Introduction

The global landscape of vaccine acceptance has been significantly impacted by the COVID-19 pandemic.1 It was paramount to investigate COVID-19 vaccine acceptance at the second phase of the vaccine roll-out to the public.1 While vaccine hesitancy is a historical phenomenon, the novel nature of COVID-19, coupled with the rapid development of vaccines, amplified concerns and uncertainties.1 Despite concerted efforts to accelerate vaccine development and distribution through initiatives like the Access to COVID-19 Tools (ACT) Accelerator, vaccine uptake faced challenges in many regions globally.1 The WHO and the United Nations Children’s Fund have raised concerns regarding a decline in COVID-19 vaccination rates, exacerbating the situation as vaccine hesitancy has progressively increased in several countries globally.1 2

Over 158 candidates of COVID-19 vaccine were developed at the time of the roll-out through the ACT Accelerator led by the WHO and partners, a global collaboration to accelerate development, production and equitable access to COVID-19 tests, treatments and vaccines.3 The first vaccine pillar led by CEPI, Gavi (the Vaccine Alliance) and the WHO, to facilitate collaboration and increase the speed of the search for an effective COVID-19 vaccine and ensure its global delivery.3 At the same time, supporting the building of manufacturing capabilities and buying supply, ahead of time, so that COVID-19 vaccine doses can be distributed fairly in the places of greatest need, worldwide.3 Hence, vaccination is our best chance towards ending the pandemic.

Previous studies on vaccine acceptance and theories of health behaviour, such as the health belief model or protection motivation theory, have identified many factors that influence the acceptance or uptake of a pandemic vaccine.4 This includes the risk perception of the disease, perception of vaccine safety and efficacy, general vaccination attitude, past vaccination history, recommendations from doctors, price, vaccination convenience and sociodemographic characteristics.4 Although high acceptance was recorded among healthcare workers during the first phase of the vaccination, vaccine acceptance has faced challenges since the discovery of vaccines and that of COVID-19 is not likely to be an exception especially in the community.5–9 Factors influencing vaccine acceptance include age, gender, income, level of education, trust in the government and employer’s recommendation.10–14 Evidence of willingness to vaccinate against COVID-19 was insufficient in 19 countries that were surveyed, highlighting the need for targeted interventions to increase and sustain public acceptance of an incoming COVID-19 vaccine.10 Our aim was to determine the level of awareness, uptake and acceptability of COVID-19 vaccine among the public during the vaccine roll-out period in Zamfara state, Nigeria.

MethodsStudy area

This study was conducted in Gusau and Bungudu local government areas (LGAs) of Zamfara state, Nigeria. Gusau is the state capital and a major urban centre, while Bungudu is a rural LGA with an agriculturally driven population. Both areas face significant challenges in terms of healthcare access, poverty and limited educational opportunities.

Zamfara state is one of the north-western Nigerian states dominated by the Hausa Muslims. It has 14 LGAs within a land area of 37 931 km2, a population of about 5 307 154, with farming being the most common occupation. All the 14 LGAs in Zamfara state have a varying degree of security threat in the form of banditry and kidnapping majorly affecting healthcare, economic activities and education.

Study design

This was a cross-sectional study.

Study population

The study participants were drawn from household heads residing in Gusau and Bungudu LGA of Zamfara state while those living in security compromised areas were excluded from the study.

Sample size determination

The sample size was calculated for the state with the household as the sampling frame. At 95% confidence level, 5% margin of error, population proportion of 80.0% (0.8) from a previous Nigerian vaccine acceptance study,5 and an unlimited population size, Z is 1.96.

z is the z score

ε is the margin of error

n is sample size

is the population proportion

Using non-response rate of 7% in a similar Nigerian study5

Sampling technique

A multistage sampling was used. In the first stage, nine wards were selected in Gusau and Bungudu LGA after excluding security compromised wards. In the second stage, 10 settlements were randomly selected from each ward. In the third stage, the first household in each settlement was randomly selected and the remaining households were selected by skipping 10 households. The heads of the household were interviewed after getting consent, 11 households were selected in each settlement, and a total of 990 households were selected for the study.

