A study on willingness to take the COVID-19 vaccine at a tertiary institution community in Johannesburg, South Africa

Background South Africa is aiming to achieve herd immunity against the coronavirus disease 2019 (COVID-19) by the first quarter of 2022. The success of the COVID-19 vaccination roll-out depends primarily on the willingness of the population to take the vaccines. Aim This study aimed to examine the willingness to take the COVID-19 vaccine, along with the factors of concern, efficacy and preferences of the individual, which may increase the willingness to be vaccinated. Setting This study was conducted at the University of the Witwatersrand, Johannesburg, amongst adult students and academic and professional staff. Methods A cross-sectional online survey from 27 July 2021 to 14 August 2021 was conducted. We performed descriptive and inferential analysis to determine the factors associated with willingness to take the COVID-19 vaccine. Results A total of 2364 participants responded to a survey link and 82.0% were students, 66.8% were in the 18–29 years age band and females represented 64.0%. A total of 1965 participants (83.3%) were willing to receive a COVID-19 vaccine, the most preferred vaccines were Pfizer (41%) and J&J (23%), local pharmacy (29%) and General Practitioner (GP) (17%) were the preferred places for vaccination and the trusted sources of information on COVID-19 vaccines were the general practitioners (40.6%) and specialists (19.2%). Perceptions that vaccines are safe (adjusted odds ratio [aOR] = 31.56, 95% confidence interval [CI]: 16.02–62.12 for affirmative agreement) and effective (aOR = 5.92, 95% CI: 2.87–12.19 for affirmative agreement) were the main determinants of willingness to taking a COVID-19 vaccine. Conclusion It is imperative to reinforce the message of COVID-19 vaccine safety and efficacy and to include the GPs and the community pharmacies in the vaccination roll-out in South Africa.

vaccinated (89%) and the lowest was in Russia (42%). The South African results showed a 61% intent on being vaccinated. 9 Evidence suggests that conspiracy beliefs and misinformation increase vaccine hesitancy and decrease the willingness to be vaccinated. 10 It's been reported that people who believe in such misinformation are four times less likely to be willing to be vaccinated and more likely to trust the information on social media. 10 The Wellcome Global Monitor 2018 11 reported that most people think vaccines are safe and effective. They further reported that almost 75% of the people, when it comes to health advice, would trust a doctor or nurse over anyone else (family, friends, leaders or famous people) in deciding to be vaccinated or not. The Kaiser Family Fund (KFF) COVID-19 vaccine monitor 12 survey found that most people in the United States, turned to doctors, nurses and other healthcare providers when deciding to be vaccinated. A follow-up survey found that 75% of people wish to obtain their COVID-19 vaccine at their doctor's office, with 15% of the people preferring pharmacies, followed by a hospital (9%). 13 Trust relationships exist not only in interpersonal relationships, such as the doctor-patient relationship but also with institutions. A trust-based relationship occurs when there is vulnerability on part of the truster because of a knowledge imbalance and a lack of power within health institutions and the importance of one's health and well-being. 14 There is limited evidence concerning the people's willingness to take the COVID-19 vaccine in resource-constrained countries. In this study, we examined the willingness to take the COVID-19 vaccine, along with the factors of concern, efficacy and preferences of the individual that may increase the willingness to be vaccinated.

Study design
The study was a cross-sectional online survey of adult students and staff members at the University of the Witwatersrand, Johannesburg, South Africa.

Study population and setting
The study population was adults studying or working at the University of the Witwatersrand. All students, academicians and the administrative and service staff older than 18 years were eligible to participate in the study. This university has a campus in Braamfontein and Parktown with approximately 40 129 registered students, 1642 academic staff members and nearly 3960 administrative and service staff members.

Sample size and sampling
Using the OpenEpi software 15 the sample was calculated based on the following assumptions, Wits community (students, academicians, administrative and service staff) population (N = 42 000), 50% of the participants will be willing to undertake COVID-19 vaccine, marginal error 5%, design effect of 2 and the 70% online survey non-response rate resulted in the estimated minimum sample size of 1300. We expected the distribution of participants to be at the ratio of 9:1 (1170 students and 130 staff). All adult students and staff were invited to participate in the study.

