Data collection.

Primary data collection occurred during April and May 2016, over two iterative phases, by Guinean research assistants trained in qualitative interviewing skills. All participants for semi-structured interviews were purposefully selected through a criterion-based sampling strategy [22]. This type of strategy is used in qualitative research as a way to deliberately select particular settings, persons, or activities in order to provide information that cannot be obtained readily otherwise [21]. Data were collected until data saturation, or repetition of information, was reached among key themes [23]. Phases of data collection are described below.

Phase 1. Understanding policy and management perceptions.

In-depth interviews were conducted first among ‘key informants’ using a semi-structured interview guide “S1 File” covering 9 domains (Table 1). We interviewed in-country stakeholders from the primary organizations that were involved in addressing nutrition as part of the Ebola response, including but not limited to Government officials, United Nations bodies, non-Government organizations (NGOs), and Medical Facilities Administration. We collected data from both decision makers and managers.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Semi-structured interview guide content.

https://doi.org/10.1371/journal.pone.0202468.t001

Phase 2. Exploring community and front-line health worker perceptions.

Community members were then interviewed as ‘informants’ using a second semi-structured guide “S2 File” with 6 domains.These individuals included community leaders, EVD survivors, family members of EVD victims, midwives, and health workers who were at the frontlines of the response.

All interviews were digitally recorded in French or a local language, each lasting approximately between 30–60 minutes, in convenient, private locations chosen by participants. Field notes were taken immediately following interviews by interviewers for reflection on key questions and commentary on transcript quality or sampling considerations [19].

Phase 3. Synthesizing findings vis-à-vis relevant indicators of secondary data.

Secondary data sources with nutrition-related data from the Ebola response were collected and reviewed for relevant information to triangulate qualitative data collected in phases 1 and 2. Secondary data originated from sources provided by UNICEF in the form of unpublished reports, as well as a web-based literature search using keywords around Ebola and nutrition.

Implications of unpreparedness.

First, key informants highlighted the initial inter-agency coordination challenges, which reportedly slowed down the overall response and stemmed from having “no response plan.” Second, key informants acknowledged Guinea’s resource-constrained health system with weak epidemiological surveillance system and administration infrastructure to respond to this type of disease outbreak. Key informants emphasized the importance of the international community in supporting Government through resource mobilization indicating that response quality was largely “an issue of means”.

Third, while key informants praised the importance of improved communications over time, they also explained that initially, there were ineffective communications at multiple levels. At the national level, key informants discussed the delicate nature of declaring a national outbreak, citing potential consequences resulting from decreased foreign investment and potential disruption of social calm, affecting the overall well-being of the state. At the organizational level, data indicate there was weak coordination leadership among actors that typically work within the same health system but not often in tight collaboration with one another. And at the community level, initial messages had unintended consequences. For instance, some communications intended to promote care-seeking behavior had the opposite effect based on the framing of the messaging. Community members were reportedly unclear why they were told to go to the hospital when Ebola was thought to be incurable.

“So, at the start, it was not easy, even at the communication level. From the moment we started saying that it (Ebola) was not a disease we could cure, it created a psychotic atmosphere among the population. So then, people said, ‘Since it’s a disease that cannot be cured, why do you want us to go to the hospital?’”

                                        -Key informant, Nutrition Staff, Conakry

Similarly, there was confusion around the response slogan, “End Ebola in 60 Days,” which community members reportedly perceived to mean that the biomedical community had a cure but just were not providing it to them at that time.

Social attitudes impacting all levels of Ebola response.

Thus, ineffective communications at multiple levels reportedly contributed to heightened feelings of confusion, stigma, and fears underlying Ebola. They also exacerbated existing levels of distrust and skepticism among the local community toward the biomedical community, contributing to reduced positive care-seeking practices and greater reliance on traditional practices.

