instruments
An expert panel of primary care professionals, epidemiologists, and physicians was consulted for the pilot-testing and validation of the survey. The survey was available in eight languages and evaluated the association between the presence of chronic conditions and mental illnesses and perceptions related to COVID-19 vaccination intention. Participants provided information on (1) their demographic and socio-economic status: age, sex, race, years of education, type of residence, living status, work/study status, health insurance coverage, welfare benefits, and prior history of influenza vaccination; (2) the presence of chronic conditions: cardiovascular disease (examples: coronary heart disease, heart failure, cardiomyopathy), hypertension, type 2 diabetes, immunodeficiency (or taking medication, such as corticosteroid, that suppresses the immune system), chronic disease of the respiratory system (examples: asthma and chronic bronchitis), chronic liver disease, chronic kidney disease, cancer during past five years, and sickle cell disease; (3) the presence of mental illnesses: depression, mania/bipolar disorder, psychotic disorders (example: schizophrenia), anxiety disorder, post-traumatic stress disorder (PTSD), eating disorder, obsessive compulsive disorder (OCD), substance use disorder (SUD), attention-deficit hyperactivity disorder (ADHD), somatoform disorder, personality disorder, autism spectrum disorder, cognitive disorder, or dementia; and (4) their perceptions of the COVID-19 vaccine in relation to intention: country of production for the vaccine, recommendation from family doctor, recommendation from the Ministry of Health, whether the vaccine has been in use for two years or more, whether there are serious side-effects caused by the vaccine, whether the vaccine is used in other countries, the likelihood of being infected with COVID-19, the vaccine's ease of access (availability out-of-hours or in pharmacies), if the vaccine is free of charge, whether restrictions on group gatherings or travel would be lifted if a high level of vaccination is met. The full questionnaire can be found in the Supplementary Materials.

Statistical Analysis
The data were entered into the IBM Statistical Package for Social Sciences (SPSS) version 25 software to conduct statistical analyses. A descriptive analysis of study participants was performed based on their demographic details and socio-economic status. The associations between vaccination intention and the presence of chronic disease, mental illness, and perceptions related to the COVID-19 vaccine were examined. Vaccination intention was the outcome variable, whilst all other factors were considered as the explanatory variables. Potential associated factors with p values less than 0.20 in the bivariate analysis were included in a binary logistic regression model.
Separate binary regression models were set up to examine the factors listed above. The predictor variable, vaccination intention, was tested for association with the explanatory variables whilst controlling for potential confounding demographic and socio-economic factors. All p values less than 0.05 were considered statistically significant.

Participant Characteristics
A total of 2459 responses were collected. The respondents had a mean age of 29.31, with a standard deviation of 11.07 years. Most of the respondents were female (n = 1526; 62.1%), were of Asian descent (n = 1616; 65.7%), had more than 12 years of education (n = 1995; 81.1%), lived in an urban area (n = 1559; 63.4%), lived with family (n = 1979; 80.5%), and worked full-time (n = 1076; 43.8%). More than half of the participants had full health insurance coverage (n = 1336; 54.3%), whilst 31% were receiving welfare benefits (n = 768; 31.2%). Approximately two-thirds of the respondents had a previous history of influenza vaccination. Detailed demographic and socio-economic statistics of the respondents are presented in Table 1. The number of respondents from each country, the style of the local communication, and the organisation of the national health system of the enrolled participants for major countries are listed in Supplementary Table S1.

Summary of Major Findings
The aim of the study was to evaluate the association between COVID-19 vaccine intention and demographic characteristics (including socio-economic background and physical and mental health conditions) and vaccine perceptions. More than 15% of the respondents reported that they were not willing to receive the COVID-19 vaccine. The correlational analysis indicated that age, gender, years of education, study and employment status, health insurance coverage, welfare benefits, and a previous history of influenza vaccination were significantly associated with a high level of vaccine intention. Additionally, all physical and mental health conditions were associated with a low level of vaccine intention, with the following ones exhibiting significantly lower levels: (1) sickle cell disease; (2) past history of cancer; (3) cognitive disorder or dementia; and (4) personality disorders. Furthermore, respondents' perceptions of the vaccine were associated with a high level of vaccine intention, particularly (1) recommendations from authorities and doctors; (2) the feasibility of getting the vaccine at the desired time and place; (3) whether the vaccine is free of charge.