Data collection procedures and ethical considerations
After permission was obtained from the University of the Witwatersrand registrar, the registrar's office sent the survey link to all students and staff. The email contained information on the study and participants had to consent before they could access the survey questionnaire. The questionnaire was based on the number of factors, which includes the acceptability of the COVID-19 vaccine, 16 the influence of social media, 17 personal preferences 13 and accessibility and convenient sites that feels safe. 18 The questionnaire had theoretical and face validity. The online survey ran from 27 July 2021 to 14 August 2021 and was closed thereafter. The survey responses were captured in Research Electronic Data Capture (REDCap) database. 19 The data were stored in a password-protected cloud server. No identifying information was collected from the participants.

Data management and analysis
Data from REDCap was exported to Stata version 14 for management and analysis. The data set contained categorical variables and therefore the descriptive statistics frequencies (proportion) is summarised. The Chi-square test was used to determine the bivariate association of factors associated with willingness to undertake the COVID-19 vaccine. It was decided a priori that age and gender will be forced in the multivariable logistic regression model regardless of the bivariate analysis. The p-value and other factors were entered into the logistic regression multivariable model if the bivariate analysis p-value was < 0.20. The authors reported adjusted odds ratios (aORs) and considered a p-value of < 0.05 statistically significant.

Ethical considerations
Approval to conduct the study was received from the Human Research Ethics Committee (Medical) of the University of the Witwatersrand (M210495).

General characteristics of the study population
The university registrar shared the survey link with 40 129 students, 1642 academic staff and 3960 professional and administrative staff. The survey response rate was 2364/45 731 (5.0%). Table 1 shows that almost all participants responded to all survey questions 2330 (98.6%). Amongst all participants (n = 2364), 82.0% were students, 66.8% were in the 18-29 years age band and females represented 64.0% of all participants. Table 2 further shows that 395/2360 participants (16.7%) indicated that they do not wish to receive COVID-19 vaccines. Amongst the 83.3% who indicated their willingness to take the COVID-19 vaccines -30.7% had vaccinated (fully or partially). The most preferred COVID-19 vaccine types by the participants were Pfizer (41.4%) and J&J (23.3%) vaccines. The least preferred ones were Sinovac (0.8%), AstraZeneca (1.9%) and Sputnik (1.8%). Amongst those willing to take the vaccines their main motivations were concern about their personal well-being (60.4%) and concerns for the well-being of others (61.3%). Side effects (58.0%) and short timelines of the vaccine trials (34.4%) were major concerns of the COVID-19 vaccines reported by the participants.

Perceptions on safety and effectiveness of the COVID-19 vaccines
Of the participants (n = 2359) who responded to the question on the safety of COVID-19 vaccines, 1134 (48.1%) agreed affirmatively that the vaccines were safe and 35.7% of the respondents somewhat agreed that the vaccines were safe. On the other hand, 2357 participants who responded to the vaccine effectiveness question, 1459 (61.9%) agreed affirmatively that the COVID-19 vaccines were effective and 28.4% indicated that they somewhat agreed (Table 2). Table 3 describes the preferred places for vaccination and sources of information. Participant's preferences were a local pharmacy (28.8%), general practitioner (17.2%), hospital (9.9%), local health clinic (8.3%), workplace (7.7%) and vaccination site run by the government (7.8%). The trusted sources of information on COVID-19 vaccines were the general practitioners (40.6%), specialists (19.2%), the Government Department of Health (12.6%) and the medical aids (6.7%). The participants of this study postulated that the main sources of information were by the Government Department of Health (60.3%), via mainstream media (Radio, television, newspapers) (56.4%), social media (48.4%) and by medical aids (26.0%).

Factors associated with willingness to take the COVID-19 vaccines
Overall, the prevalence of willingness to take the COVID-19 vaccines was 83.3%. All reported sociodemographic factors,   (Table 5).

