This study used a cross-sectional design. The calculated sample size was 1635 using the formula (z²pq/e²)*Deff*NR. The z score for a 95% CI was 1.96, and the prevalence of willingness to accept a COVID-19 vaccine in an earlier study was 32.5% (p) [23]. We used a margin of error, e=0.03, for the sampling variation design effect (Deff)=1.6 and a 10% nonresponse rate. The ratio for face-to-face and web-based platform surveys was 2:1, considering the country’s digital divide. The overall participation rate was 93.1% (face-to-face survey: 91.9% and web-based platform survey: 97.7%), and 112 respondents did not consent to participate in the study (web-based platform survey: n=11 and face-to-face survey: n=101). Finally, 26 respondents who were unaware of the COVID-19 vaccine were excluded from the sample. Thus, the final sample for this study was 1497 for analysis (web-based platform survey: n=475, 31.7% and face-to-face survey: n=1022, 68.3%). As the GoB intended to initiate a mass vaccination program from February 7, 2021, we fixed February 1, 2021, to February 7, 2021, for data collection time.

Data were collected using a web-based platform through face-to-face interviews through Google Forms, where the questionnaire was developed in Bengali. Google Forms collects web-based data by prioritizing user friendliness and accessible design. This study conducted a pretest to validate and further improve the survey questionnaire. During the pretest, we tested the technical functionality and usability of the questionnaire. This helped us to identify and address missing questions, response codes, chronology, navigation, and technical performance. Due to the government’s restrictions on movement, safety protocols, and public hesitancy to meet data collectors at their homes, we decided to collect web-based data first to quickly reach a large audience. The survey link was circulated among the research team members’ networks via email, Facebook, WhatsApp, and other platforms. The respondents were also asked to share the link with their networks to reach as many people as possible. We used a single-response-per-participant option for the web-based survey to prevent duplicate entries. The web-based questionnaire contained 12 sections, each displayed on a separate screen for clarity. Upon completion, the respondents received a confirmation message indicating the successful submission of their responses. Additionally, the questionnaire included the respondents’ options to request the study results. Data collection for this study was completed between February 1 and 7, 2021 (web-based data were collected on February 1 and 3, 2021). After 3 days, the data were checked for divisional and age- and sex-specific distributions. To ensure representativeness of the sample, face-to-face interviews were conducted to determine the population’s national representation in terms of age, sex, residence, division, and marital status. Data were collected from 2 randomly selected districts in each of 8 divisions. Four days were allocated to collecting data through face-to-face interviews, and the interviewers maintained proper health measures during the interviews.

For offline data collection, the selection criteria to participate in this study were at least 18 years and older of age and knew about the COVID-19 vaccine, with an additional criterion for the web-based survey respondents’ reading and writing ability.

At the beginning of the data analysis, we used poststratification weighting techniques to reduce biases stemming from sampling, thereby making our study representative of the national population of Bangladesh. We used a weight-adjustment technique for the age variable using the following formula:

where ωi is the weight-adjusted factor, pi is the relative proportion of population characteristics, and si is the proportion of sample characteristics. For example, the sample proportion for individuals aged 18‐24 years was 0.289, whereas the national population proportion, according to Bangladesh’s 2022 Population and Housing Census, was 0.201. Thus, using the formula, we calculated a weight of 1.43 for the age group 18‐24 years. Similarly, we calculated the weight of each age group to ensure that our study population was representative of the national population in terms of age distribution. The weighted samples were used for further analysis.

We first used univariate descriptive statistics (percentage, mean, and SD) for all variables to obtain the background characteristics of the participants and the prevalence of WTP for the COVID-19 vaccine. At the bivariate level, we obtained differentials in WTP for the COVID-19 vaccine using a chi-square test and point biserial correlation. Statistically significant variables (P≤.05) at the bivariate level were entered into a hierarchical logistic regression model to assess the correlates of WTP for the COVID-19 vaccine. We used only descriptive analysis (mean and median) for WTP the highest amount of money. Data were analyzed using SPSS (version 26; IBM Corp).

Ethics approval was obtained from the Bangladesh Medical Research Council (registration 39131012021). Voluntary participation was encouraged, and no incentives were provided to the participants. First, the aims, objectives, potential scopes, and implications of the study’s findings were communicated to the participants, and they then provided written consent. The respondents then participated in the study. Privacy and confidentiality were maintained throughout the study by ensuring that all participant data were anonymized and securely stored. Only authorized research personnel had access to the data, and the participants’ identities were kept confidential.

