Background
Cholera results from ingesting pathogenic strains of the bacterium
Vibrio cholerae in contaminated water or food [
1]. Although cholera should not be fatal, if untreated, case-fatality rates (CFR) for severe cholera may be as high as 50 % [
2]. An estimated 1.4 billion people are at risk for cholera in endemic countries [
3]. Controlling cholera remains a significant challenge in Sub-Saharan Africa. Access to safe water and sanitation remain low in the region, about 61 and 30 %, respectively [
4]. Needed development requires major investments in infrastructure that proceed very slowly.
In the interim, oral cholera vaccines (OCVs) can save lives in epidemics and in endemic areas. The World Health Organization (WHO) recommends OCVs as a short-term control strategy for high-risk populations to complement long-term water and sanitation improvements [
5]. Two safe OCVs—Shanchol™, with a protective efficacy of 66 % [
6], and Dukoral®, with 79 % direct protection [
7]—are currently available for international use. Efficacy is not enough, however, for vaccines to be effective. People must also be willing to accept them. Local social and cultural ideas about illness, vaccines and community preferences are critical considerations. Past programme experience provides valuable lessons that underscore the priority of social and cultural aspects of vaccine acceptance and effective vaccine action [
8‐
11]. A recent review of vaccine hesitancy suggests community effectiveness may depend on particular features of setting, health problem and vaccine [
12].
Making the benefits of immunization, including new and underutilized vaccines, available to all regardless of where they are born, who they are, or where they live is a vision of the Decade of Vaccines (2011–2020) [
13]. In 2012, the World Health Assembly approved the Global Vaccine Action Plan as a framework to achieve this vision and a strategic objective of the plan emphasised the importance of understanding community demand and trust in vaccines. The decision made by Gavi, the Vaccine Alliance (Gavi) to contribute to a global stockpile for OCVs during 2014–2018 reflects increasing priority for use of cholera vaccines in endemic settings [
14]. However, not enough is known about community acceptance of OCVs, especially across populations, and this information is critical for effective vaccine implementation. Furthermore, although some sociocultural features may have consistent effects on across settings, others are specific to particular local settings. Systematic comparison of local studies clarifies consistent and distinctive effects of sociocultural factors on vaccine acceptance that may not be apparent from findings of the local studies. This interest motivated the comparative analysis reported here.
This report presents findings from comparison of sociocultural determinants of anticipated OCV acceptance across the three settings in Sub-Saharan Africa: Southeastern Democratic Republic of Congo (SE-DRC), Western Kenya (W-Kenya) and Zanzibar. Studies of sociocultural aspects of cholera and determinants of anticipated OCV acceptance were undertaken in three cholera-endemic settings in Africa [
15‐
17]. A comparison of sociocultural features of cholera illness experience and meaning have been reported by Schaetti and colleagues [
18]. Common and distinctive sociocultural determinants of anticipated OCV acceptance that may affect uptake and effectiveness of OCVs in cholera-endemic areas of three countries in Sub-Saharan Africa are presented in this cross-setting analysis. Knowledge of comparative community interest and setting-specific determinants of OCV acceptance, gleaned from cross-cultural study, can guide policy for effective implementation of OCV. With the development of a global stockpile, findings to help guide use of OCVs are likely to be timely.
Discussion
Experience with OCV in vaccination campaigns has been steadily increasing [
30]. To the best of our knowledge, this analysis is the first review of common and distinctive sociocultural determinants of anticipated OCV acceptance across multiple settings in Africa. Comparable research methods enabled a systematic meta-analytic approach. The findings identified patterns that would be unapparent in the individual studies, such as the identification of relevant determinants in all three populations. For example, the finding that knowledge of ORS for home-treatment of cholera was negatively associated with OCV acceptance at the high price was a unique finding from this meta-analysis, and it was not apparent (significant) from any of the individual country-specific studies. Our analysis is based on a systematic comparison of the three data sets, rather than a simple comparison of summary findings reported in the three published papers. The quantitative associations, derived and presented through forest plots, show how priority symptoms, perceived causes and options for help-seeking may influence OCV acceptance positively or negatively across different populations. Some factors have common effects across populations and others are setting-specific, indicating the value of local study to enable locally effective vaccine action. Although our methods are not a traditional meta-analysis, use of meta-analytic techniques highlight key sociocultural determinants common to three African settings and the importance of studying them.