Discussion
Key findings of this study were that more than 80.0% of participants declared their willingness to be vaccinated. They also had a clear preference for Pfizer and Johnson and Johnson vaccines. Mainstream media and social media were sources of information and safety, and efficacy were prominent determinants of willingness to vaccinate. The trusted sources of information on COVID-19 vaccines were the general practitioners (40.6%), specialists (19.2%). Several studies have found a three-way relationship between the factors of perceived severity, perceived susceptibility and perceived benefits, which influenced the willingness to take the COVID-19 vaccine. 2,21,22,23 Although most of the study population felt these vaccines were safe and effective, the concerns around side effects and the rapid clinical trials to test these vaccines are possible barriers; the primary motivation to take the vaccine included personal well-being, as well as the well-being of others. Thus, these factors meet two criteria of the triad (perceived susceptibility and perceived benefits) and will influence their willingness to take the vaccine.
The willingness may also have been influenced because the survey included participants in a tertiary education facility with better access to information and developed reasoning skills. A study in Thailand showed that increased health education and increased willingness in the uptake of influenza vaccinations, 24 whilst a study in America suggested that people who were more knowledgeable about herd immunity, were also more willing to take vaccinations. 25 A study in Nigeria found a high level of hesitancy amongst medical and nursing students, however marginally more under nurses, who had lower risk perception and knowledge of  Vaccine hesitancy is an important hurdle in the willingness to be vaccinated, thus hesitancy is an essential component to address for a successful vaccination roll-out. 27 Razia and colleagues 28 proposed a 5-C model to influence vaccine hesitancy. This model consists of: Confidence (importance, safety and efficacy of vaccines); Complacency (perception of low risk and low disease severity); Convenience (access issues dependent on the context, time and specific vaccine being offered); Communications (sources of information); and Context (sociodemographic characteristics).
In this study, although the main sources of health information concerning the COVID-19 vaccines were obtained via the department of health, social media (Facebook, Twitter, Instagram), mainstream media (radio, television and newspapers), and the medical aids, the most trusted source of the health information was from general practitioners and Other sources 322 13.6 †, Each category was a standalone question.
followed by specialists. This is new information in the South African context. Although many study participants obtained their vaccine-related information from social media, the study participants based in a tertiary institution seem discerning in interpreting the information and not be swayed by the conspiracy theories that abound on social media. 10 The trusted source of information result is also in keeping with Opel et al., 29 who found that a child's healthcare provider is a trusted resource in the parental decision to vaccinate their child. In vaccine-hesitant parents living in the United States, they cite the child's healthcare provider as the key to changing their minds.
Most of the study population were keen on obtaining two types of COVID-19 vaccines currently offered in SA.
Willingness to vaccinate may decrease if the other vaccine products become available in the market and people do not have a choice of the vaccine they will receive.
In countries where the COVID-19 vaccination has been successfully rolled-out to the majority of the population, primary healthcare and community-based teams played a central role in rolling out the vaccination programme. 30,31,32 In this study, to increase the willingness to be vaccinated, most people preferred being vaccinated at their local pharmacy and their general practitioner, as opposed to large vaccination sites, which currently exist.
The findings of this study should be interpreted considering the following limitations. Firstly, a poor response rate (5%) introduces the selection bias of participation of those using the university email, internet access (for those working from home) and those interested in the topic. Secondly, a potential risk of a social desirability response of over-reporting willingness to take the COVID-19 vaccines but the online anonymous survey would have mitigated this bias. Thirdly, even though the university communitycomprises individuals from all walks of life and geographical areas in the Gauteng province, the findings may not be entirely generalisable to the general population but university communities in South Africa. Finally, because of the cross-sectional nature of the study, we are only able to determine the factors associated with willingness to take the COVID-19 vaccine.

Conclusion
This study identified that the safety and efficacy of the COVID-19 vaccines are significantly associated with the willingness to be vaccinated against COVID-19. There is an urgent need to reinforce the communication around the safety of vaccines and effectiveness for the individual and the community, through doctors, media and other approaches to reduce vaccine hesitancy and increase the willingness to be vaccinated. The policymakers should consider incorporating the community pharmacies and the GPs in the primary careled vaccination roll-out, in addition to the current large vaccination sites, to achieve herd immunity as soon as possible.