The background characteristics of the respondents are presented in Table 1, which shows that 340 of 1497 (22.7%) respondents in this study were between 31 and 39 years. Of the 1497 participants, 808 (54%) were male. In terms of religion, 1297 (86.7%) participants were Muslim, and 1061 (71%) participants were married. In total, 1129 (75.3%) had at least a secondary education and higher education. A total of 979 (66.4%) were from rural areas, and the highest percentage of respondents (n=474, 31.7%) were from the Dhaka division (Table 1). In total, 395 (26.4%) were homemakers. The mean number of household members was 4.95, and they had a mean collective monthly family income of 38,500 BDT (a currency exchange rate of 1 BDT=US $0.008 is applicable). Most of the participants (n=1074, 71.8%) identified themselves as having perfect health status (Table 1).

The prevalence of WTP for the COVID-19 vaccine was 772 (51.6%) participants. However, 725 (48.4%) participants refused to pay any money for the COVID-19 vaccine.

Table 2 shows that WTP varied significantly (P≤.05) by religion, marital status, education, place of residence, administrative division, occupation, household income, and coronavirus infection status of respondents, family, and friends. For instance, respondents who identified as others than Muslim, unmarried, had master's and MPhil or PhD-level education, resided in an urban area, were from the Sylhet division, employed in a professional job, and had an income of 40,000 BDT and above had higher WTP at the bivariate level analysis (Table 2).

Table 3 presents the point biserial correlation among WTP and independent variables. Knowledge about the COVID-19 vaccine and vaccination process, preventive behavioral practices related to COVID-19, subjective norm toward the COVID-19 vaccine, and anticipated regret regarding getting infected by COVID-19 had a statistically significant positive correlation with WTP. On the other hand, COVID-19 conspiracy beliefs, attitude toward the COVID-19 vaccine, and perceived behavioral control against COVID-19 vaccination were negatively correlated with WTP.

The significant variables at the bivariate level were entered into a hierarchical logistic regression model to assess the correlates of WTP for the COVID-19 vaccine. Three models were developed for this study. The first model was constructed using the socioeconomic and demographic variables. The second model was built using the variables from the first model along with knowledge of the COVID-19 vaccine and vaccination process, COVID-19 vaccine conspiracy theories, behavioral practices related to COVID-19, and health-related variables. Furthermore, the final model included all variables from the second model as well as components of the TPB. The final model showed that education, administrative division of Bangladesh, knowledge about the COVID-19 vaccine, behavioral practice related to COVID-19, attitude toward the COVID-19 vaccine, subjective norm toward COVID-19 vaccine, perceived behavioral control against COVID-19 vaccination, and anticipated regret regarding COVID-19 infection were statistically significant (P≤ .05) predictors of WTP (Table 4).

According to the final model, respondents of other religions were 40% more likely to pay for the COVID-19 vaccine than Muslims (adjusted odds ratio [aOR 1.4, 95% CI 0.97‐2.01). Similarly, education was a statistically significant predictor of WTP, where respondents who had a graduate degree (aOR 1.98, 95% CI 1.14‐3.45) and master’s and MPhil or PhD degrees (aOR 1.93, 95% CI 1.07‐3.480) had a higher WTP for the vaccine compared to the respondents who had no education. Division was a significant predictor of WTP, which showed that respondents from Chattogram (aOR 2.27, 95% CI 1.61‐4.68), Dhaka (aOR 2.72, 95% CI 1.66‐4.45), Khulna (aOR 3.37, 95% CI 1.84‐6.17), Rajshahi (aOR 3.36, 95% CI 1.19‐5.91), Rangpur (aOR 3.22, 95% CI 1.70‐6.09), and Sylhet (aOR 4.70, 95% CI 2.53‐8.74) had a higher WTP compared to respondents from Barisal. The findings also showed that WTP increased with an increasing level of knowledge about the COVID-19 vaccine (aOR 1.09, 95% CI 1.02‐1.15), behavioral practices related to COVID-19 (aOR 1.11, 95% CI 1.06-1.17), higher subjective norm toward COVID-19 vaccine (aOR 1.25, 95% CI 1.08-1.46), and higher anticipated regret regarding being infected by COVID-19 (aOR 1.17, 95% CI 1.04-1.32). On the other hand, a more negative attitude toward the COVID-19 vaccine (aOR 0.91, 95% CI 0.88-0.95) was associated with decreased WTP.

Although the logistic regression models initially included significant variables at the bivariate level, some variables lost their significance in the multivariate context. These included marital status, place of residence, occupation, income, and COVID-19 infection–related health status of the respondent, family, and friends. Compared to rural areas of residence, both urban and city corporations remain insignificant across the model. The respondent’s family income, notably when it exceeded 10,000 BDT, was negligible in the adjusted models. Among the administrative divisions, the Mymensingh division showed insignificant effects across all models. The COVID-19 infection–related health status of the respondents, family, and friends showed an insignificant association with the outcome variable.

