Predicting the intention to receive the COVID-19 booster vaccine based on the health belief model

COVID-19 vaccine boosters are recommended because the protection provided by previous doses eventually decreases, posing a threat to immunity. Some people, however, remain hesitant or unwilling to get vaccinated. The present study sought to investigate factors associated with the intention to receive the COVID-19 booster vaccine based on (1) the constructs of the Health Belief Model, and (2) trust in healthcare workers and science. A sample of 165 adults with two doses of the COVID-19 vaccine were recruited using convenience sampling. Data was collected using an online survey from November 2021 to January 2022. The survey included questions about participants' socio-demographic details, health beliefs, trust, history of COVID-19 vaccination, and the intention to receive a third dose of the COVID-19 vaccine. Ordinal logistic regression analysis showed that higher perceived benefits, severity, and trust in healthcare workers, and lower perceived barriers predicted higher willingness to get a third dose of the vaccine whereas perceived susceptibility and trust in science did not. Understanding the factors and health beliefs that underlie vaccine hesitancy are vital when developing effective interventions with the aim of increasing uptake of COVID-19 booster vaccines.


Introduction
The first COVID-19 vaccine was clinically authorised for use in the United Kingdom in December of 2020 (UK Department of Health and Social Care & Hancock, 2020).At the time of present study's data collection, a third dose of the COVID-19 vaccine was recommended for all adults aged 18 and above (UK Health Security Agency, 2021).At present, seasonal boosters are recommended to those over the age of 65, frontline health and social care workers, those who belong in clinical risk groups, those aged 12-64 who are household contacts of immunosuppressed individuals, and those aged 16-64 who are carers or work in care homes for older adults (UK Health Security Agency, 2023).Booster doses are recommended because the protection provided by the previous doses eventually decreases, and also because of a heightened immune evasion by new variants of the virus (Cowling et al., 2022;Pérez-Then et al., 2022).Immunity provided by the booster vaccine reduces rates of hospitalisation, death, and burden on the healthcare system (Andrews et al., 2022).Despite the recommendations, some people remain hesitant to receiving vaccines.Therefore, it is important to investigate the possible causes of vaccination refusal or hesitancy.

Vaccine hesitancy and acceptance
Vaccine hesitancy can be defined as the refusal or delay in acceptance of vaccination despite the availability of vaccines (MacDonald, 2015), and it has been established as one of the top ten threats to global health in 2019 by the World Health Organization (2019).
Several sociodemographic factors might contribute to vaccine hesitancy, including gender.According to reviews, women seem to express more hesitancy towards the COVID-19 vaccine than men (Kafadar et al., 2023;Limbu et al., 2022;Patwary et al., 2022).Other factors associated with COVID-19 vaccination intention are age, profession, and educational level.Reviews have reported older adults being more willing to get vaccinated against COVID-19 compared to younger individuals (for a review of reviews, see Kafadar et al., 2023).Studies have also shown that those working in the healthcare field are more willing to get vaccinated compared to the general population (Kafadar et al., 2023;Limbu et al., 2022).Greater willingness to receive COVID-19 vaccinations has also been reported amongst those with higher levels of educational attainment (Kafadar et al., 2023).In contrast, factors associated with increased COVID-19 vaccination hesitancy include lower income levels, being of non-white ethnicity, and having conservative religious beliefs (Kafadar et al., 2023).

