Introduction

With over 28000 cases and 11000 deaths, the 2014–2016 Ebola epidemic in West Africa was declared by the World Health Organization (WHO) as a public health emergency of international concern and a threat to global security and international peace [1]. Liberia was the most affected, recording 37% and 42% of total cases and deaths, respectively [2]. Some reasons posited for the outbreak of Ebola in Liberia have been associated with the decimation of the healthcare system and underdevelopment resulting from the country’s 14-year civil war [3–5], mistrust of government and state institutions [6–8], and the slow response from the WHO and the international community [9–11]. Before the Ebola outbreak in Liberia, the country’s healthcare system was still recovering from the 14-year civil war and was, therefore, inadequate to handle the transmission and death rates. Fewer and ill-equipped health infrastructure and inadequate medical personnel meant that public health interventions to contain the Ebola epidemic rested on collaborative efforts of the community, government, and international organizations [4,12].

International response strategies by the WHO, the United States Centers for Disease Control and Prevention (CDC), and other international community members provided financial and technical support to the Liberian government. These included devising and implementing multiple response strategies, such as rigorously implementing the WHO-recommended non-pharmaceutical interventions (e.g., case isolation, quarantine, encouraging sanitary practices of washing hands and burial, and social distancing) [13]. Non-pharmaceutical interventions have proven effective in controlling disease spread, as showcased during Ebola and the recent COVID-19 pandemic [14].

In addition to the international response, local community members, including women’s groups, were particularly influential during the all-hands-on-deck Ebola response systems. Community leaders leveraged their influence, social networks, and community mobilization to implement community-level Ebola containment strategies that have been credited with helping to end Ebola in Liberia. In particular, women’s groups have led and engaged in Ebola containment practices by devising locally made equipment and using it to manage Ebola victims in families and communities [15]. Women’s groups served as contact tracers and case finders in their communities. They engaged in Ebola education and sensitisation programs through house-to-house campaigns and radio/town crier services to raise awareness of Ebola safe practices [15]. Indeed, Sharon Abramowitz and colleagues [16] have discussed the gendered labour distribution of Ebola prevention in communities, highlighting how women’s groups supported community mobilisation efforts by screening for Ebola symptoms and enforcing Ebola quarantine measures. Perry [1] also described how a female chief mobilised her community to build a 3-bedroom structure to care for Ebola victims. Further, Leymah Gbowee and Ruth Caesar, who have both played critical roles in forming women’s groups in Liberia, documented the various caregiving roles women’s groups played as first responders during the Ebola epidemic in Liberia [17,18].

Despite the significant contribution of women’s mobilisation efforts to ending Ebola, much has not been discussed on how community members perceived their presence and roles in helping manage the Ebola crisis in the country. Additionally, the current literature seems to credit community male leaders and external partners [19,20] without fully exploring how residents linked the presence of women’s groups to Ebola mitigation efforts in their respective communities. To contribute to bridging this knowledge gap, we pose the following research question: How did the presence of women’s groups influence perceptions of Ebola response strategies in Ebola-affected communities in Liberia? We hope that the findings from this study will contribute to Liberia’s pandemic preparedness efforts and the broader discussion of achieving SDG 3.3, which aims to end infectious diseases in Liberia and globally.

Materials and methods

Measures

Our outcome variable, ‘Community response strategies to Ebola,’ was derived from the question, ‘How did your community respond to the 2014 Ebola outbreak? Participants could choose one or multiple response strategies as needed, including a) my community did nothing, b) Ebola education/awareness, c) community surveillance, d) lobby for Ebola prevention supplies, and e) Other, in which case the participant specified. We constructed the ‘community response’ variable and created three categories (i.e., no response, single response, and multiple response see Table 1 for percentage distribution). Single responses category indicates participants choosing only one of the above options (e.g., education, community surveillance, lobbying, or other), while those that selected two or more options (e.g., education and community surveillance or education, surveillance, lobbying) were categorized as multiple responses. “My community did nothing” was categorized as no response. We coded responses as follows: no response = 0, single response = 1, and multiple responses = 2.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Sample characteristics.

