key: cord-0749733-zefgp6e3
authors: Dodd, Rachael H; Pickles, Kristen; Cvejic, Erin; Cornell, Samuel; MJ Isautier, Jennifer; Copp, Tessa; Nickel, Brooke; Bonner, Carissa; Batcup, Carys; Muscat, Danielle M; Ayre, Julie; McCaffery, Kirsten J
title: Perceived public health threat a key factor for willingness to get the COVID-19 vaccine in Australia
date: 2021-08-05
journal: Vaccine
DOI: 10.1016/j.vaccine.2021.08.007
sha: cb4d57848622f357f4f8637292c64ff15502a208
doc_id: 749733
cord_uid: zefgp6e3

BACKGROUND: Vaccination rollout against COVID-19 has begun across multiple countries worldwide. Although the vaccine is free, rollout might still be compromised by hesitancy or concerns about COVID-19 vaccines. METHODS: We conducted two online surveys of Australian adults in April (during national lockdown; convenience cross-sectional sample) and November (virtually no cases of COVID-19; nationally representative sample) 2020, prior to vaccine rollout. We asked about intentions to have a potential COVID-19 vaccine (If a COVID-19 vaccine becomes available, I will get it) and free-text responses (November only). RESULTS: After adjustment for differences in sample demographics, the estimated proportion agreeing to a COVID-19 vaccine if it became available in April (n=1146) was 76.3%. In November (n=1941) this was estimated at 71.5% of the sample; additional analyses identified that the variation was driven by differences in perceived public health threat between April and November. Across both surveys, female gender, being younger, having inadequate health literacy and lower education were associated with reluctance to be vaccinated against COVID-19. Lower perceived susceptibility to COVID-19, belief that data on the efficacy of vaccines is ‘largely made up’, having lower confidence in government, and lower perception of COVID-19 as a public health threat, were also associated with reluctance to be vaccinated against COVID-19. The top three reasons for agreeing to vaccinate (November only) were to protect myself and others, moral responsibility, and having no reason not to get it. For those who were indifferent or disagreeing to vaccinate, safety concerns were the top reason, followed by indecision and lack of trust in the vaccine respectively. CONCLUSIONS: These findings highlight some factors related to willingness to accept a COVID-19 vaccine prior to one being available in Australia. Now that the vaccine is being offered, this study identifies key issues that can inform public health messaging to address vaccine hesitancy.

offered, this study identifies key issues that can inform public health messaging to address vaccine hesitancy.

Keywords: COVID-19, vaccination, intentions, concerns, perceived risk HIGHLIGHTS  Perceived public health threat is associated with intentions to vaccinate  Those believing the efficacy of vaccines is made up were less willing to get vaccinated  To protect myself and others was the top reason for getting the vaccine Vaccines are crucial to developing herd immunity, protecting those who are most vulnerable to serious consequences of COVID-19, and to enable easing of national and international travel restrictions and opening up of the economy.

Willingness to get a COVID-19 vaccination (before it became available) has varied considerably across countries over the course of the pandemic. Between April and July 2020, willingness to vaccinate was shown to range from 57.6% in the US, 3 to 64% in the UK 4 , 74% in New Zealand, 5 and 85.8% in Australia. 6 Our research in April showed inadequate health literacy and lower education were associated with a reluctance to be vaccinated 6 and demonstrated an evident need to address health literacy, language and cultural needs of the community in public health messaging about COVID-19. 7 Australian 8 and New Zealand 5 data have shown the most commonly reported reason to get vaccinated were to protect family and self, whilst safety about the vaccine was a chief concern.

Although the vaccine is free, rollout and uptake might still be compromised by concerns about COVID-19 vaccines, so it is important to investigate these. As our previous research during earlier months of the pandemic demonstrated high intentions towards vaccine uptake, 6, 8 it was important to reassess intentions since restrictions in Australia have been relaxed and the immediate threat of COVID-19 has diminished. This study aimed to examine vaccine willingness in the Australian population at two distinct time points in the pandemic:

April 2020, during national lockdown, and November 2020, when there were virtually no cases of COVID-19 in Australia.