The health belief model and booster vaccination intention
The Health Belief Model (HBM) is a conceptual framework aimed to understand health behaviour and possible causes of non-engagement in recommended preventative health behaviours (Rosenstock, 1974).The HBM includes constructs that determine the likelihood of a person to adopt certain health behaviours.These constructs include perceived susceptibility (belief regarding the likelihood of getting a disease), perceived severity (belief regarding how severe the disease is and what are the social, medical, and clinical consequences if left untreated), perceived benefits (how much the individual believes that the preventative health behaviour will reduce the risk of disease), and perceived barriers (belief regarding how much obstacles or psychological costs stand in the way of the suggested preventative health behaviour) (Champion & Skinner, 2008).Other constructs have later been included: cues to action (Rosenstock, 1974), health motivation (Becker, 1974), and self-efficacy (Rosenstock et al., 1988).
The HBM has been used to predict the intention to receive the first two doses of the COVID-19 vaccine.However, the results are at times inconsistent, which might be explained by different populations and sampling methods.For example, some authors (Huynh et al., 2021;Mahmud et al., 2021;Mahmud et al., 2022;Shmueli, 2021;Wong et al., 2021;Yu et al., 2021) found perceived severity to predict COVID-19 vaccination intention whereas some did not (AI-Metwali et al., 2021;Chen et al., 2021;Lin et al., 2020;Tao et al., 2021).
At the time of this study, the literature on the topic of COVID-19 booster vaccination intention and the HBM was highly limited with only one published peer-reviewed study.Lai et al. (2021) used the HBM to examine factors associated with the intention to receive the third dose of the COVID-19 vaccine amongst those who had received two doses of the vaccine in China.They found that perceived high benefits and low barriers were predictive of third dose vaccination intention, however, susceptibility, severity, self-efficacy, and cues to action were not.Although recently more research has been published indicating that the HBM can predict booster vaccination intention (Alobaidi & Hashim, 2022;Qin et al., 2022;Wang et al., 2022;Wirawan et al., 2022;Zhang et al., 2022), booster vaccination hesitancy has not yet been studied in European populations as the studies have mostly focused on Asian samples.Due to the limited number of studies on the continual use of booster vaccines internationally, more research is needed.

Trust in healthcare workers and science
Previous studies have reported that individuals experienced issues with trust during the pandemic regarding the healthcare system and concerns over the safety and efficacy of COVID-19 vaccines (e.g., Ayappan et al., 2022;Beller et al., 2022;Piltch-Loeb et al., 2021).Specifically, trust in science and healthcare workers have consistently been found to be predictive of COVID-19 vaccination intention for the first two doses of the vaccine (Attwell et al., 2021;Giannakou et al., 2022;Giannouchos et al., 2021;Hromatko et al., 2021;Murphy et al., 2021;Tran et al., 2021;Travis et al., 2021;Wong et al., 2021).
Similar results have also been reported for the COVID-19 booster vaccine.Lee et al. (2022) investigated booster vaccination intention and found that higher levels of trust in science predicted increased willingness to get vaccinated.Similar results on the influence of trust in science on vaccination intention were reported by two other studies (Miao et al., 2022;Toro-Ascuy et al., 2022).Toro-Ascuy et al. (2022) looked at the influence of trust in science and trust in healthcare workers on participants' attitudes towards the third dose of the vaccine and possible annual vaccinations.They found that greater trust in both science and healthcare workers increased willingness to receive the booster and annual doses.Similar results on the influence of trust in healthcare workers or the healthcare system on vaccination intention have been reported by a few other studies (Alshahrani et al., 2023;Ben-David et al., 2022;Miao et al., 2022).Thus, it appears that trust in both science and healthcare workers can increase uptake in booster vaccinations.Although the number of studies investigating the influence of trust in healthcare workers or science on booster vaccination intention is currently limited, the results are promising and warrant further investigation.
To date, the HBM has been successfully used in explaining vaccination intention (Limbu et al., 2022), but it cannot fully account for variation in vaccination intention alone.Therefore, trust in healthcare workers and trust in science will be examined alongside the HBM constructs with the aim of enhancing explanatory power.