https://doi.org/10.1371/journal.pone.0325496.t001

The key independent variable, ‘women’s groups,’ came from the question, ‘Do you have women’s groups in this community?’ Responses were coded as follows: no, 0, and yes, 1. We added other theoretically relevant control variables specific to the study context, including demographic, socioeconomic, and place-based variables. The demographic variables included in this study were gender, age, marital status, and religion. Education, occupation, and household wealth index constituted the socioeconomic variables. Following the criteria used by the Demographic and Health Survey (DHS) conducted in Liberia and other African contexts [42–44], we collected variables on the economic conditions of participants, such as household assets like radio and television, electricity, access to running water, ownership of livestock, and many others. Using principal component analysis, we constructed a composite variable from these responses and recategorized it into five wealth categories (i.e., poorest, poor, middle, richer, richest). Reporting household income this way provides a fairly accurate representation of living conditions. Finally, location and place-based variables included rural/urban status, county of residence, and the presence of a health facility. All variables with a value of 1 in Tables 2 and 3 serve as the reference category against which we tested our outcome variable in the analyses. Table 1 contains the sample distribution and characteristics of the control variables.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Bivariate multinomial logit models estimating women’s groups’ influence on Ebola epidemic response in Liberia.

https://doi.org/10.1371/journal.pone.0325496.t002

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Multivariate multinomial logit regression estimating women’s groups’ influence on the Ebola epidemic response in Liberia.

https://doi.org/10.1371/journal.pone.0325496.t003

Results

Discussion

This paper explored the perceived influence of women’s groups’ mobilisation initiatives in response to the Ebola outbreak in Liberia. Overall, we found that knowledge of the presence of women’s groups in communities affected people’s perception of reporting an Ebola response. Community response to Ebola was either single (i.e., one intervention only) and/or multiple (i.e., two or more interventions). Participants in communities with women’s groups reported higher odds of observing both single and multiple responses. This finding aligns with calls for integrating health equity and community engagement in public health interventions during epidemics and pandemics, especially in resource-constrained post-conflict contexts [48,49]. Two possible explanations can be provided for this observed relationship. The first is the widespread recognition of Liberian women as peacebuilders and gender equality advocates locally and internationally, resulting from their decades-old use of maternal and social activism [50,51], and the second is the trust-based relationships women’s groups have built over the years with communities [52,53]. We believe that because of the social and cultural visibility of women’s groups, their presence may have brought the spotlight to those communities, affecting the level of support communities received from national and international Ebola response teams. A recent study by Anfaara et al. [15] reported how women’s groups formed partnerships with national and international organisations (e.g., UN Women) to implement Ebola interventions in affected communities. Women’s groups reported receiving funds from these organisations to purchase and distribute Ebola supplies during the Ebola outbreak, including hand sanitisers and handwashing equipment. While others reported receiving technical training as contact tracers to help identify Ebola patients in their communities. These activities might have increased their visibility in communities, thereby influencing how community members perceived their role in Ebola strategies. Additionally, it is possible that women’s groups’ contribution to peace and development in Liberia gave them an added advantage as trustworthy community partners, contributing to adherence and compliance with the Ebola control measures. Research on community engagement, trust, and compliance with public health measures during disease outbreaks and pandemics has been widely documented [3,6,7,13].

Demographic characteristics, such as gender, age, marital status, and religion, affected how participants perceived Ebola response strategies in their communities. Men reporting lower odds for multiple responses corroborate with the literature about the gendered involvement and impacts of the Ebola outbreak in Liberia. Several studies have documented the burden of care work on women and how women’s caregiving duties made them susceptible to Ebola infection [10,16,54], as well as the social and economic impacts women faced at the peak of the epidemic [55]. We believe that the gendered implications of Ebola care work and social burdens on women may have led to targeted interventions for women as a group compared to men. For instance, targeted interventions aimed at increasing antenatal access and utilization for pregnant women during the Ebola outbreak have been reported [56,57].