An online survey was conducted with two independent, cross-sectional samples at two different time points using the web-based survey platform Qualtrics. This study was approved by the University of Sydney Human Research Ethics Committee (2020/212).

The survey was distributed Australia-wide. Data used for this study were collected between 17-22 April 2020, when national stage 3 restrictions (colloquially referred to as 'lockdown' at that time) had been in place for 3 weeks (i.e. only leaving home for essential reasons) and between 4-18 November 2020, when restrictions were considerably eased across Australia as no locally acquired cases were recorded for the first time since June 2020.

Participants were aged 18 years and older, able to read and understand English, and currently residing in Australia. Participants were recruited via Dynata, who have more than 600 000 online Australian panel members aged older than 18 years who have consented to participate in online research. Panel members were sent an email invitation to participate in the study and received points for completing the survey, which they could redeem for gift vouchers, donations to charities, or money. For the April sample, we set quotas across age, sex and residential state to aim for equal representation across these variables where possible. As the April sample was overrepresented by those having attended University, we purposively set quotas to recruit a nationally representative sample by age, gender and education in the November sample.

Participants completed sociodemographic questions of age, gender, educational status, and residential state and postcode. The area-level socioeconomic status of participants were derived from participants' postcodes. 9 We assessed health literacy using a single item Participants were asked to respond on a seven-point Likert scale about intentions to have a potential COVID-19 vaccine (If a COVID-19 vaccine becomes available, I will get it). In November, the participants were also asked to give a reason for their choice (free-text response). Details of all measures are included in our baseline survey paper on health literacy disparities. 7

Statistical analyses were conducted using Stata/IC v16.1 (StataCorp, College Station, TX).

Descriptive statistics were generated for demographic characteristics, COVID-19 beliefs, and willingness to get a COVID-19 vaccine, and compared between cross-sectional (April and November) samples using independent t-tests for continuous variables and chi-square tests for categorical variables. Differences in willingness to get a COVID-19 vaccine between samples were examined using ordered logistic regression, controlling for demographic variables (age group, gender, education, area-level socioeconomic status, and health literacy adequacy) that may have differed due to sampling at different time points (base model).

Exploratory analyses using ordered logistic regression were then conducted to identify potential factors associated with vaccine willingness by collectively adding COVID-19 belief variables (personal risk belief, vaccine efficacy beliefs, confidence in government, institutional trust, and perceived public health threat) into the base model (full model). To minimise potential multicollinearity due to conceptual overlap between the categorical variable "Belief that the threat of COVID-19 is greatly exaggerated" and continuous variable "Perceived public health threat", the continuous variable was retained and included in the regression model. Participants who responded strongly agree, agree or somewhat agree on the outcome variable were recoded as 'agree'; strongly disagree, disagree, somewhat disagree as 'disagree'; and neither agree nor disagree was standalone. This decision was made due to a smaller number of responses within individual response categories at the lower end of the response scale, and to simplify interpretability of the outcome.

Free-text responses were analysed using content analysis, 12 Table 1 . Respondents sampled in November were selected to be nationally representative of the Australian population (by age, gender and education), and as such were younger (p<.001) and had a more even distribution of education level compared to the April sample which had a high proportion of university-educated adults (p<.001). A slightly greater proportion of the November sample had adequate health literacy compared to the April sample (p=.018). Relative to the group sampled in April, a greater proportion of the November sample believed they were unlikely to get sick with COVID-19 (p=.005), that data about the efficacy of vaccines (in general) is made up (p=.003), perceived the general public health threat of COVID-19 to be lower (p<.001), and that the threat of COVID-19 is greatly exaggerated (p<.001). Willingness to get a COVID-19 vaccine if available was also lower in November compared to April (p<.001).