Data collection tools

A standardised semistructured electronic questionnaire was used to collect data on the sociodemographic characteristics, knowledge on COVID-19, uptake, awareness, and acceptability of COVID-19 vaccine. We adapted a questionnaire from another COVID-19 acceptance study in China.4

Data collection procedure

Data were collected between 12 October and 20 December 2021. Data collectors were trained to administer the questionnaire in both Hausa and English. The questionnaire was pretested in a ward that was not included in the main study. Three experts assessed both the face and internal validity of the questionnaire. The modified questionnaire was used to collect data from the household heads. A household was visited a second time on the same day before replacement in the case where the selected household head was absent. The respondents with poor knowledge or no awareness of COVID-19 or its vaccine were sensitised by the data collectors before completion of the survey.

Data analysis

Data were analysed using Epi Info and Microsoft Excel. Eleven (11) age variables were missing, and imputation of the mean age (47) was used to complete the missing data.

Descriptive statistics were done to calculate the frequencies and proportions of the sociodemographic variables; age, sex, education, tribe, religion, marital status, employment, owning mobile phone, owning a radio, owning a television (TV), and having a chronic illness, awareness of COVID-19 pandemic, awareness of COVID-19 vaccine, total knowledge score, uptake of COVID-19 vaccine, acceptance of COVID-19 vaccine, testing for COVID-19, recommending COVID-19 vaccine to friends, allowing family to receive COVID-19 vaccine, preferred route of vaccine administration, receiving any other vaccines in the past, the nine trust variables, total trust score, the seven attitude variables, total attitude score and level of confidence in individuals and organisations. Data were summarised in tables and figures.

Inferential statistics were executed using bivariate and multivariate analysis. Our objective was to develop an association model to assess the relationship of risk factors (nine sociodemographic, three acceptance, total trust score category, total attitude score category and three awareness variables) with COVID-19 vaccine acceptance. To address this objective, we conducted a cross-sectional study to identify the predictors of COVID-19 vaccine acceptance. In the unadjusted bivariate analysis of the instrument variables, univariate selection with p<0.05 was used to select variables for the multivariate analysis. We used a multivariable logistic regression model and the fully adjusted approach to explore the interaction of the effects between COVID-19 vaccine acceptance and other variables. The results are summarised in tables with unadjusted ORs (crude ORs), adjusted ORs, fishers exact 95% CIs, and considered a p value of <0.05 as statistically significant.

Variable measurement

The attitude of the respondents towards COVID-19 was measured using a five-point Likert scale with the options strongly agree, agree, neutral, disagree and strongly disagree given the points 5, 4, 3, 2 and 1, respectively, in positively framed questions and the reverse in negatively framed questions. Also, the level of trust was measured with a five-point Likert scale with options very much, much, some, little and very little given the points 5, 4, 3, 2 and 1, respectively, in positively framed questions and the reverse in negatively framed questions. The mean attitude was categorised into positive, neutral and negative, while trust was categorised into high, moderate and low. Knowledge was assessed by awarding 1 mark to each correct answer with 19 as the highest possible total score. The total knowledge scores were categorised into low, moderate and high.

Ethical considerations

Informed written consent was obtained from all participants, who were informed of their right to withdraw from the study at any time. No identifying information was collected, and all data were anonymised to ensure participant confidentiality. Participants who lacked knowledge of COVID-19 or its vaccine were provided with relevant information to ensure they could make an informed decision about participation. All study data were securely stored on an encrypted, password-protected computer, and only authorised personnel had access to it. Regular audits were conducted to ensure compliance with data security protocols.

Patient and public involvement

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Results

This study identified the level of uptake, acceptance and factors associated with COVID-19 vaccine in Zamfara state during the vaccine rollout in September 2021. We approached 990 respondents, 91.9% (910) of them gave a written consent and responded to our questions, and we had a non-response rate of 8.1% (80). Among the 8.1% (80) that did not respond, 98% of them were men, 77% of them were within the age category of 30–39 years, and 91.3% (73) of them were from Gusau LGA, which is an urban area. Of the 910 respondents, 11 had missing age that was addressed by imputation of the mean 47.