The present study
Overall, past research has shown that the HBM, trust in science, and trust in healthcare workers each contribute to predicting vaccination intention for the first two doses of the COVID-19 vaccine.However, previous studies examining the intention for the first dose of the COVID-19 vaccine using the HBM are at times inconsistent.Although more research on booster vaccination intention has recently been published (Alobaidi & Hashim, 2022;Qin et al., 2022;Wang et al., 2022;Wirawan et al., 2022;Zhang et al., 2022) at the time of data collection, the research was limited with only one study examining the intention using HBM (Lai et al., 2021).Furthermore, previous literature on booster vaccination intention at the time of data collection had only focused on non-European populations (Lai et al., 2021).The research investigating the role of trust on booster vaccination intention is currently highly limited, warranting more research.Moreover, no study has yet examined how the HBM incorporated with trust in healthcare workers and science predicts third dose COVID-19 vaccination intention.This research is needed because additional doses of the vaccine are recommended to maintain immunity against COVID-19 (Cowling et al., 2022;Pérez-Then et al., 2022).Therefore, it is important to understand the causes of vaccine hesitancy to increase the uptake of maintenance doses of the vaccine.The current study aims to investigate the reasons behind an individual's intention to receive a third dose of the COVID-19 vaccine using the HBM together with trust in healthcare workers and science.
Based on previous research (Alobaidi & Hashim, 2022;Ben-David et al., 2022;Lai et al., 2021;Lee et al., 2022;Miao et al., 2022;Qin et al., 2022;Toro-Ascuy et al., 2022;Wang et al., 2022;Wirawan et al., 2022;Zhang et al., 2022) it is predicted that first, there will be a positive relationship between perceived severity, susceptibility, benefits, and the intention to receive a third dose of the COVID-19 vaccine.Second, there will be a negative relationship between perceived barriers and the intention to receive a third dose of the COVID-19 vaccine.Third, there will be a positive relationship between trust in healthcare workers and the intention to receive a third dose of the COVID-19 vaccine.Lastly, there will be a positive relationship between trust in science and the intention to receive a third dose of the COVID-19 vaccine.

Design
This study utilised a correlational cross-sectional design where the predictor variables were the HBM constructs (perceived susceptibility, severity, barriers, and benefits), trust in healthcare workers and in science.The outcome variable was the intention to receive a third dose of the COVID-19 vaccine.

Participants
Participants were recruited using convenience sampling.This method of sampling was used due to time constraints and budgetary factors, as this research was unfunded.The inclusion criteria were that the participants were over the age of 18 and had had two, but not three, doses of the COVID-19 vaccine.Participants were excluded if they had had any other number of COVID-19 vaccinations than two.Based on a G-Power analysis of six predictor variables the aim was to recruit 149 participants.After data collection, the sample consisted of 167 participants.Two participants were excluded after having reported receiving only one dose of the COVID-19 vaccine.The final sample included 165 participants (n = 121 female, n = 40 male, and n = 4 non-binary or other).The participants' ages ranged from 18 to 71 years (M = 29 years, SD = 11.1).The sociodemographics of participants are further summarised in Table 1.

Measures
The scale used to measure health beliefs was adapted from The Health Belief Model Questionnaire by Coe et al. (2012).The scale comprised of 17 questions (4 for severity, 3 for susceptibility, 7 for barriers, and 3 for benefits).The questions were measured on a fivepoint Likert scale ('Strongly disagree' to 'Strongly agree').Trust in healthcare workers was assessed with an adapted version of the Trust in Physician scale (Anderson & Dedrick, 1990).The scale included 10 questions measured on a five-point Likert scale ('Strongly disagree' to 'Strongly agree') and had good internal consistency (α = 0.85-0.90;George & Mallery, 2003).The scale used to measure trust in science and scientists was adapted from the Trust in Science Scale by McCright et al. (2013) and had an average to good internal consistency (α = 0.79-0.88;George & Mallery, 2003).The scale comprised of 4 questions on trust in scientists measured on a five-point Likert scale ('Completely distrust' to 'Completely trust'), and 3 questions on trust in science measured on a five-point Likert scale ('Strongly disagree' to 'Strongly agree').
The survey included demographical questions on age, gender, education, ethnicity, and country of residence.Ethnicity was measured based on the recommendations for social surveys conducted in Scotland (Fletcher & Valentova, 2011).
The participants' COVID-19 vaccination history was measured with a question asking how many doses of the COVID-19 vaccine the participant had received (0, 1, or 2).Those who replied 0 or 1 were excluded from the study (n = 2) as the study was only interested in those who had received two doses and thus, may be eligible for the third vaccine dose.The participants were asked about their intention to receive a third dose of the COVID-19 vaccine, if offered.This was measured on a 1 to 5 Likert scale ('Not at all confident' to 'Very confident').