Although anticipated acceptance may not perfectly reflect actual acceptance, observed priority for OCVs indicate that these communities desire benefits from such vaccination initiatives. The finding that fewer determinants of anticipated acceptance were identified for the high priced vaccine (two), compared with the medium price (six), clearly shows that increased cost imposes an economic barrier making other features of acceptance and demand irrelevant.
Paradoxically, SE-DRC has the greatest number willing to purchase the medium-price and high-price OCVs. People in W-Kenya and Zanzibar are economically better off as seen from gross domestic product per capita [
31] and self-reported reliability of income among study respondents [
18]. The seeming contradiction of greatest willingness to purchase OCVs among those with least economic resources may be explained by the serious trouble caused by cholera in SE-DRC, where public health facilities are often inaccessible or non-functional. Another point worth noting is that vaccines in SE-DRC are usually provided for free. The ability to pay is often overestimated when the scenario is hypothetical and respondents do not have to actually make the payment from their own pockets [
32,
33]. The finding indicates community priority for a desired vaccine, rather than capacity to pay or prospects for effective uptake at the high price. Zanzibar has the lowest anticipated OCV acceptance at all prices. This may be an unintended consequence of a more accessible and effective public health system there compared to the other settings. Cholera camps instituted during an outbreak are accessible to most of the population who anticipate a fairly rapid response from local authorities [
16]. Hence, the priority to pay for a vaccine may be reduced when timely life-saving treatment is assumed to be readily available compared with SE-DRC, where such confidence in lacking.
Findings suggest that when vaccine price is high, motivation to purchase OCV appears low among those with knowledge of feasible treatment options such as ORS. Vaccination and ORS seem to be competing interventions in the public mind. Zwisler et al. [
34] found substantial satisfaction with ORS in treating diarrhoea among caregivers in Kenya and likely re-use of ORS in treatment if it had ever been used before. The marginal value of an OCV that users consider costly may be more limited in areas where ORS is well-known and widely used. Furthermore, priority for treatment may be valued more highly than prevention.
All study respondents were adults, and anticipated OCV acceptance was higher among younger adults. Lack of education in our study was associated with OCV non-acceptance. Other studies report a significant positive association between education and cholera-related knowledge [
35]. Youth and better educated community residents may be a resource for vaccination campaigns to mobilize for community awareness of the benefits of vaccines.
Household size imposes economic constraints; more mouths to feed leaves less money available for other expenses, even if desired, including vaccines. OCVs are especially important for larger households which are more likely to be crowded and burdened by limited sanitation. Sharing a latrine with many households is a reported risk factor for cholera in Kenya [
36]. Economic limitations affecting the most vulnerable segments of the community with the least resources highlights the priority of making OCVs available without cost to users. If provided at a low price, incentives or discounts for larger families may increase vaccine uptake.
Contrary to expectations, knowledge of dehydration symptoms decreased the priority of OCVs for prevention. Symptoms of dehydration, which are clearly related to cholera for health professionals, do not seem to be core features of a vaccine-preventable formulation of cholera in the community. Symptoms of dehydration may be linked in local perceptions to other forms of diarrhoea making a “cholera” vaccine less relevant. Although most respondents in the three studies identified the illness of the vignette as cholera or its local language equivalent (>85 %) [
15,
17,
27], its link to dehydration appears less well understood.
In SE-DRC, social and economic vulnerability are interrelated, and both may constrain access to vaccines for those who may need it most. In Zanzibar, cholera-related stigma appears to motivate OCV acceptance, presumably to avoid stigma. However, vaccine acceptance was impeded by local magico-religious ideas, possibly reflecting a conflict between public health and interests of local healers.