Furthermore, people in the 29–48 age groups also reported lower odds of witnessing multiple responses. We explain this finding within the contextual dynamic of age-related roles and responsibilities of Liberians. Like elsewhere in Africa, Liberian adults within the 29–48 age categories are mostly the main providers and caretakers of their nuclear and extended families, who may have borne the brunt of providing for their families during Ebola [58]. While the working-age category ranges between 15 and 65 years in most developed countries, the situation in Liberia is quite different. The country is relatively young, with about 89% dependency ratio relative to the working population [59]. With a high dependency ratio, pressure to financially support dependent family members may have been placed on the 29–48 age cohorts who are within the active working population. We argue that taking care of bread-and-butter issues and Ebola-related medical costs might have been the primary goal of these adults during Ebola; evidence suggests that disruptions in economic activities heightened the suffering and poverty levels of the working few [60,61]. It is possible that people within the 29–48 age groups were preoccupied with providing for their families; as such, any intervention that was not financial may not have been perceived as an Ebola response for these people. For instance, Pellecchia et al. [62] have documented how the Liberian government and its international partners initiated an economic support package for poor households. This partially explains why people 49 and older may have indicated witnessing a single response.

Similarly, people who were formerly married reported lower odds for multiple responses, while those who identified with traditional religion reported lower odds for single and multiple responses. Health interventions, including outbreak response in SSA contexts and Liberia, tend to favour those in marriage unions. Legal unions, such as marriage or common law, are often considered the primary unit for deploying health interventions due to the likelihood of influencing positive behaviour change or compliance with a health intervention [32,63]. However, the unintended consequence is the potential for unmarried or formerly married individuals to be overlooked. During the Ebola outbreak in Liberia, for instance, while interventions focused on reducing Ebola spread through safer household dynamics and the impact of Ebola on increasing sexual and gender-based violence cases among married people [55], interventions targeting formerly married people were limited to social assistance, including food coupons and financial assistance, especially in rural areas [64,65]. This may have resulted in formerly married people reporting lower odds for multiple responses.

Regarding religious differences, it is documented that closing religious centres and banning some traditional practices was effective in controlling the spread of Ebola in affected communities [66]. Research shows that traditional beliefs and misconceptions about Ebola and unsafe burial practices greatly affected the spread of Ebola in Liberia [67]. Consequently, rigorous education and awareness campaigns on disease etiology and prevention, safe burial practices, and social distancing targeted religious groups, including those of traditional religion who predominantly congregate in specific geographic areas. Therefore, participants reported single or multiple responses depending on the religious activities and location, as highlighted by our findings.

Socioeconomic factors, including education, household wealth, and occupation, affected perceptions about Ebola responses. Except for education, people in all wealth categories, petty traders, and farmers reported lower odds for an Ebola response. As alluded to, the economic impact of Ebola exacerbated the living conditions of the poorest socioeconomic groups, most of whom are either employed in the informal sector as petty traders or in the agricultural sector as farmers [55]. The relationship between household wealth and occupation explains why the economic impact of Ebola in Liberia was pronounced. With most of Liberia’s economy relying on the informal sector, Ebola prevention strategies such as quarantine and social distancing affected markets where the majority of workers do business, thereby exacerbating the poverty situation in the country. We posit that interventions targeting this demographic may not have been sufficient, which may be reflected in their response. The education variable was significant for people with primary and high school education, who reported higher odds for both single and multiple Ebola responses. While this finding may seem counterintuitive and require further investigation, a possible explanation for this relationship may involve how public health interventions work in Africa. Most health interventions target people with lower educational status because of their perceived increased risk of non-compliance with safety measures [43,68]; consequently, Ebola response interventions may have specifically targeted these groups, resulting in how they perceive the response interventions.

Our location and place-based variables yielded some important findings that support the notion that geography tends to influence perceptions of health interventions. Communities with a health facility and people living in urban areas reported higher odds for single and multiple responses, respectively. In the context of the outbreak in Liberia, most health facilities were overcrowded as healthcare professionals prioritised attending to Ebola victims [69]. People who had access to health facilities or lived close to them would experience multiple Ebola responses, either through their role as healthcare recipients or targets of Ebola education and prevention strategies. Residents may have perceived the activities in and around healthcare facilities as response strategies. Likewise, due to the populous urban settlements in Liberia, urban dwellers may have received much attention regarding the Ebola response, owing to the urban bias in the delivery of healthcare services [70]. For instance, West Point, the biggest urban slum in Monrovia, was a target for multiple Ebola responses sometimes resulting in clashes between the community and military personnel who were there to enforce social distancing as an Ebola response measure [71].