Procedure
The study was approved by the University of the West of Scotland ethics committee 164).An online survey was conducted from November 2021 to January 2022 on QuestionPro.The study was advertised on social media platforms, and participants were also recruited from amongst second year undergraduate psychology students at the University of the West of Scotland as part of their module.At the beginning of the survey, participants provided informed consent.After the survey, they were debriefed to provide educational information about the study, contact details for support services and the researchers, and data withdrawal processes.

Statistical analyses
Percentages for demographic information (gender, age, ethnicity, education, country of residence) and the intention to receive the third dose of the vaccine were calculated.Reverse scoring was implemented for the items that required it (e.g., items 1, 5, and 10 in Trust in Physician Scale; Anderson & Dedrick, 1990).The data were analysed using a cumulative ordinal logistic regression with proportional odds, where the predictor variables were susceptibility, severity, barriers, benefits, trust in healthcare workers and trust in science.The outcome variable was the intention to obtain a third dose of the COVID-19 vaccine.

Vaccination intention
The majority of participants (84.20 %) reported some level of confidence (4 or 5 on a 1 to 5 scale) towards getting the vaccine.Only 12.07 % indicated hesitancy or refusal, and 3.64 % indicated neither confidence nor lack of confidence (see Table 2 for mean scores).Note.Measured on a 1-5 scale where higher scores indicate higher intention or beliefs.

A cumulative odds ordinal logistic regression with proportional odds
was run to determine the effects of perceived severity, susceptibility, benefits, barriers, trust in science and healthcare, on the intention to vaccinate.A correlation matrix indicated no multicollinearity between the predictor variables.The assumption of proportional odds was not met for trust in science.Also, the proportional odds ratios for perceived severity, susceptibility, and barriers differed for participants with the lowest intention to vaccinate (score of 1) compared to everyone else (scores of 2-5).Lastly, the proportional odds ratios for trust in healthcare were quite different for participants who had the highest intent to vaccinate (score of 5) compared to everyone else (1-4).Thus, caution should be exercised when interpreting the results from these independent variables or participants.The Pearson goodness-of-fit test indicated that the model was a good fit to the observed data, χ 2 (650) = 534,210, p = 1.00.The final model statistically significantly predicted the dependent variable over and above the intercept-only model, χ 2 (6) = 69.484,p < .001.
Increases in perceived severity were also associated with increases in the odds of intending to vaccinate, with an odds ratio of 1.759, 95 % CI [1.055, 2.931], χ 2 (1) = 4.969, p = .030(Fig. 3).In contrast, an increase in perceived barriers was associated with a decrease in the odds of intending to vaccinate, with an odds ratio of 0.447, 95 % CI [0.229, 0.874], χ 2 (1) = 5.539, p = .019(Fig. 4).Trust in science and perceived susceptibility did not make statistically significant independent contributions to the model.