After adjustment for age group, gender, education, socioeconomic status, and health literacy adequacy, there was evidence of a difference in vaccine willingness between samples ( Table   2 ). Predicted probabilities (estimated at the sample means for all covariate values) suggested a lower proportion of individuals being willing to agree to a vaccine in November (71.5%) compared to April (76.3%), representing an absolute difference of 4.7% (95%CI: 1.4 to 8.0%; see Figure 1 ). The adjusted odds of having a higher level of willingness to get a COVID-19 vaccine in the November sample was 0.78 (95% CI: 0.66, 0.93; p=.006) times that of the April sample.

>Table 2< >Figure 1<

The results from the full ordered logistic regression model are shown in Table 3 for both time points. Older age (i.e., individuals aged 56 to 90 years, relative to all younger age groups), university education (relative to high school education or less), adequate health literacy, higher confidence in government, trust in institutions, and greater perception of COVID-19 as a public health threat were associated with increased odds of being more willing for vaccination (ie in a higher vaccine willingness category). Female gender (relative to male gender), a low personal perceived risk of COVID-19, and belief that the data on efficacy of vaccines is largely made up were associated with reduced odds of being in a higher vaccine willingness category. After controlling for these factors, there was no longer statistical evidence of a difference between timepoint samples in willingness to vaccinate (aOR = 0.96, 95%CI: 0.80, 1.17; p=.70).

To better understand which (if any) of the additional covariates entered into the model were accounting for the differences between timepoints observed from the base model, a leaveone-out approach was employed whereby the additional covariates were individually removed from the model and the resulting coefficient for timepoint compared to the full model. Model coefficients remained consistent using this backwards removal approach, except when perceived public health threat of COVID-19 was removed. When this covariate was omitted, there was statistical evidence (p=.039) of lowered odds of higher vaccine willingness in November (aOR: 0.82, 95%CI: 0.68, 0.99) compared to April. Thus, it appears differences in the perceived public health threat between April and November were driving the observed difference in vaccine willingness (see Figure 2 ).

Of the 1941 participants who provided a response to the question ('If a COVID-19 vaccine becomes available, I will get it') in November, 1799 (93%) provided a written response explaining their choice. A total of 41 codes were generated to capture responses to the question, with 20 'agree' codes and 21 'disagree' codess. >Table 4< Table 4 (full table Appendix 

This Australian data from two cross-sectional samples at two time points during the COVID-19 pandemic demonstrates that from the knowledge, attitude and belief measures included in the study, willingness to vaccinate against COVID-19 is predominately driven by the perceived public health threat of COVID-19 in Australia. In the November sample, when cases were low, restrictions were eased and life was much more 'normal', vaccine willingness was lower than in the April sample, during national lockdown. This relationship has been shown more recently in Australia, with an increase in people wanting to be vaccinated following the latest outbreak in Melbourne. 14 These findings support health behaviour models such as the Health Belief Model 15 and Protection Motivation Theory 16 that perceived risk or severity of a disease is a key driver of intentions, and has been shown in previous COVID-19 vaccination research that having chronic respiratory disease and feeling at increased risk, is associated with greater willingness to be vaccinated. 17 We also found that 18-23% of respondents across the samples believed the threat of COVID-19 is exaggerated and 29-34% believed the effectiveness of vaccines (in general) is made up.

Being female, having a belief that oneself is not likely to get COVID-19, and belief that the data on efficacy of vaccines is largely made up were associated with being less willing to have the COVID-19 vaccine. Older age (55+ years), university educated, adequate health literacy, higher confidence in government, and greater perception of COVID-19 as a public health threat were associated with greater willingness to be vaccinated against COVID-19.

The top reason given for being willing to have the COVID-19 vaccine was to protect myself and others, with concern about the safety of the vaccine being the primary reason for being less willing to have the COVID-19 vaccine.