While these challenges were not explained by community members in the context of nutrition specifically, recurring themes indicated that the Ebola challenges were in fact linked to nutrition at multiple levels of influence:

“Everyone was disturbed, at work…everyone no longer had that courage, everyone was paralyzed…the farmer could not properly do his farming, the fisherman, as well, could not go to fish…even government workers (couldn’t work). Because the epidemic was here, there was no more trust between the people, for example…when someone felt sick it was difficult for him to go to the hospital…they felt like if they go to the hospital and it happens to be this disease (Ebola), then it’s over for him…even between husbands and wives there was this suspicion. When the epidemic began to disappear, people started their old behaviors, started to go and work, some started farming, but the impact was there because people could not afford things. The whole year of 2015, the impact was there… before Ebola, the number of children that had malnutrition… (compared to) after Ebola, (well) the number tripled up…it was severe malnutrition.”

                                        Informant, Community Member, Nzérékoré

The community member above summarized several effects of Ebola at basic, underlying, and immediate levels of nutrition. Despite this finding, key informants explained that there was a low importance placed on nutrition during the outbreak among both health professionals and community members.

Household food insecurity stemming from market disruptions.

Less food availability. Ebola disrupted national and local food markets in several ways, according to participants. First, participants explained that Ebola reduced food availability from impacts on livelihoods because many individuals, particularly women, could no longer engage in typical agricultural practices due to falling ill. Second, the national trade restrictions on food commodity imports and exports also contributed to less availability of food, explained key informants. The review of secondary data also supports this relationship between trade bans and decreased food availability, particular in the Nzérékoré region where compared to the previous year, 24% and 51% of the respondents from non-affected and affected areas, respectively, had to modify the ways they typically procured basic food items due to such market disruptions [33].

Reduced food access. Key informants explained that higher food commodity prices increased because of market disruptions.

“Here (in Conakry) we used to pay 40,000 to 50,000 GNF [4.4–5.5 USD]for a daily meal and we would send some to our friends, but finally (because of the outbreak) it was necessary to pay over 100,000 GNF [11.0 USD] for the same meal, because of price increases of everything in the markets.”

                                        Key informant, Hospital Supervisor, Conakry

Empirically, secondary data show that some but not all commodity prices increased nationally. For instance, while rice prices increased, those of corn, yam, and cassava declined. For the most affected areas, substantial price drops occurred from a lack of demand stemming from people’s lack of access, as well as fear and stigma associated with gathering in crowded market areas [33–35].

Qualitative data revealed also how individual livelihoods were disrupted. Garnished wages and lowered incomes, coupled with higher prices, meant reduced food access for households. Ebola-affected areas, where food commodities once were sold, saw entire local food market closures.

“It was difficult…this is a zone that produces a lot of pineapples and tomatoes. When people knew that there was Ebola into our zone, there was no business. The women that used to sell by the road side were not selling anymore because vehicles were not stopping anymore due to the Ebola epidemic.”

                                        Informant, Community Member, Friguiagbe

Coupled with local market disruptions, there were also food access challenges stemming from the mobility restrictions placed upon households where one or several members had been diagnosed with Ebola. Households had fewer coping strategies for food insecurity from mobility restrictions. Typical strategies, such as inter-household food sharing (i.e., food trade between households), were made very difficult with quarantine sanctions imposed.

“First of all, we went several months without going to markets, our markets were spoiled, and it is through the small villages that we (typically) did transactions so we could get what we needed to eat…but these village members were forbidden from moving from one place to another…we had nothing…there were people who could hardly get anything to eat.”

                                        Informant, Community Leader, Beyla

Changes to IYCF practices.

Increased levels of household food insecurity and fewer coping strategies therefore were factors that contributed to changes to IYCF practices because many households now lacked the ability to access nutritious complementary foods and were in positions where others could not help due to the mobility restrictions put in place. And at the individual level, infants and young children whose mothers had been afflicted with Ebola were advised to no longer engage in exclusive or continued breastfeeding. Also, informants ascribed underlying fears of contagion for preventing typical community interactions related to food, explaining that improving food security was one of the primary areas were the nutrition response was so helpful because organizations helped where the community could not.

