key: cord-0749970-yz0rw66x
authors: Orangi, S.; Pinchoff, J.; Mwanga, D.; Abuya, T.; Hamaluba, M.; Warimwe, G.; Austrian, K.; Barasa, E.
title: Assessing the level and determinants of COVID-19 Vaccine Confidence in Kenya
date: 2021-06-16
journal: nan
DOI: 10.1101/2021.06.11.21258775
sha: a546bf54b0970778972bd32db5b63a39736d9603
doc_id: 749970
cord_uid: yz0rw66x

The government of Kenya has launched a phased rollout of COVID-19 vaccination. A major barrier is vaccine hesitancy; the refusal or delay of accepting vaccination. This study evaluated the level and determinants of vaccine hesitancy in Kenya. We conducted a cross-sectional study administered through a phone-based survey in February 2021 in four counties of Kenya. Multivariate logistic regression was used to identify individual perceived risks and influences, context-specific factors, and vaccine-specific issues associated with COVID-19 vaccine hesitancy. COVID-19 vaccine hesitancy in Kenya was high: 60.1%. Factors associated with vaccine hesitancy included: older age, lower education level, perceived difficulty in adhering to government regulations on COVID-19 prevention, less adherence to wearing of face masks, not having ever been tested for COVID-19, no reported socio-economic loss as a result of COVID public-health restriction measures, and concerns regarding vaccine safety and effectiveness. There is a need for the prioritization of interventions to address vaccine hesitancy and improve vaccine confidence as part of the vaccine roll-out plan. These messaging and/or interventions should be holistic to include the value of other public health measures, be focused and targeted to specific groups, raise awareness on the risks of COVID-19 and effectively communicate the benefits and risks of vaccines.

perceptions of the pharmaceutical industries, concerns on vaccine safety and/or the source of the vaccine, lack of confidence in the government, and vaccine costs [11] [12] [13] [14] . Vaccine hesitancy is however context-specific and varies across time and place [15] . Kenya reports a high vaccine confidence against childhood diseases, with 89% reporting vaccines as safe, 87% reporting them as being effective, and 97% perceived importance for childhood vaccination [16] . However, there is limited evidence and understanding of the public willingness to accept; and the confidence they place on the COVID-19 vaccine in Kenya, which mainly targets the adult population. This study, therefore, aims to determine the level of COVID-19 vaccine hesitancy in Kenya and report on its determinants.

This study employed a cross-sectional study design to administer a knowledge, attitudes, and practices survey through a phone-based platform. The survey was conducted in February 2021.

The survey was administered to participants sampled from households in four existing Population Council prospective cohort studies across four counties; Nairobi, Wajir, Kilifi, and Kisumu [17] .

Specifically, in Nairobi County the target population was from five urban informal settlements: 2,565 households in Huruma and Kibera enrolled in the Adolescent Girls Initiative-Kenya (AGI-K) study [17, 18] and 4,519 households in Dandora, Kariobangi, and Mathare enrolled in the NISITU program study [17, 19] . In Wajir County, an arid region in north-eastern Kenya, the target population were households from 79 villages in Wajir East, Wajir West, and Wajir South sub-counties enrolled in the AGI-K study [17, 18] . In Kilifi County, a coastal region in Kenya, households from three sub-counties (Ganze, Kaloleni, and Magarini) enrolled in the Nia Project formed the target population [20] . Lastly in Kisumu County, located in Western Kenya, the target populations were households in Nyalenda area (one of the largest informal settlements in Kisumu) and Kolwa East (a peri-urban area) who were enrolled in the PEPFAR DREAMS study cohort [17] . This is illustrated in Table 1 . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

Households with available phone numbers were randomly sampled from the four existing cohorts using a ratio of 1:3 for male to female interviews. Due to the nature of the sampling frame described above, the randomly sampled participants were from households with at least one adolescent. Households that solely constituted adult residents or adults and very young children only were not eligible for inclusion in the initial cohorts and therefore not represented in this study.