Discussion
The current study aimed to investigate the reasons behind vaccination intention for the third dose of the COVID-19 vaccine.It was anticipated that the four HBM constructs together with trust in healthcare workers and in science would predict vaccination intention.The hypotheses were partially supported.The results showed that participants who perceived COVID-19 to be more severe were more likely to intend on getting vaccinated.Moreover, those who perceived the vaccine to be more beneficial and those who considered there were fewer barriers towards getting the vaccine were more likely to intend on getting vaccinated.Participants with higher trust in healthcare workers were also more willing to get the third dose of the COVID-19 vaccine.In contrast, no support was found for the relationship between perceived susceptibility to COVID-19, trust in science, and vaccination intention.
Regarding the HBM constructs, perceived severity, barriers, and benefits, but not susceptibility, were found to be predictive of vaccination intention for the third dose of the COVID-19 vaccine.These findings are consistent with Kabir et al. (2021) and Wong et al. (2021) who found the same constructs to predict vaccination intention for the first doses of the COVID-19 vaccine.However, somewhat different results have been reported by several other studies investigating the intention for the first    dose of the COVID-19 vaccine.For example, some authors found perceived susceptibility to be predictive of vaccination intention whereas the current study did not (e.g., Huynh et al., 2021;Mahmud et al., 2021).Moreover, some did not find perceived severity to be a predictive factor whereas the current study did (e.g., AI-Metwali et al., 2021;Chen et al., 2021).However, the previous studies were conducted at an earlier stage of the pandemic compared to the current study, which might explain the different results.Studies have shown that health beliefs can change through experience with an illness, as shown with influenza.Individuals with a recent influenza infection had higher levels of perceived susceptibility compared to others (Shahrabani & Benzion, 2012).
The results from this study contribute to the growing body of research investigating the intention to receive the third dose of the COVID-19 vaccine, however, again showing discrepancies in results for perceived severity and susceptibility.Similar to the current study, Lai et al. (2021) and Qin et al. (2022) found perceived barriers and benefits to predict vaccination intention for the third dose.They, however, did not find severity to be a predictive factor like the current study did.Most of the previous studies on third dose intention found all four constructs to predict vaccination intention, including perceived susceptibility, which the current study did not (Alobaidi & Hashim, 2022;Wang et al., 2022;Wirawan et al., 2022;Zhang et al., 2022).These discrepancies in results between the current and previous studies may be due to different populations, and therefore, culture.All of the previously reported studies were conducted in Asia with collectivist cultures, where the needs of the group precede the ones of the individual.In contrast, the current study was carried out with a Western European sample which often have lower collectivistic and higher individualistic perspectives.Studies have shown a higher COVID-19 vaccine acceptance in collectivist cultures (Leonhardt & Pezzuti, 2022).
The results of this study add to previous literature demonstrating that higher levels of trust in healthcare workers and the healthcare system predict higher willingness to get vaccinated against COVID-19 (Toro-Ascuy et al., 2022;Tran et al., 2021;Wong et al., 2021).Contrary to what was predicted, no relationship was found between trust in science and COVID-19 vaccination intention.This result is in contrast with previous studies (Attwell et al., 2021;Hromatko et al., 2021;Lee et al., 2022;Murphy et al., 2021;Toro-Ascuy et al., 2022;Travis et al., 2021).This discrepancy between the current study and previous studies investigating the booster dose (Lee et al., 2022;Toro-Ascuy et al., 2022) could be due to measurement differences as different instruments were used to measure trust between the studies.Differences in operationalisation of trust can result in different findings (Furr, 2011).

Study strengths and limitations
The current study contributes to a novel area of booster vaccination intention using the HBM.Moreover, to date, no other study in this research area has investigated vaccination intention using the HBM combined with trust in healthcare workers and science, producing novel information about the factors underlying vaccine acceptance and hesitancy for the booster dose.
Notwithstanding the contributions of this study, this research is limited in, first, the use of sampling method.Convenience sampling was used to recruit the participants, which limits the generalisability of the results to the wider population and therefore, the results of this study must be interpreted with caution (Sedgwick, 2013).The participants were largely recruited from a university so the majority of them were highly educated.Studies have shown a relationship between higher education and higher vaccination intention (Wang et al., 2021).Therefore, the participants might have already had a more positive attitude towards the vaccine which might have influenced the results.
Second, the internal consistency of the constructs benefits and barriers from the Health Belief Questionnaire (Coe et al., 2012) were either not reported or unacceptable (George & Mallery, 2003).Therefore, the findings around benefits and barriers need to be treated with more caution in comparison to the other constructs.This scale was chosen as there was a lack of HBM scales measuring vaccination intention that had all survey items available for use or had information on its internal consistency.More research is needed to develop more robust HBM questionnaires for COVID-19.
Third, the current study used a cross-sectional design where intention to vaccinate was measured at a single time point, prohibiting causal inferences on how vaccination intention later translates into vaccination uptake.This phenomenon where intention does not necessarily translate into future behaviour is called the 'intention-behaviour gap' (Conner & Norman, 2022).The gap has been reported with different vaccines such as the HPV vaccine.For example, one study found intention to vaccinate to predict only 10 % of HPV vaccination uptake (Juraskova et al., 2012).However, the literature suggests that the 'intention-behaviour gap' may not be as prominent for the COVID-19 vaccine, as studies have reported a later vaccine uptake of 81.8 %-94 % amongst those who first intended on getting vaccinated (Griffin et al., 2022;Wang et al., 2022).
Lastly, the use of self-reported data may have been influenced by social desirability, where participants may have answered the questions in a way that is seen favourably by others (American Psychological Association, n.d.).COVID-19 vaccines can be a controversial topic with the public having highly polarised opinions (Jain et al., 2022).Strong levels of encouragement are coming from the government, healthcare providers, and other individuals of the public to get vaccinated, but some remain hesitant or refuse it altogether.Therefore, in the current study, this pressure may have resulted in some participants answering the item regarding vaccination intention in a way that would be seen as favourable by others.However, it is important to note that social desirability may have been decreased by the anonymity of the online survey (Chang & Krosnick, 2009;Lee et al., 2015).

Implications and future research
The results showed that those who perceive getting the COVID-19 booster vaccine as more beneficial and those who had fewer barriers towards receiving it, are more likely to intend on getting vaccinated.The results also found that those who perceive COVID-19 to be more severe are more likely to intend on getting the third dose of the vaccine.These results can be used when aiming to reduce hesitancy and improve the uptake of COVID-19 vaccinations.For example, these results can be used in campaigns promoting the benefits of the COVID-19 vaccine and to decrease any perceived barriers the public may have towards getting the vaccine.Campaigns delivering information about the severity of COVID-19 may also be beneficial in increasing vaccination intention.
Aside from promoting the benefits of the vaccine and aiming to lower the barriers towards getting the vaccine, other factors might be considered as well.The results showed that higher levels of trust in healthcare workers predicts higher likelihood of getting vaccinated.Vaccination uptake could therefore be improved by increasing the public's trust in healthcare workers.Previous research has shown that trust in healthcare personnel can be increased by reducing wait times in the healthcare systems and making healthcare services more accessible (Murray & McCrone, 2015).Trust can also be increased by good healthcare personnel interpersonal skills such as listening to the patient empathetically (Murray & McCrone, 2015).However, due to the current understaffing in the UK healthcare sector (UK Parliament, 2022), good interpersonal skills might be tested which might decrease trust in healthcare workers.Improving the working conditions of healthcare workers could therefore improve the public's trust and willingness to vaccinate.
Current research on booster vaccination intention using the HBM has largely focused on populations in Asia.Therefore, more research is needed including other populations from other regions of the world.Future research might also focus on those eligible for the seasonal COVID vaccinations such as those aged 50 and older and those in risk groups to investigate their perspectives on continued vaccination.Other populations to consider could be those who still have not received their first dose of the vaccine to explore their motives for complete vaccination refusal.

Conclusion
This study sought to investigate the reasons behind vaccination intention for the third dose of the COVID-19 vaccine using the HBM and trust.The results from this study suggest that amongst those with two doses of the vaccine, higher perceived severity of COVID-19, higher perceived benefits of the vaccine, higher levels of trust in healthcare workers, and lower barriers to obtaining the vaccine predict higher intention of getting a third dose of the vaccine.The results have implications for public health campaigns aiming to promote booster vaccination uptake and for possible campaigns for future additional booster doses.

Declaration of competing interest
None.

Fig. 1 .
Fig. 1.Mean scores on trust in healthcare for participants with each level of intent to receive the third covid vaccine Note: error bars represent standard error.

Fig. 2 .
Fig. 2. Mean scores on perceived benefits for participants with each level of intent to receive the third covid vaccine Note: error bars represent standard error.

Fig. 3 .
Fig. 3. Mean scores on perceived severity for participants with each level of intent to receive the third covid vaccine Note: error bars represent standard error.

Fig. 4 .
Fig. 4. Mean scores on perceived barriers for participants with each level of intent to receive the third covid vaccine Note: error bars represent standard error.

Table 1 .
Sociodemographic characteristics of participants.

Table 2
Health belief, trust, and vaccination intention scores amongst respondents.