Descriptive statisticsSociodemographic characteristics

The respondents had a median age of 48 years (IQR: 37–55), with 78.1% (711) being men, and 90.3% (822) aged 30 years or older. Among the 897 respondents, 88.0% (789) had received formal education at least up to the primary school level, while 12.0% (108) had no formal education. The majority of respondents were of Hausa ethnicity and Muslim faith, with both groups representing 97.9% (891) of the sample. Additionally, 95.5% (869) of respondents were married.

In terms of employment, 89.2% (812) were employed, with the largest proportion (44.6%, 362) working as farmers. Ownership of communication devices was common; 64.3% (585) owned a mobile phone, 90.8% (826) owned a radio and 31.3% (285) owned a TV. Furthermore, 11.2% (102) of the respondents reported having at least one chronic illness. Table 1 provides further details on the sociodemographic characteristics of the respondents.

Table 1

Sociodemographic characteristics of the respondents

Awareness, knowledge, uptake and acceptance of COVID-19 vaccine, previous history of any vaccination

Table 2 presents the data on COVID-19 vaccine uptake and acceptance among respondents who had not yet received the vaccine. The vast majority, 95.7% (793), were aware of the pandemic, and 83.8% (695) were aware of the availability of the COVID-19 vaccine. However, 99% (785) of respondents demonstrated an individual knowledge score below 11.4, representing less than 60% of the total possible score.

Table 2

COVID-19 vaccine awareness, knowledge, uptake and acceptance among the respondents

At the time of the survey, only 8.9% (81) of respondents had been vaccinated. Among the 829 unvaccinated individuals, 10.1% (84) expressed willingness to receive the vaccine during the week of the interview, 38.4% (318) indicated they would do so in the following week, 27.4% (227) within the next month, 6.4% (53) within the next year, while 12.2% (101) stated they would refuse the vaccine, and 5.6% (46) were unsure. Additionally, a majority (68.2%, 565) had never been tested for COVID-19.

In terms of vaccine recommendation, 86% (713) of respondents reported they would recommend the vaccine to friends, and 92% (763) indicated they would allow their family members to receive the vaccine. Regarding vaccine preferences, 64% (436) had no preference for the route of administration, and 67% (456) had no preference for the location of vaccine administration. A significant proportion, 86% (713), had received other vaccinations in the past.

Attitude, trust and confidence

Regarding trust in health authorities in table 3, 33.1% (301) of respondents had ‘very little’ trust in their doctors, while 25% (227) reported ‘much’ trust in patent drug sellers. Furthermore, 31.4% (286) had ‘some’ trust in the local primary healthcare department, and 21.3% (194) had ‘much’ trust in the state health department. Notably, 23.7% (216) expressed ‘very much’ trust in the Nigeria Centre for Disease Control, while 35.4% (322) had ‘very much’ trust in the National Primary Healthcare Development Agency (NPHCDA). Similarly, 24.1% (219) reported ‘very much’ trust in the Presidential Task Force on COVID-19. However, 33.3% (303) and 33.5% (305) expressed ‘little’ trust in religious and traditional leaders, respectively. Overall, 51.8% (430) had a low total individual trust score of less than 3 representing <60% of the total trust score.

Table 3

Trust in individuals, leaders, health institutions and attitude towards COVID-19

As shown in table 3, a significant portion of respondents expressed neutral or negative attitude towards the perceived severity of COVID-19. Specifically, 28.5% (259) were ‘neutral’ on the statement that their health would be severely damaged if they contracted COVID-19, and 31.2% (284) were ‘neutral’ on whether COVID-19 is more severe than the common cold. Additionally, 40.8% (371) disagreed with the statement that COVID-19 would cause serious damage in their community, while 38.7% (352) disagreed with the statement that COVID-19 has never existed in Zamfara state, and 40.4% (368) disagreed with the claim that COVID-19 no longer exists in the state. Most of the respondents (62.4%, 517) had a moderate total attitude score.