The data collection tool was a knowledge, attitudes, and practices survey that collected information on 1) socio-demographic background information 2) the knowledge, attitudes, and practices reported by households concerning COVID-19 3) the barriers to adoption of non-pharmaceutical interventions for COVID-19 prevention 4) the social, economic, education and health effects of COVID-19 prevention measures on adults 5) the level and determinants of vaccine hesitancy. The latter being the focus of this study.

The data collection tool was designed using validated measures where possible such as the WHO SAGE vaccine hesitancy tool, existing COVID-19 vaccine hesitancy tools [21, 22] , and was also informed by local Kenyan researchers. Questions on the determinants of vaccine hesitancy were categorized into three broad groups adapted from the WHO SAGE vaccine hesitancy tool [21] . The first category reported on individual and group influences. These included questions on the individual's perceived risk of getting COVID, the ease of following government regulations, societal perception of having COVID-19, the individual adherence to wearing masks and having ever been tested for COVID, and The tools were in English and translated to Swahili, Dholuo and Somali, piloted, and administered by local interviewers. Data was collected using Open Data Kit.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint

Ethical approval for this study was obtained from both Population Council Institutional Review Board (p936) and AMREF Ethics and Scientific Review Committee (P803/2020). Before data collection, verbal informed consent was obtained from all participants 18 years and over. Participants were told they could terminate the survey at any time or refuse to answer specific questions. Participants were informed beforehand that they would be reimbursed 100 Kenyan shillings (~US$1) for their time (transferred via M-PESA mobile money).

The socio-demographic characteristics of the sample population were described by computing descriptive statistics. A cross-tabulation analysis was performed to determine the level of vaccine hesitancy among the respondents' sociodemographic characteristics using chi-squared tests.

Multivariate logistic regression analyses, accounting for the counties as clusters, were performed to compute the adjusted odds ratio (aOR) with a 95% confidence interval. Vaccine hesitancy was the dependent variable and was dichotomized as either accepting (i.e. very likely to get the vaccine) or hesitant (showing some level of hesitancy i.e. somewhat likely, somewhat unlikely, or very unlikely to get the vaccine). Socio-demographic characteristics, individual risks and perceptions, contextual factors, and vaccine-specific issues were included as predictor variables for vaccine hesitancy. The description of the dependent and independent variables is illustrated in supplementary table 1. Predictor variables were included in the multivariate model if found to be significant at a 0.05 significance level in the crude logistic regression. Multicollinearity of the variables was assessed using variance inflation factors and the Hosmer-Lemeshow goodness-of-fit test was used to ensure that the model adequately fit the data. The significance level was set at <0.05 and Stata version 15.0 (Stata Corporation, College Station, Texas, USA) was used for the data analyses.

The socio-demographic characteristics of the 4,136 respondents who participated in this study are shown in Table 2 . The mean age of the respondents was 40.8 years (SD 12.6) and the average household size of the respondents was 7.5 (SD 4.4). Most of the respondents were female (67.2%), residents of rural counties (56.0%), married (72.7%), had either no schooling or primary school level education) as the highest education level (70.7%), and were from the lowest wealth tertile (36.7%).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint Among the respondents, the amount participants were willing to pay for the vaccine (if it was not available for free) was reported at USD 3.23 (KES 323.06 (SD 1407.69)) and the overall reported hesitancy towards the COVID-19 vaccine was 60.6% (n=2,507). The overall level of vaccine hesitancy and across the counties is illustrated in figure 1.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint 

The multivariate logistic regression analysis for socio-demographic factors, individual influences, contextual factors, and vaccine-specific factors as predictors of vaccine hesitancy among the respondents is illustrated in is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 16, 2021. 

This study set out to evaluate the level and determinants of COVID-19 vaccine hesitancy in Kenya. We make several reflections from the findings. First, the overall level of COVID-19 vaccine hesitancy reported is high (60.61%), as compared to childhood vaccine acceptance in Kenya: Specifically, reported perceived childhood vaccine safety, effectiveness, and relative importance of common childhood vaccines is greater than 87% in Kenya [16] . This highlights the need for the Kenyan government to prioritize interventions to address vaccine hesitancy and improve vaccine confidence as part of its vaccine roll-out plan; examine the socio-demographic factors, individual effects and perceived risks, context, and vaccine-specific issues that affect vaccine hesitancy. This level of hesitancy is consistent with findings elsewhere. For instance, a survey done across 19 countries reported 71.5% of the respondents were very likely or somewhat likely to accept an available COVID-19 vaccine that is proven safe and effective, with differences in vaccine acceptance ranging from 90% in China to 55% in Russia [23] . In sub-Saharan Africa, surveys have reported 84.6% of Cameroonians, 52% of South Africans, and 50% of Zimbabweans to be hesitant or would reject the COVID-19 vaccine [13, 24, 25] .