The mean and median WTP the highest amount of money among the respondents were 780.39 BDT and 300 BDT (IQR 150-500 BDT), respectively, for the COVID-19 vaccine. The WTP ranged from 1 BDT to 30,000 BDT.

Figure 1 illustrates the mean WTP for collective monthly family income. There was an increasing trend in WTP, accompanied by rising collective monthly family income, except for the lowest income category (less than 10,000 BDT). Respondents from the 10,000‐20,000 BDT income group had a mean WTP of 391.73 BDT. On the other hand, the mean WTP of the group with an income of 40,000 BDT and above was 1256.10 BDT.

A successful vaccination program depends not only on safe and effective vaccines but also on vaccine hesitancy and WTP. The objective of this study was to assess the prevalence of WTP for the COVID-19 vaccine and its correlates. The study found that 51.6% (n=772) of participants were willing to pay for the vaccine, with an average WTP of 780.49 BDT and a median of 300 BDT (IQR 150-500 BDT). Our findings reveal a lower prevalence and WTP compared to a previous study conducted in Bangladesh. A survey conducted earlier reported that a higher proportion of respondents (68.4%) were willing to pay for the vaccine, along with a higher median WTP of US $7.08 [21]. This difference may be attributed to differences in the sampling methodologies. An earlier study collected web-based data, which likely led to an overrepresentation of individuals who are more educated, financially better off, and digitally literate—characteristics associated with higher WTP. By contrast, our study used both web-based and offline (face-to-face) data collection methods, thereby capturing a more socioeconomically and educationally diverse sample, potentially leading to a more representative estimate of WTP in the general population.

Similarly, there were also relatively modest mean and median values among respondents from low- and middle-income countries who were willing to pay for the COVID-19 vaccine (mean WTP of US $5 in Ethiopia and US $15 for an 80% effective vaccination in Iran) [26]. Another study conducted in Indonesia revealed that the WTP for a COVID-19 vaccine booster dose in this low- and middle-income country remains relatively low, with only 66.2% of the respondents expressing WTP. Among them, the majority (63.5%) indicated a WTP within the range of US $6.71‐$33.57) [3]. As concerns about vaccine resistance and budgetary limitations create the expectation that the government will fully subsidize the vaccine, this expectation contributes to a lower willingness to accept and pay for the vaccine. However, the lower prevalence of WTP is a concern for attaining herd immunity, as it is estimated that 60% of the population must be immunized under a vaccination system to ensure the effectiveness of any vaccine [3].

Our findings suggest that religion has a significant impact on the WTP for the COVID-19 vaccine in Bangladesh. Muslims had a significantly lower WTP than other religions (Hindu, Christian, and Buddhist), and the existing literature supports these findings [3]. The Muslim community is known for its fatalistic view of health and its appreciation, acceptance, and patience regarding their current situation [27,28]. The majority of the population in Bangladesh is Muslim. Muslims may hold various religion-related beliefs, have lower perceptions and levels of knowledge about health and vaccines, and exhibit trust issues toward vaccines developed by non-Muslim countries. These factors can influence their decision to purchase a vaccine and contribute to a lower WTP for the COVID-19 vaccine [29-32].

Our study demonstrates that education has a significant predictive value for WTP for the COVID-19 vaccine [19,20]. In our study, respondents with a graduate and master's or MPhil or PhD-level of education had a higher WTP than those with no education. Educated individuals are more likely to be conscious of their health and more willing to prevent diseases through preventive care such as vaccination [26,33,34]. Education also influences knowledge of health and vaccines, and individuals with a low level of expertise are less likely to develop a WTP for vaccines [34,35]. Educated individuals are more likely to have higher incomes, which gives them greater purchasing power.

In our analysis, household income was found to have a significant positive association with WTP for the COVID-19 vaccine, particularly among higher-income groups. However, this association became statistically nonsignificant after adjusting for behavioral practices related to COVID-19 prevention. This suggests that behavioral practices may mediate the relationship between income and WTP. In our study sample, individuals with higher incomes were more likely to engage in protective behaviors against COVID-19. Consequently, the independent effect of income on WTP appeared to be mediated or absorbed by these behavioral practices, leading to the observed nonsignificant association after adjustment. However, income has proven to be one of the most influential factors in determining WTP for both COVID-19 and non–COVID-19 vaccines [8,18-20].