Our findings showed that perceived public health threat of COVID-19 was associated with intentions to vaccinate, and belief that the threat of COVID-19 is exaggerated was associated with less intention to vaccinate, in both samples. If people are less likely to have a COVID-19 vaccine due to reduced perception of public health threat in Australia, then this could potentially threaten the government aim of achieving 95% uptake of the COVID-19 vaccine. 18 We know from previous infectious disease outbreaks that perceived risk is influential in people taking preventative measures, 19 and people also need to believe that the behaviour, in this case having the vaccine, will be effective in reducing their risk. 20 Previous research has unsurprisingly shown that having personal or direct experience with COVID-19 is associated with greater perceived risk, and perceived risk during the COVID-19 pandemic has been highest in the UK compared to US, Australia, Germany, Spain, Italy, Sweden, Mexico, Japan and South Korea. 21 Given the association between believing vaccine efficacy is made up and reduced intention to vaccinate in our study, there is a need for strategies that reduce people's complacency about the public health threat of COVID-19, as well as correct misperceptions about vaccine efficacy, safety and importance. Strategies could also aim to communicate the cost of not vaccinating to society rather than the individual, particularly for those younger age groups who perceive themselves at less risk and are therefore less willing to get the vaccine.

Previous research has shown that people's prior beliefs are influential in trusting facts, but not in response to communication of uncertainty. 22 This is encouraging as this means being transparent about the uncertainties about the vaccine should not undermine people's trust in the facts or who is communicating these.

Surveillance, based on a review of COVID-19 acceptance literature, include addressing doubts about the pandemic threat by explaining complex concepts in ways that are easy to understand, and addressing low perceived risk by emphasising the broad range of benefits of the vaccine. 23 Similar strategies to those used for getting tested for COVID-19 could also be adapted in public communication, for example using celebrities in television advertisements.

Findings from our previous national surveys have concluded that any communication needs to be designed for those with lower health literacy and education and appropriate for culturally and linguistically diverse groups and Aboriginal and Torres Strait Islander people. 7, 8 Safety concerns were the top reason for participants to be indifferent or disagree with having the vaccine. Since this data was collected, concerns have been raised in the early stages of the vaccine rollout internationally about whether the AstraZeneca vaccine is implicated in thromboembolic events, and some European countries temporarily suspended the use of the vaccine while these events were investigated. 24 In Australia, Pfizer is now the recommended vaccine for those under 50 years of age due to the potentially increased risk following AstraZeneca vaccine in this age group. 25 Although the likelihood of these events occurring is low, suspension of the AstraZeneca vaccine and a change in guidelines has likely impacted public confidence globally and it often takes time for this to recover. 26 Safety concerns were the top reason we found for those unwilling to be vaccinated, so it is of upmost importance to restore public confidence through being transparent about the decisions made and the data collected monitoring the side effects, particularly through the AusVax program. 27 Those chosen to communicate about these issues need to do so with empathy, and not dismiss concerns about the vaccine. 26 Trust in the federal government and in doctors is high in Australia. 6 Given GPs and other health professionals will play a key role in administering the vaccine, restoring public confidence and alleviating concerns may fall to them, so ensuring strong communication channels between doctors and the government is vital.

Our finding that women have lower intentions to be vaccinated than men, has been found in previous research in Australia 28 and in 35/60 studies in a recent systematic review looking at gender differences in intentions to vaccinate against COVID-19. 29 Potential reasons for this finding could be due to women who are pregnant, breastfeeding, or planning children being concerned about vaccine safety, or women have been shown to have less trust in health professionals 30 , although these reasons were not evident in the free text responses of the current study. Due to the exclusion of pregnant women in the large vaccine trials, 31 there is no recommendations about vaccinating women who are pregnant, breastfeeding or planning a pregnancy, but decision aids are available to help women make an informed choice about the COVID-19 vaccine. 32 It is pertinent that we conduct further research with younger women to understand their reasons for vaccine hesitancy.

The current study is strengthened by presenting findings from two large cross-sectional samples. It is important to note that the two samples are different participants and were sampled to different quotas, with the November sample being more representative of the Australian population. While differences have been controlled for where possible, there could also be some unmeasured differences that could not be controlled for statistically.