Second, demographic and socio-economic factors that determine vaccine hesitancy include age, level of education, and socio-economic status. In contrast to some studies [22, [26] [27] [28] , we report a higher likelihood of vaccine hesitancy between those of a higher age group (36-57 years) as compared to younger people (18-35 years). However, our findings align with some settings such as Ireland, where those aged 35-44 years were more likely to be vaccine hesitant or resistant than accepting [27] .

Consistent with other studies a higher level of education was associated with less likelihood of vaccine hesitancy [26, 28] . Findings on socio-economic status are mixed, with individuals in the 2 nd tertile less likely to be hesitant and those in the 3 rd tertile being more hesitant, when compared to those in the 1 st tertile. However, with the exception of the rural pastoralist site, income was not associated with vaccine hesitancy. Other studies reported less vaccine hesitancy among higher income groups [13, 27] . These findings illustrate variations among certain sub-groups in the population and hence the need for focussed messaging and campaigns aimed at specific target groups that are more likely to be vaccine hesitant.

Third, attitudes and level of compliance with other NPI COVID-19 measures appear to be associated with hesitancy. For example, those who reported difficulty in adhering to government regulations regarding COVID-19 prevention, reported wearing face masks less often now than when the pandemic started were more likely to be vaccine hesitant. Other studies have also reported an increased likelihood of vaccine hesitancy amongst those who are less adherent to measures put in place to control COVID-19 [29] . Reduced adherence could also be linked with reduced trust in institutions or government.

Findings from a global survey found that those who reported having trust in their governments were . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint less likely to be vaccine hesitant [28] . This finding highlights the need for public health messaging of vaccination to be holistic and include information about the value of other public health measures. As vaccine rollout in Kenya may take some time, NPI preventive measures must continue to avoid a resurgence of infections.

Fourth, individuals perceived risk of COVID-19 and its impact on their lives and livelihood also determined their level of hesitancy. For instance, those who had tested for COVID-19 previously, and those who had experienced substantial socio-economic impacts of COVID-19 were less likely to be hesitant. This finding is similar to findings in Saudi Arabia and France, where individuals with an increased perceived risk of COVID-19 were less likely to be vaccine hesitant [22, 26] . This underlines the need to raise awareness about the consequences and risks of COVID-19 and to effectively communicate the value of vaccines.

Fifth, consistent with other studies, those who had concerns over COVID-19 vaccines' side-effects and effectiveness were more likely to be vaccine hesitant [26, [28] [29] [30] . This highlights the prevailing environment where there is heightened concern about the effectiveness and side effects of COVID-19 vaccines. Our findings emphasize the importance of a holistic, dynamic, transparent, and consistent public health messaging in improving vaccine hesitancy. Attention should be placed on building trust in the vaccine [31, 32] . Additionally, reassurance of the capabilities of the regulatory bodies in ensuring safety and effectiveness should be emphasized [28] . This should be accompanied by open access, real time safety data at a national and regional level, and risk-based assessments that inform decision making.

This study focused on one dimension of access, namely acceptability. However, we recognize that other dimensions of access such as the availability of COVID-19 vaccines ensured through procurement and supply-side factors, and the affordability of the vaccines need to be considered over and above vaccine hesitancy as we aim to achieve herd-immunity, and the control of the pandemic. Further, this study has several limitations. First, the sample was drawn from existing Population Council cohorts whose households all have adolescents. Therefore, the sample is not representative of the four counties included in this study and the results are not generalizable to the full population. Second, the study was cross-sectional and reflects the level and determinants of vaccine hesitancy, as of February 2021. This was before the actual COVID-19 vaccine rollout in Kenya that started in March 2021. Conducting a longitudinal study would have provided more information on the change in vaccine hesitancy and its drivers, which could also inform the tailoring of messages over time. Third, there is also a need for qualitative studies to further explore the drivers and deterrents of COVID-19 vaccine uptake and the factors that may improve or compound COVID-19 vaccine acceptance. Despite these reported weaknesses, the study provides important insights on the COVID-19 vaccine hesitancy in certain . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint locations of Kenya and provides implications to policymakers on possible avenues of improving vaccine hesitancy in Kenya.