The study findings revealed a divisional disparity in WTP for the COVID-19 vaccine in Bangladesh. These differences may have occurred because of the socioeconomic and demographic differences among these divisions [36]. Barisal and Mymensingh are the 2 divisions with the highest poverty headcount ratios, whereas Sylhet, Dhaka, Khulna, and Chattogram have significantly lower poverty headcount ratios [37]. Our findings show that Sylhet has a 4-fold higher WTP, which may be attributed to economic and cultural factors. The household income and financial capacity of residents in Sylhet are high because of the strong remittance flow from the United Kingdom and the Middle East, which enhances their ability to pay for health services [38]. Additionally, the distinct cultural norms of the people of Sylhet may contribute to a higher WTP, as studies suggest that perceived norms and the personal environment are positively related to vaccination intention [38]. Households in Barisal and Mymensingh, with comparatively low incomes, may have a lower WTP for the COVID-19 vaccine. Our analysis reveals diverse WTP across divisions; however, a further qualitative study is essential to examine the behavioral and socioeconomic factors that underlie these disparities in WTP.

Apart from socioeconomic and demographic status, knowledge about the COVID-19 vaccine and preventive behavioral practices was associated with higher WTP, and the literature supports our study’s findings [7,39]. Increased knowledge about the COVID-19 vaccine indicates a more reliable understanding of its safety and effectiveness, which in turn improves trust in the vaccine; thus, WTP may increase. Again, people with higher preventive behavioral practices tend to be more educated and health-conscious, so they are expected to have a higher WTP toward any vaccine [40].

The study findings indicate that components of TPB are statistically significant predictors of WTP. The main argument of TPB is that behavioral intention is the most important determinant of health behavior, specifically in the case of WTP [41]. Our findings revealed that respondents with a more negative attitude toward vaccines also had a lower WTP. Attitude acts as a personal evaluation of a behavior; as a result, the intention to pay decreases with a more negative attitude [42]. Similarly, perceived behavioral control suggests that performing a health behavior is not solely within the respondent’s control; in this case, it is challenging to register. In our study, difficulty in web-based registration acts as a structural barrier that reduces WTP for the COVID-19 vaccine. On the other hand, our findings showed that subjective norms regarding COVID-19 vaccination and anticipated regret of contracting COVID-19 were positively associated with WTP for the vaccine. In our study, respondents whose family members approved of their decision to take the vaccine and highly regretted contracting COVID-19 had a higher WTP. This is also a part of subjective norms, as it provides permission from family members or friends for health behavior [41]. Thus, intention toward health behavior related to a disease or vaccine is driven by attitudes toward the disease or vaccine, subjective norms, perceived behavioral control, and other factors [41,43].

This study aimed to assess the prevalence and correlates of WTP for the COVID-19 vaccine in Bangladesh, which can help the GoB and policy makers promote a successful vaccination program on a larger scale for the general population by addressing economic challenges. However, some limitations of this study should be considered prior to the present findings. Our study used a cross-sectional design, which is limited in its ability to generate causal inferences because of temporal issues. Nonprobability sampling was used to reach the study population. We collected self-reported data on health status and other sociodemographic variables, which may have been subject to recall bias. Furthermore, we collected data using both web-based and face-to-face methods. We acknowledge the potential for selection bias that may have resulted from these data-collection methods. Web-based respondents were more likely to be younger, urban, educated, and technology-savvy because of the digital nature of the data-collection platform, which may have ultimately resulted in an overrepresentation of these groups. To mitigate this, the majority of the data (n=1022, 68.3%) were collected through face-to-face interviews to ensure the representativeness of the sample and to determine the population’s national representation in terms of age, sex, residence, division, and marital status. However, it cannot be represented in terms of education, occupation, or income status. Finally, we acknowledge that WTP for a vaccine is context-dependent. Our study’s results may be influenced by the unique sociodemographic and cultural dynamics that emerged during the data collection.

The study findings suggest that the government should introduce targeted educational campaigns aimed at specific demographics, such as religious communities, less educated groups, or those with lower incomes, to address their lower WTP for the COVID-19 vaccine. These campaigns need to be culturally appropriate to increase WTP. Health promotion materials and awareness campaigns as part of the behavior change communication program should be developed to increase knowledge about the COVID-19 vaccine. It will also increase preventive behavioral practices and reduce negative attitudes toward vaccines and vaccine-related conspiracies. Here, mass media can be an effective platform to circulate accurate messages of the COVID-19 vaccine, and community leaders, along with religious leaders, can also be incorporated to mitigate religion-related mistrust and misconceptions. Policy makers should reconsider the web-based registration procedure for vaccine uptake, as it poses a structural barrier to WTP for the COVID-19 vaccine. Based on our findings, an easy alternative system should be introduced for the mass population to achieve the sustainability of the vaccination program. The government may offer small incentives to those who choose vaccination, which will be particularly helpful for lower-income groups. We suggest that policy makers consider a subsidization program that considers socioeconomic stratification, with a focus on highlighting lower-income groups to mitigate the catastrophic income challenge associated with WTP for the COVID-19 vaccine. Otherwise, a reasonable price should be fixed so that the COVID-19 vaccine is affordable. This will help to achieve the highest vaccine coverage and run a successful vaccination program without economic hardship.