In figure 1, respondents ranked their confidence in various individuals and organisations to manage the COVID-19 pandemic. A majority (37%) ranked doctors as their top choice, 29% ranked patent drug sellers second, 38% ranked the state health department third, 31% ranked the Nigeria CDC fourth and 30% ranked the NPHCDA fifth. The Presidential Task Force was ranked sixth by 38%, religious leaders were ranked seventh by 55%, and traditional leaders were ranked eighth by 57%. The detailed ranking results are shown in figure 1.

Figure 1

The level of confidence in individuals and government bodies (N=910) shows the confidence rankings assigned by respondents to various individuals and government bodies regarding health-related advice or decisions. The bars represent the percentage of respondents selecting each rank, colour-coded as follows: dark green for first choice, light green for second, pale green for third, yellow for fourth, blue for fifth, peach for sixth, pink for seventh and red for eighth choice. Entities ranked include ‘Your own doctor’, ‘Traditional Leaders’, ‘State Health Department’, ‘Religious Leaders’, ‘Presidential Task Force’, ‘Patent Drug Seller’, ‘Nigeria Centre for Disease Control (NCDC)’ and ‘National Primary Healthcare Development Agency (NPHCDA)’. The figure highlights that respondents show the highest confidence in their own doctors, with 37% ranking them as their first choice, while traditional and religious leaders were ranked lower, indicating less trust.

Bivariate and multivariate logistic regressionBivariate analysis

The bivariate analysis in table 4 indicated several factors were significantly associated with COVID-19 vaccine acceptance at p<0.05. Respondents aged 30 years and older had higher odds of accepting the vaccine (OR: 3.39, 95% CI 2.06 to 5.57, p<0.001) compared with those under 30. Similarly, having a chronic illness was strongly associated with vaccine acceptance (OR: 5.16, 95% CI 1.89 to 19.66, p<0.001). Ownership of a TV (OR: 23.42, 95% CI 8.77 to 87.99, p<0.001) and a mobile phone (OR: 62.15, 95% CI 29.79 to 129.67, p<0.001) were also significantly associated with vaccine acceptance.

Table 4

Bivariate analysis associations of COVID-19 vaccine acceptance in Zamfara state

Awareness of the COVID-19 pandemic (OR: 6.55, 95% CI 3.31 to 12.99, p<0.001) and awareness of the COVID-19 vaccine (OR: 1.84, 95% CI 1.19 to 2.84, p=0.009) were positively associated with vaccine acceptance. Previous testing for COVID-19 was also a significant factor (OR: 1.55, 95% CI 1.03 to 2.33, p=0.040).

Higher levels of trust were strongly linked to vaccine acceptance. Respondents with medium-to-high trust had much higher odds of accepting the vaccine (OR: 23.35, 95% CI 11.27 to 48.38, p<0.001) compared with those with low trust. A medium-to-high total attitude score was similarly associated with vaccine acceptance (OR: 3.27, 95% CI 2.26 to 4.72, p<0.001).

Vaccination with other vaccines in the past (OR: 13.92, 95% CI 8.92 to 21.71, p<0.001), willingness to allow family members to receive the COVID-19 vaccine (OR: 12.81, 95% CI 7.38 to 22.26, p<0.001), and recommending the vaccine to friends (OR: 5.04, 95% CI 3.30 to 7.71, p<0.001) were all strongly associated with COVID-19 vaccine acceptance.

Multivariate analysis

The multivariate analysis in table 5 depicted individuals aged 30 years and older were significantly more likely to accept the COVID-19 vaccine compared with those younger than 30 (aOR=2.39, 95% CI 1.16 to 4.94, p=0.018). Ownership of a mobile phone (aOR=25.35, 95% CI 11.23 to 57.23, p<0.001) and a TV (aOR=3.72, 95% CI 1.09 to 12.69, p=0.036) was also strong predictor of vaccine acceptance. Although awareness of the COVID-19 pandemic and its vaccine increased the likelihood of vaccine acceptance, these associations did not reach statistical significance at the multivariate level.

Table 5

Multivariate analysis: factors influencing COVID-19 vaccine acceptance in Zamfara state

Higher levels of trust in public health institutions and key individuals were significantly associated with vaccine acceptance. Respondents with medium-to-high trust had more odds of accepting the vaccine compared with those with low trust (aOR=7.41, 95% CI 3.10 to 17.74, p<0.001). Similarly, individuals with medium-to-high (positive) attitudes towards COVID-19 (the disease or pandemic) were significantly more likely to accept the vaccine compared with those with low (negative) attitudes (aOR=1.82, 95% CI 1.06 to 3.11, p=0.029).