Participants were only asked for their reasons behind their intentions in the November survey and therefore we cannot draw any conclusions on whether these reasons have changed over the course of the pandemic. This survey measured participants intentions to have the COVID-19 vaccine in the future and not actual behaviour and therefore further research since the rollout of the vaccine is required. Health literacy is a complex construct and to reduce participant burden we used a screening tool rather than a more comprehensive measurement tool. This would not have been sensitive enough to detect differences in health literacy or its association with outcomes. Responses not related to survey question 14 3.8 "I can see into the future" # Appendix 1 shows the full list of codes and their frequency *some free-text responses were allocated more than one code

Vaccinations in United Kingdom

Tracking Australia's COVID vaccine rollout numbers

Attitudes Toward a Potential SARS-CoV-2 Vaccine: A Survey of U.S. Adults

COVID-19 vaccination intention in the UK: Results from the COVID-19 Vaccination Acceptability Study (CoVAccS), a nationally representative cross-sectional survey. Hum Vaccines Immunother

Willingness to vaccinate against COVID-19 in Australia

Health literacy and disparities in COVID-19-related knowledge, attitudes, beliefs and behaviours in Australia

Concerns and motivations about COVID-19 vaccination

Socio-Economic Indexes for Areas (SEIFA)

Brief questions to identify patients with inadequate health literacy

Can This Patient Read and Understand Written Health Information?

Content Analysis

Measuring nominal scale agreement among many raters

Victorian lockdown results in national "urgency" for people wanting the COVID-19 vaccine

The Health Belief Model and Sick Role Behavior

A Protection Motivation Theory of Fear Appeals and Attitude Change

Chronic diseases: Perceptions about COVID-19 risk and vaccination. medRxiv

Why should I get vaccinated for COVID-19? Accessed

Public Perceptions and Knowledge of the Ebola Virus, Willingness to Vaccinate, and Likely Behavioral Responses to an Outbreak

Demographic and attitudinal determinants of protective behaviours during a pandemic: A review

Risk perceptions of COVID-19 around the world

The effects of communicating uncertainty on public trust in facts and numbers

National Centre for Immunisation Research and Surveillance. Recommendations for Developing COVID-19 Vaccine Communication Materials

Covid-19: European countries suspend use of Oxford-AstraZeneca vaccine after reports of blood clots

Australian Government Department of Health. ATAGI statement on AstraZeneca vaccine in response to new vaccine safety concerns

Restoring confidence after the covid-19 Oxford/AstraZeneca vaccine pause will be an uphill battle

National Centre for Immunisation Research and Surveillance. COVID-19 Vaccines

Intention to vaccinate against COVID-19 in Australia

Gender differences in the intention to get vaccinated against COVID-19 -A systematic review and meta-analysis. SSRN

Patient factors that affect trust in physicians: a cross-sectional study

Pregnant People's Paradox-Excluded From Vaccine Trials Despite Having a Higher Risk of COVID-19 Complications

Australian Government Department of Health. COVID-19 vaccination decision guide for women who are pregnant, breastfeeding or planning pregnancy

Methodology, Investigation, Resources, Data Curation Rachael Dodd, Kristen Pickles

Australian Capital Territory 25 (2.2%) 44 (2.3%)

Adequate Health Literacy ^918 (80.1%)

2%) 231 (11.9%) Neither disagree or agree 186 (16.2%) 361 (18.6%) Agree

Assessed using the Single Item Literacy Screener, those responding 'quite a bit/extremely' were categorised as having adequate health literacy; 'not at all/a little bit/somewhat' as inadequate health literacy $ IRSAD: Index of Relative Socio-Economic Advantage and Disadvantage 9 ; a quintile value of 1 represents most disadvantaged (least advantaged) and a quintile value of 5 represents most advantaged

Other / Prefer not to say" was included in the model

We thank Dynata for recruitment services and all study participants.

Estimates were determined at covariate sample means. Error bars indicate the 95% confidence interval.Competing interests: All authors declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: