Though vaccines, in the form of successful mass immunization programs, have saved millions of lives and improved health and wellbeing across the world (1), historically, such successes have constantly been challenged by a minority, yet a significant proportion of vaccine-hesitant individuals and groups for a variety of environmental, cultural, political and psychological reasons (2–5). The World Health Organization has identified vaccine hesitancy as one of the top ten global health threats (6), where vaccine hesitancy has been defined as a delay in acceptance or refusal of vaccination despite its availability (5).

Most studies on vaccine hesitancy come from the Western, educated, industrialized, rich, and democratic (WEIRD) countries (7–9). Thus, very little is known about the contexts of developing, low- and middle-income countries (10). However, the world has witnessed an insurgence of vaccine hesitancy globally during the pandemic of Coronavirus Disease 2019 (COVID-19) (11). A systematic review found that many studies reported a COVID-19 vaccine acceptance rate below 60%, ranging from the lowest 23.6% in Kuwait to the highest 97% in Ecuador (9).

It has been recommended that 60–72% immunity at the population level would be necessary to halt the Coronavirus transmission and community spread (12). The Government of Bangladesh has aimed to achieve an 80% coverage of the COVID-19 vaccination program (13). However, evidence shows that the creation of herd immunity through vaccination is a challenging task when there is vaccine hesitancy (4, 5, 12). Thus, it is crucial to explore vaccine behavior. Unfortunately, there is an apparent dearth of research on COVID-19 vaccination behavior in Bangladesh. The studies conducted in Bangladesh have reported a vaccine hesitancy rate between 29 and 50% (14–17). However, the findings of these studies are not representative of Bangladesh in general as their sample size was small (15, 16), or the data were collected using the online platform (14–16), and there exists a digital divide across the country (18).

Along with quantifying the prevalence of vaccine hesitancy among the population, it is crucial to understand the determinants of the individual decision-making process that result in delay or omission of vaccination (19, 20). Studies found that vaccine hesitancy is predominantly the outcome of the individual decision-making process, which is influenced by individual's feelings about the vaccination or a particular vaccine, barriers, and enablers to vaccinate (10, 19, 20). Thus, it is crucial to understand which psychological drivers determine to delay or refusal of the vaccination (19, 20) so that targeted interventions can be designed to reduce vaccine hesitancy and increase vaccine demand (5, 21).

The health belief model (HBM) is one of the most widely used models in vaccination behavior, particularly in influenza (22), swine flu (23), ebola (22), hepatitis B (24), and COVID-19 (25–27). The key argument of HBM is that the likelihood of an individual adopting a particular health behavior (e.g., getting COVID-19 vaccine) is determined by the perceived susceptibility and severity of illness or disease (e.g., COVID-19), along with the belief in the effectiveness of the recommended health behavior (e.g., COVID-19 vaccination) (28). Thus, the model is comprised of, as applied to COVID-19 and its vaccine, perceived severity of and perceived susceptibility to COVID-19, perceived benefits of and perceived barriers to getting a COVID-19 vaccine, and cues to action which include implicit or explicit incentives or situations that serve to motivate vaccination, such as information from mass media (25).

In contrast, the theory of planned behavior (TPB) argues that behavior is driven by the intention to carry out the behavior, ultimately determined by an individual's “belief structure” (29). As applied to the context of COVID-19 vaccine, belief structure is comprised of attitude toward COVID-19 vaccine (i.e., its perceived necessity, benefit, and effectiveness), subjective norms (i.e., whether significant others support getting a COVID-19 vaccine), and perceived behavioral control (i.e., to what extent COVID-19 vaccination is perceived within the individual's control) (30). However, new components are continually being added to the TPB framework to increase its usefulness (23). For example, Gallagher and Povey (31) found that the addition of “anticipated regret” substantially increased the predictive value of TPB in terms of older adults' intention to get a seasonal influenza vaccination.

On the other hand, Betsch et al. (10) have incorporated and expanded existing vaccination behavior measures and proposed a framework of 5C psychological antecedents of vaccination. It includes confidence (trust in vaccine effectiveness, safety, and necessity, and the system that delivers it), complacency (perceiving the disease as low risk), constraints (perceived low vaccine availability, affordability, accessibility, and other barriers to vaccinating), the calculation (analyzing pros and cons of vaccination), and collective responsibility (willingness to take the vaccine for protecting others via herd immunity). The 5C psychological antecedents have explained a greater extent of vaccination variance than other existing models, though it has not been tested alongside the HBM and the TPB (10).

However, in Bangladesh, the studies conducted to understand vaccine hesitancy-related behavior either did not use any theory (14, 17) or used only the health belief model (15, 16). Therefore, this study aimed to examine the predictability of the HBM, the TPB, and the 5C psychological antecedents to understand vaccine hesitancy, which will ultimately help the Government of Bangladesh design the campaign to reduce hesitancy behavior and increase the uptake of the COVID-19 vaccine.

Given the dearth of research on COVID-19 vaccination behavior in Bangladesh, this study aimed to determine the prevalence of COVID-19 vaccine hesitancy and explore its psychological determinants among Bangladeshi adults. To fulfill that intent, we have utilized three of the most widely used tools in the field of vaccination behavior- the HBM, the TPB, and the 5C psychological antecedents of vaccination, and compared their predictability in terms of predicting the COVID-19 vaccine hesitancy. As a result, we found a 41.1% prevalence of the COVID-19 vaccine hesitancy among our study respondents, a significantly higher estimate than found by Kabir et al. (31%) (15), Ali and Hossain (32.5%) (17), and Mahmud et al. (38.8%) (16). This higher estimation may, in part, be explained by the fact that the existing studies (15–17) conducted a rapid assessment of the situation, which may suffer from respondent selection bias. The existing studies also had a small sample size and conducted an online survey (15, 16). In contrast, data were collected through online and face-to-face interviews from a nationally representative sample covering all eight administrative divisions in our study. In that sense, our study findings provide a more accurate estimate of COVID-19 vaccine hesitancy that is generalizable in the context of the adult population living in Bangladesh.

We found that after controlling the effects of socio-economic and demographic variables, level of knowledge related to COVID-19, its vaccine and vaccination process, and level of preventive practices toward COVID-19, the TPB has the highest predictive power (adjusted R² = 0.43), followed by the 5C model (adjusted R² = 0.32) and the HBM (adjusted R² = 0.31) in terms of explaining total variance in COVID-19 vaccine hesitancy among the adults of Bangladesh. This finding is particularly unique because, to date, no study compared the predictability of these three behavioral frameworks or models in terms of predicting vaccine hesitancy, especially in Bangladesh. Moreover, given the newness of the 5C scale (10), the available literature is still scanty that tests this model's predictive validity (34, 35). However, various other studies that adopted only the HBM and the TPB validate our findings that TPB constructs are a better predictor of vaccination behavior (36), specifically in the intention to vaccinate against swine flu (23) human papillomavirus (37) or in the context of COVID-19 vaccination (25, 26, 30).

According to the findings of the HBM constructs, an increase in perceived benefits of the COVID-19 vaccine, along with increasing perceived severity of and perceived susceptibility to COVID-19, significantly reduced the vaccine hesitancy. On the other hand, an increase in perceived barriers to getting vaccinated acted as a significant vaccine hesitancy promoter. These findings essentially correspond to other studies related to influenza vaccination (38), though Lin et al. (27) found perceived susceptibility was not a significant predictor of the COVID-19 vaccine hesitancy in China. In terms of cues to action, though COVID-19 infection status (of both self and family members) were non-significant predictors at a 5% level of significance. We found that social media (e.g., Facebook) or online news portals as the source of information about the COVID-19 vaccine was the significant predictor, and respondents who heard about the COVID-19 vaccine from social media (e.g., Facebook) or online news portals or blogs were less hesitant. Taken together, these findings suggest that imparting adequate and proper information about the COVID-19 vaccine to the public, along with solid evidence of the safety, efficacy, and benefits of the COVID-19 vaccine, can be a crucial strategy to reduce vaccine hesitancy and increase its demand and actual uptake. In that case, social media and online news portals may act as more effective means than printed newspapers to disseminate COVID-19 related information, as found in our study.

According to the findings of the TPB constructs, an increase of negative attitude toward the COVID-19 vaccine and a decrease in perceived behavioral control significantly increased the COVID-19 vaccine hesitancy. Other studies support these findings (25, 36), though perceived behavioral control was found non-significant predictor in other contexts (23, 39, 40). However, subjective norms in family members' support for having COVID-19 vaccination significantly reduced vaccine hesitancy, corresponding to other studies (25, 36). Consistent with earlier research (23, 31, 41), anticipated regret was a significant predictor of COVID-19 vaccine hesitancy. This finding implies that an intervention to increase the COVID-19 vaccination uptake should circulate the message that it is better to get vaccinated than regret later. In addition, alternative measures should be devised to reduce the barriers related to COVID-19 vaccination, such as online registration to get the COVID-19 vaccine. This provision is inconvenient, especially for older persons, people living in rural areas, and those who do not have internet access (42).

Finally, according to the 5C model, more substantial confidence and higher collective responsibility were significantly associated with reduced COVID-19 vaccine hesitancy, whereas increased complacency and calculation significantly increased vaccine hesitancy. These findings are supported by other studies in COVID-19 and other contexts (34, 35), though calculation and constraints were non-significant predictors of COVID-19 vaccine hesitancy among the nurses in Hong Kong (35). In our study, the constraint related to the COVID-19 vaccination was a non-significant predictor of COVID-19 vaccine hesitancy. These findings suggest that public confidence in the vaccine and the health system that delivers the vaccination service are crucial. Widespread misinformation, conspiracy beliefs, and superstitions regarding the COVID-19 vaccine and its potential health hazards have been found to diminish public trust (43) that need to be addressed through proper communication. Extensive information searching about the subjective utility of vaccination, as evidenced in earlier studies (19, 20, 44), might have resulted in more vaccine hesitancy as more calculation increased COVID-19 vaccine hesitancy among the respondents, along with complacency. These findings warrant the urgency of re-iterating the risk communication and health benefits message of getting vaccinated to mass people.

This study provides evidence that theoretical frameworks like the TPB, the HBM, and the 5C psychological antecedents can explore the psychological determinants that influence a person's vaccination decision-making process. Among the frameworks of determinants, the TPB has the highest predictive power in determining the vaccination decision. These findings can be used to craft targeted interventions to reduce vaccine hesitancy and increase vaccine uptake. Thus, this study's findings will steer Bangladesh's vaccination campaign and those alike to reach the targeted coverage of the COVID-19 vaccination program and, thereby, paving the way for successful prevention of the never-ending pandemic COVID-19.

The dataset presented in this study can be found at https://doi.org/10.17632/jzvbvvknkv.1.

The studies involving human participants were reviewed and approved by the National Research Ethics Committee of the Bangladesh Medical Research Council (BMRC). The patients/participants provided their written informed consent to participate in this study.

MBH conceptualized the study. MBH, MAH, and MZA analyzed and interpreted the data and drafted the manuscript. MSI, SS, MMF, SR, and AAM revised the manuscript critically for valuable intellectual content and approved the final version to be published. All the authors designed the study, collected the data, and remained in agreement to be accountable for all aspects of the work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.