Furthermore, individuals who had previously received other vaccines (aOR=2.20, 95% CI 1.09 to 4.43, p=0.027) and those who had been tested for COVID-19 (aOR=2.00, 95% CI 1.10 to 3.66, p=0.023) were significantly more likely to accept the COVID-19 vaccine. However, willingness to recommend the vaccine to friends or allow family members to be vaccinated was not significantly associated with personal vaccine acceptance.

Discussion

This study assessed the uptake and acceptance of the COVID-19 vaccine among household heads in Zamfara state, northern Nigeria, while examining sociodemographic, attitudinal and behavioural factors associated with vaccine acceptance. Our findings provide important insights into the factors influencing COVID-19 vaccination in a typical northern Nigerian population known to exhibit traits of vaccine hesitancy and rejection, contributing to a growing body of literature on vaccine hesitancy in low and middle-income countries (LMICs).

Despite the high level of awareness regarding the COVID-19 pandemic and the availability of vaccines (95.7% and 83.8%, respectively), the actual vaccine uptake was remarkably low, with only 8.9% of respondents having received the vaccine. This finding is consistent with other studies conducted in sub-Saharan Africa, USA, Kuwait, Jordan, Italy, Russia, Poland and France where the uptake was low.13 15–17 This highlighted the gap between awareness and action in terms of vaccine uptake. The high level of awareness in Zamfara state may be attributed to extensive media campaigns, yet this does not seem to translate into vaccination behaviours, a pattern that has been observed in other parts of Nigeria.18 Zamfara state is one of the northern Nigerian states with documented history of vaccine rejection.19–21

Studies on acceptance of COVID-19 vaccine have been conflicting due to many reasons; population, method of questionnaire administration, knowledge and even country. Our study considered different levels of hesitancy by avoiding using the binary yes or no question (used by several studies) to assess acceptance.15 16 22 Any person who was aware of COVID-19 vaccine availability, had access to it, and did not take it had some element of hesitancy.23–26 Those who responded they would take the vaccine within the week were non-hesitant. Those that would take the vaccine the following week were ‘mildly hesitant’, those that responded they would not take it till next month were ‘moderately hesitant’, those that said they would take it next year were ‘severely hesitant’. Then we had the ones that refused to take the vaccine, classified as ‘total rejection’, and those that did not know were classified as ‘unsure’. Like this study, COVID-19 hesitancy was reported in several LMICs by other studies.27

This study demonstrated that age, having a chronic illness, and access to media (ownership of a phone and TV) were significant predictors of vaccine acceptance. Household heads aged 30 years and older were more than two times as likely to accept the vaccine compared with younger individuals (aOR=2.39, 95% CI 1.16 to 4.94). This aligns with the existing literature that suggests older adults perceive themselves as being at higher risk for severe COVID-19 outcomes, which enhances vaccine acceptance.12 Additionally, individuals with chronic illnesses (though not statistically significant) were more likely to accept the vaccine (aOR=3.21, 95% CI 0.93 to 11.05), reinforcing the well-established association between perceived susceptibility to illness and vaccine uptake demonstrated in another study in Ethiopia.11 Previous studies in Nigeria have shown that older adults, and those with comorbidities, have more severe disease and outcome, ‘self-preservation’ may play a role in the acceptance exhibited by these groups.28–30 Owning a TV and mobile phone were also strongly associated with vaccine acceptance (aOR=3.72 and aOR=25.35, respectively). Access to media likely plays a crucial role in disseminating information and countering misinformation. This finding emphasises the importance of leveraging mass media platforms to enhance public health messaging in remote areas, especially given that misinformation has been a major barrier to vaccine acceptance in Nigeria.