Explanations and Comparisons with Previous Literature
The findings of the current study align with previous findings which demonstrated that certain demographic and socio-economic variables were predictive of vaccine intention and hesitation. Within the past literature, there has been a focus on men [10], the elderly population [11], and those with pre-existing health conditions [12], as these are considered vulnerable subgroups who might be more susceptible to contracting COVID-19. Men may suffer more dire consequences after contracting COVID-19 due to gender differences in lifestyle habits, such as higher levels of tobacco smoking and alcohol consumption [10]. Commonly reoccurring demographic and socio-economic characteristics associated with high levels of vaccine intention in our study included belonging to older age groups, higher educational attainment, being currently employed, a history of influenza vaccination, and health insurance coverage [13]. The findings are inconsistent across the literature, with some studies finding men to possess a stronger sense of vaccine intention [14], whilst others indicated that women were more likely to receive the vaccination [15]. An association be-tween COVID-19 vaccination intention and education has previously been reported [11,16], which was further supported by this study's finding that higher educational attainment indicated an increased likelihood of vaccine intention. This may be because individuals with a higher exposure to educational material are consequently more aware of the importance of achieving herd immunity, which encourages them to take the necessary precautions to protect themselves from COVID-19.
In our sample, the presence of chronic physical and mental health conditions was associated with vaccine hesitancy and lower levels of vaccine intention. This may be attributed to a lack of trust in the vaccine development process [17], a low perceived risk associated with the disease, as well as key concerns of vaccine safety [18]. The risk of exacerbating pre-existing health conditions because of potential side effects and unknown interactions with medication has been cited as the main reason for vaccine hesitancy [18,19]. Vulnerable populations, such as cancer patients, exhibited the same concern as the general population, with vaccine side effects being the largest predictor of vaccine hesitancy [20]. The vaccine hesitation and the lower level of vaccine intention demonstrated by this subgroup are alarming given their immunocompromised state, as the risk for COVID-19 incidence, severity, and mortality is significantly higher for those with chronic health conditions.
Wariness surrounding new vaccines is not uncommon, with individuals expressing concerns of safety or scepticism of vaccine effectiveness or the perceived risk that the disease poses. Therefore, it is imperative to bolster public confidence in vaccinations by establishing its necessity to accomplishing herd immunity to protect the wider community at large. The core perceptions that influenced higher levels of vaccine intention in this study were advice from doctors and authorities, the feasibility of receiving the vaccine at the desired location and time, and the vaccine cost. In the past literature, there have been mixed findings on an individual's confidence in the COVID-19 vaccination in that it may vary greatly under the recommendation of authorities and healthcare professionals. Some may be more reassured given their trust in said professionals or in the healthcare system, [16,21] though others may express a strong distrust of government bodies and thus doubt the information campaigns that encourage getting vaccinated [14]. Moreover, vaccine intention is impacted by the possibility of obtaining the vaccine at one's convenience. People living in developing countries, subgroups such as ethnic minorities or individuals with lower socio-economic backgrounds, or those living in rural areas may experience greater difficulty in accessing medical care in general and subsequently have reduced access to the vaccine. Additionally, the cost of the vaccine and whether it is costless may contribute to strengthening vaccine intention. If the vaccine is free of charge, the likelihood of improving vaccination rates increases, as individuals are encouraged to seek it out due to its availability [22]. However, in some cases, this is not a decisive factor in decision making, as individuals continue to express strong resistance and low levels of vaccine intention, despite it being readily available [23]. Previous literature has explored the difference in vaccine intention between mRNA and conventional vaccines, and it was found that the novelty of mRNA technology would reduce the acceptability of the vaccine. However, such vaccine hesitancy would be reduced with social conformity [24].
It is likely that vaccination intention was affected by the participants' country of origin, as the style of local communication and healthcare coverage varied among countries. For instance, various fees such as laboratory testing, community isolation, and hospitalisation are covered in the Philippines and Thailand during the pandemic. In contrast, there is not a single nationwide system of health insurance in the US, which would lead to higher medical expenses for those who were not fully covered. A more detailed description of the healthcare coverage, national health systems, and styles of communication can be found in Supplementary Table S1b.

Limitations
Various countries from different regions were involved in the survey, signifying a global collaboration from researchers in different study sites. The survey had been validated and pilot-tested by an expert panel made up of not only professionals with public health backgrounds and epidemiologists but also general practitioners. Nonetheless, there are limitations that need to be noted: (1) the response rate of the survey cannot be evaluated under the use of a consecutive sampling strategy, since the number of participants who received the survey invitation is unknown. (2) The survey responses were submitted over a span two years. The incidence and mortality rate of COVID-19 cases are time-sensitive, and it is difficult to evaluate the interaction between the perception of vaccination, various government policies, and the severity of the pandemic under different waves of COVID-19.
(3) Selection bias may exist due to the use of a non-random sampling strategy, as the online nature of the survey may have made it difficult to attract older participants who are not regular internet users or cannot access the internet. Therefore, the generalisation of the study's findings to other settings and populations in the future should be done with caution.

Conclusions
In this study, individuals with a lower intention to be vaccinated were identified with multivariable logistic regression models. The results demonstrated that people with any types of chronic physical or mental disorders were less willing to get vaccinated, especially those with sickle cell disease, cancer, and cognitive disorders. Since these groups of people often have a weaker immune system, they are more vulnerable to the serious symptoms of COVID-19. Given that participants were more willing to get vaccinated upon recommendations from the government and their family doctors, the authorities should work with the health sector to provide advice to vulnerable people with chronic conditions. Furthermore, people without health insurance coverage and people receiving welfare benefits were less willing to receive vaccination, which is possibly attributable to the cost of the vaccination or the difficulty of making time for the vaccination after long working hours. It is recommended that the vaccination is provided free of charge and that vaccination centres are set up in more convenient locations with later service hours to address these barriers. Since some participants were concerned about the safety of the vaccine in terms of the country of production, serious side effects, and the use of the vaccination in other countries, it is imperative that governments provide ample and factual information on the vaccination to the general public for reference. Further research exploring incentives to motivate people to get vaccinated could be beneficial to ensure that vulnerable groups are protected.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/vaccines10091539/s1, Supplementary material 1 -Questionnaire; Table S1 Number of responses by country; Table S1b Organization of the national health system and style of local communication by country; Table S2 Socio-demographic factors associated with COVID-19 vaccination intention (sensitivity analysis with regions).  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions. All data that were analysed during this study are included in this published article.