In Zanzibar, religious influences appear less enabling for OCV acceptance. Although no active resistance from religious leaders is foreseen in Zanzibar, engaging religious leaders for vaccine action in all settings is important to build alliances and pre-empt opposition that may affect uptake [
37], as indicated by notable opposition to polio vaccines in Nigeria [
9,
38]. On the other hand, prayer and religious influences of a predominantly Christian population in SE-DRC and W-Kenya may promote vaccine use.
In W-Kenya, lower anticipated acceptance at the rural compared to the urban site, may result from urban–rural income disparities in W-Kenya, which was the only setting where fewer rural than urban respondents reported reliable and dependable incomes (
p < 0.001) [
15]. Access and uptake would appear to be more sensitive there to the effect of price. In Zanzibar, however, rural respondents were more likely to accept the high-price OCV. These findings suggest that urban–rural differences in vaccine acceptance may vary across settings based on local conditions and priorities.
Limitations and strengths
Data for this analysis were collected between 2008 and 2010 during a period of high cholera burden in all three settings. More recent WHO data indicate a persisting cholera burden in DRC (33,661 cases). In United Republic of Tanzania however, fewer cases (286) were reported, and in Kenya no cases of cholera were reported in 2012 [
39]. The decline in cholera cases in Kenya was attributed to effectiveness of water, sanitation and hygiene (WASH) interventions by public health officials (Personal communication, Public Health Officials in Kisumu and Siaya. Conversations during a dissemination activity conducted by the research team at study sites in Western Kenya, 2013). More recently, however, rapid spread of cholera has been noted in Kenya with 3301 cases reported between December 2014 and May 2015 [
40]. High CFR of 2 % have been noted country-wide with some counties reporting CFR as high as 7.6 %. Zanzibar appears to have benefitted from the OCV campaign that was undertaken there. Thus, priority for use of OCVs changes with the change in cholera epidemiology. While OCVs are not indicated in settings with no more cholera, findings of this study and community priority for preventing cholera remain relevant not only for consideration in future outbreaks but also for implementation of other WASH interventions. Findings and the approach presented in this comparative analysis are relevant for settings where there is a clear rationale for use of OCVs.
Although detailed narrative data were collected during interviews that lasted approximately one hour, sample size was a limitation in the quantitative analysis. Due to practical constraints, between 356 and 379 interviews could be conducted in each setting. The sample size allowed for limited adjustment and all explanatory variables could not be adjusted against each other.
In addition to the benefits of increased power through meta-analytic techniques that enable identification of determinants that may not be apparent in individual studies, a major strength of this cross-setting analysis is the identification of sociocultural determinants of OCV acceptance that are generalizable across multiple populations. Local study is necessary to understand nuances of vaccine acceptance or hesitancy that are influenced by local culture and context. However, questions of generalizability of findings to other settings often arise. By considering sociocultural determinants of anticipated OCV acceptance from local study, but common across three distinct populations, broader and more generalizable determinants have been distilled that are useful in guiding policy for wider use and implementation of OCVs by national and global policy makers and public health professionals.
Recent developments towards setting up of a global stockpile for OCVs is based on the assumption that it would be used by the populations it is given to. The approach presented in this study, which makes it possible to distinguish common and setting-specific sociocultural factors affecting OCV acceptance is especially timely in view of opportunities arising for effective use of OCVs enabled by the development of a global stockpile. It is also relevant for enhancing coverage of other vaccines through consideration of community determinants.
The community perspective is relevant not only for OCVs but also for consideration of community-related determinants of vaccine effectiveness, such as hesitancy, demand and access. A rapid assessment of such community interests can be expected to contribute to the effectiveness of vaccine action. Based on experience with this approach for community assessment, development and validation of rapid assessment tools are needed to demonstrate the usefulness of the approach for enhancing uptake in programme settings.
Competing interests
The authors have no conflicts of interest to declare.
Authors’ contributions
Conception and design of study: NS, CS1, SM, CS2 and MGW. Analysis of data: NS, CS1 and CS2. Writing of manuscript: NS, CS1, SM and MGW. Critical review and revision of manuscript: all authors. All authors have approved the final manuscript.