Trust in health institutions emerged as a critical determinant of vaccine acceptance. Respondents with medium-to-high levels of trust in public health authorities were seven times more likely to accept the vaccine (aOR=7.41, 95% CI 3.10 to 17.74). This finding echoes previous studies in other LMICs, which have shown that trust in government and health systems is key to overcoming vaccine hesitancy.6 7 Conversely, low levels of trust in religious and traditional leaders were reported, aligning with global reports that have documented lower trust in non-health actors during the COVID-19 pandemic.31

Attitudes toward the severity and impact of COVID-19 also played a role. Those with medium-to-high positive attitudes towards the virus were more likely to accept the vaccine (aOR=1.82, 95% CI 1.06 to 3.11). This underscores the importance of addressing misconceptions about the disease. A significant proportion of respondents expressed neutral or negative beliefs about the severity of COVID-19, which could explain some of the hesitancy observed. Vaccine hesitancy is often fuelled by perceptions that the disease is not severe, especially in regions like northern Nigeria, where the direct impact of COVID-19 may have been perceived as lower than in other parts of the country.7 31–33

Previous vaccination history was a strong predictor of COVID-19 vaccine acceptance, with respondents who had received other vaccines being more likely to accept the COVID-19 vaccine (aOR=2.2, 95% CI 1.09 to 4.43). This finding highlights the role of routine immunisation as a potential pathway to increase COVID-19 vaccine uptake. Strengthening routine immunisation programmes in Zamfara state could provide a platform for future vaccine campaigns. Beyond COVID-19 vaccine introduction, strengthening immunisation programmes will likely increase the likelihood of acceptance of new vaccines for diseases like malaria, Lassa fever and dengue fever.

Interestingly, while respondents expressed a willingness to recommend the vaccine to friends and allow family to get vaccinated, these attitudes were not significantly associated with their own acceptance, which is in contrast to an online Nigeria study.34 This suggests that social desirability may play a role, where individuals advocate for vaccination publicly but remain hesitant to receive it themselves as reported by the US study where the motivations of ‘hesitant adopters’ of the COVID-19 vaccine were explored.35 This discrepancy points to the need for tailored health communication strategies that not only promote vaccine uptake but also address personal hesitancy.36

The findings from this study have several implications for public health interventions in northern Nigeria. First, efforts to increase vaccine uptake must focus on building trust in health authorities and leveraging media platforms to dispel misinformation. Second, targeted interventions should be developed to engage younger individuals, as they were found to be less likely to accept the vaccine. Third, addressing neutral or negative attitudes towards COVID-19’s severity is crucial, as perceived risk strongly influences vaccine behaviour. Public health authorities should also continue to engage local leaders, despite lower trust levels in these groups, as they remain influential within the community. Involving religious and traditional leaders in advocacy campaigns may help bridge the trust gap, as has been successfully demonstrated in other health interventions in Nigeria, such as polio eradication efforts.

A major strength of this study is its design, use of random sampling and a large sample size. A pretested electronic questionnaire was used to minimise data collection errors, and data collectors were trained to minimise bias. This study is among the first of its kind in Zamfara state capturing a population that would not necessarily have access to internet-based surveys conducted in Nigeria. Also, the inclusion of both rural areas and urban slums, captures an underserved population that is often under-represented in vaccine studies. The inclusion of household heads ensures that the data reflect the perspectives of key decision-makers within the household.

However, the cross-sectional design limits the ability to infer causality. Additionally, self-reported data may be subjected to biases such as social desirability or recall bias. Another drawback, during data collection, occurred when mobile communication was lost between the researchers and data collectors as a result of the state-wide shutdown of all communication towers by the government for more than 3 months due to an increase in security threat and banditry activities in the state. This increased the length of time for the planned data collection period by 2 months and also made the researchers to exclude some planned communities that could not be reached or suspected to be unsafe for the data collectors to visit.

This study highlights key sociodemographic, attitudinal and trust-related factors that influence COVID-19 vaccine acceptance in Zamfara state, northern Nigeria. Efforts to improve vaccine uptake in this setting must prioritise building trust in health authorities, leveraging media for accurate information dissemination and addressing misconceptions about the disease. The findings contribute valuable insights to the global effort to improve vaccine uptake, particularly in rural and low-resource settings. Future research should consider longitudinal designs to better understand vaccine acceptance over time.