Our findings highlight that almost one year into the pandemic and about a month before vaccine rollout had begun in Kenya, vaccine hesitancy was quite high. As the current vaccine rollout plan will take time, there is also a need to promote holistic public health messaging to ensure that NPI interventions such as face masks, hand washing, physical distancing, and hand sanitizer use continue. These behaviours are tied to vaccine hesitancy and confidence, clearly linking the two and the need for cohesive messaging campaigns. We find significant variation by socio-demographics and perceived risk of COVID-19 infection and economic impacts; tailored messages may be required to reach those with different concerns, levels of education, and other factors. Lastly, there is a critical need for accurate and transparent information from trusted sources to combat misinformation particularly around vaccine side effects and effectiveness.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 16, 2021. ; https://doi.org/10.1101/2021.06.11.21258775 doi: medRxiv preprint

World Health Organization WHO Coronavirus Disease (COVID-19) Dashboard With Vaccination Data Available online

Health and socio-economic impacts of physical distancing for COVID-19 in Africa

Assessing the Indirect Health Effects of the COVID-19 Pandemic in Kenya

World Health Organization A Coordinated Global Research RoadMap

World Health Organization Draft landscape and tracker of COVID-19 candidate vaccines Available online

What will it take to approve them for the rest of the world? Available online

The ins and outs of Kenya's COVID-19 vaccine rollout plan Available online

Imperial College COVID-19 Response Team

COVID-19 vaccine hesitancy in a representative working-age population in France: a survey experiment based on vaccine characteristics

Dealing with vaccine hesitancy in Africa: The prospective COVID-19 vaccine context

Addressing challenges to rolling out COVID-19 vaccines in African countries

Assessment of Vaccine Hesitancy to a Covid-19 Vaccine in Cameroonian Adults and Its Global Implication. Vaccines

Covid-19 vaccine acceptance in the Democratic Republic of Congo: A cross-sectional survey

Gallup Wellcome Global Monitor -First Wave Findings

Experiences among adults and adolescents during COVID-19 pandemic from four locations across Kenya-Study description

The Adolescent Girls Initiative-Kenya (AGI-K): Study protocol

Evaluation of the NISITU Program: a study to determine the effect of a gender attitudes and gender based violence program for adolescents in

Cluster randomized evaluation of the Nia Project: Study protocol

Measuring vaccine hesitancy: The development of a survey tool

Determinants of COVID-19 Vaccine Acceptance in Saudi Arabia: A Web-Based National Survey

El-Mohandes, A. A global survey of potential acceptance of a COVID-19 vaccine

COVID-19 Vaccine Roll-Out in South Africa and Zimbabwe: Urgent Need to Address Community Preparedness, Fears and Hesitancy

IOL 52% of South Africans don't want Covid vaccine, despite SA securing initial batch Available online

COVID-19 vaccine hesitancy in a representative working-age population in France: a survey experiment based on vaccine characteristics

Psychological characteristics associated with COVID-19 vaccine hesitancy and resistance in Ireland and the United Kingdom

El-Mohandes, A. A global survey of potential acceptance of a COVID-19 vaccine

COVID-19 vaccine hesitancy and resistance: Correlates in a nationally representative longitudinal survey of the Australian population

COVID-19 Vaccine Hesitancy among Young Adults in Saudi Arabia: A Cross-Sectional Web-Based Study

World Health Organization Behavioural considerations for acceptance and uptake of COVID-19 vaccines

A Rapid Systematic Review of Public Responses to Health Messages Encouraging Vaccination against Infectious Diseases in a Pandemic or Epidemic