Following the observed rural-urban dynamic, there were county-level distinctions in the Ebola response interventions. Except for Bomi and Margibi counties, where participants reported higher odds for a single response, the other counties reported lower odds for multiple responses. John Stanturf et al. [72] study on social vulnerability and Ebola infection found that counties in the northwestern part of Liberia, including Bomi, Margibi, Lofa, and Montserrado, recorded the highest risk vulnerability to Ebola infection, probably because of population-related dynamics. This vulnerability led to county-level Ebola intervention strategies such as the establishment of community care centers for quarantine services, Ebola education campaigns in and around hospitals, and active Ebola monitoring and prevention initiatives in Bomi and Margibi counties [69,73]. Similar initiatives were undertaken in Lofa and Montserrado counties; however, the Ebola disturbances or riots associated with implementing Ebola control measures might have affected people’s perceived response activities in Montserrado [16]. For people in Montserrado County, military and police surveillance may not have constituted an Ebola response, reflected in participants reporting lower odds for single and multiple responses.

While our study contributes to the literature on the importance of community mobilisation and engagement during disease outbreaks, our analyses and discussion may have contributed to some unintended assumptions, which we note. First, our study is limited to five of the 15 counties in Liberia, so interpretation and generalizability should be limited to these contexts. Second, while we ensured that data collection was of the utmost quality through the quality management strategies we implemented, participant responses may be affected by recall and social desirability bias. Data for this study were collected 8 years after the first Ebola case was detected and 6 years after Ebola ended in Liberia. It is possible that participants may have challenges recalling Ebola response strategies. Third, while it is essential to bring visibility to women’s social reproductive work, especially in a context where patriarchal structures and institutional barriers continue to discriminate against women in social and political forums, we acknowledge that Liberian women face multiple interconnected and structural challenges; as such our analysis and discussion may risk the valorisation and essentialisation of women’s social reproductive and caregiving functions. Traditional gender roles continue to mar women’s gender equality aspirations, with many lagging behind in political representation, formal education, and economic empowerment [74]. Gender-based violence (GBV) is on the rise despite the passage of the 2005 Rape Act and the 2019 Domestic Violence Act [75]. We also acknowledge the work of feminist political economy scholars who have cautioned us about the ‘double burden’ of care work and the feminisation of poverty that comes with valorising women’s care work, particularly in how the ‘caring economy’ benefits capitalist states [29,76,77]. That said, Liberian women’s contribution to health and well-being and the country’s development cannot be refuted. While they are mainly presented as victims of war, there is a changing tide where women are beginning to write their stories by showcasing work that positions them as agents rather than helpless victims [78]. These are the perspectives that inform the presentation of our analysis and discussion.

Conclusion and policy recommendations

This study underscores the contribution of women’s groups’ presence and mobilising activities to influence Ebola response strategies in Liberia. While local, national, and international partnerships and collaborations significantly contributed to containing Ebola in Liberia, community involvement and grassroots mobilisations have been crucial to this success. The community mobilisation efforts resulting from their trust-based relationships and global visibility benefited the communities in which women’s groups worked. As discussed elsewhere [15], the Government of Liberia and its international partners need to sustain existing community-based partnerships and collaborations while allowing space for new partnerships to develop. Activities of these local partners, including women’s groups, can be streamlined and better resourced to support Liberia’s public health infrastructure. Women’s groups can be trained and deployed to at-risk communities as community health liaisons, which can help with early detection for future pandemic preparedness. The Ebola epidemic in Liberia exacerbated the social and economic vulnerabilities of most people, affecting how they perceived response strategies. The social and economic impacts of Ebola will linger on; therefore, there is a need for a sustained social assistance program to support women and struggling families in their recovery.

Acknowledgments

We want to thank the survey participants for volunteering their time to answer our questions during the COVID-19 pandemic. We also extend our gratitude to Mr. Dackemue Dolo and all 12 survey enumerators who made the research possible. Special thanks to Dr Jason Were for his assistance in cleaning the raw survey data. Finally, we sincerely thank the three reviewers for their thoughtful, constructive feedback.