The ineffective communications at multiple levels had consequences for not only care-seeking but also IYCF practices. Data suggested that traditional food-related practices (e.g., heavy reliance on energy-dense, non-nutritious staple foods) were particularly difficult to change even in the face of increased household food insecurity and malnutrition during the outbreak, underscoring the low importance placed on nutrition and reinforcing the importance of the underlying distrust and skepticism toward the biomedical community during the outbreak.

Nutrition and food assistance.

To address household food security, SuperCereal+ and other commodities, including rice, lentils, peas, rice, beans, oil, salt, and sugar, were provided. Specialized nutritious foods, including ready-to-use therapeutic foods (RUTF) for severe acute malnutrition and replacement milks for children who could no longer breastfeed, were also provided. Initially, community fear and distrust were barriers to utilizing these food provisions, with perceptions that foods given out by health staff contained Ebola. One community member explained, “…(due to the) loss of lives, so I didn’t have the courage to eat (foods given out).”

Participants explained that more intensive and appropriate sensitization efforts increased the acceptability of the food assistance over time. The RUTF was highly acceptable by community members and care staff, with reportedly high demand but frequent stock outs. “Sometimes the quantity that was demanded by the population was often more than the stock,” said a community member from Friguiagbe.

Community members without children or who had not been directly impacted by EVD were reportedly requesting the RUTF, an indication of its high acceptability. Care staff explained that they were promoting it as a “food with medicine” to reducing suspected intra-household sharing at the expense of intended beneficiaries. Key informants noted the high costs of infant formulas for replacement feeding, which was made affordable during the outbreak due to increased resources from donors, a facilitating factor to improved response efforts.

Like most aspects of the response, acceptability around the food assistance improved over time, which was eventually an important contributor to increased community confidence and trust toward the outbreak response at large. Food assistance was perceived to be a very direct and tangible link between donors and community support.

“The foods that organizations were supplying to the community, especially to victims of Ebola, started building confidence in the community and the people. But at the beginning, it was not easy. It was very difficult. The fact that the donors came closer to the community and to the people materially and financially…sensitization had already started. The community started to believe and trust the donors and their implementing partner …”

                                        Informant, Community Leader, Kindia

Social and behavior change communications (SBCC).

SBCC were implemented to improve both behaviors of care staff and community members during the response.

Interim guidelines for organizations and front-line care staff. As part of the larger multi-level communications strategy, interim guidelines were provided to organizational care staff. Key informants offered mixed opinions toward the usefulness of the guidelines, indicating not being aware or not using them. However, some key informants did indicate that while they were useful in theory, some of the recommended practices contained within them were difficult to implement in practice. For instance, participants explained that separating one’s baby from his/her mother, especially during an outbreak, was too challenging and alternative strategies needed to be provided.

“In theory, it was easy but in practice, it was not easy. For example, isolating a child from his mother is not an easy thing to do in communities where we know that there is a reinforced bond between the child and his/her mother so it is not easy to apply that guideline.”

                                        Key Informant, Nutrition Coordinator, Conakry

Further, participants explained that such care and/or nutrition guidelines can be useful, but in fact a lot of work needs to go into trainings, sensitization, and message development, for effectively putting them into practice. Psychosocial support is also needed but was/is missing in the current response, according to participants.

Messaging and sensitization to communities. A combination of mass media (e.g., radio) and interpersonal messages (e.g., community health agents going house to house; midwives giving information during childbirth) were used as channels in SBCC during the outbreak. Early on, illiteracy, distrust, misinformation, and fear hampered effective communication efforts. “During the heart of the epidemic everything was very difficult, even at the mosque …” explained one informant who articulated the magnitude of overall confusion at that time.

After more coordinated efforts that involved pilot testing as an important stage of message development, the effectiveness of SBCC improved. Interpersonal messages were more accepted due to greater trust and familiarity associated with fellow community members who passed those messages, namely community leaders, midwives, and community health agents. The health agents were particularly instrumental in passing messages from health clinics to community members, according to interviews with midwives.

Informants also explained that community-level committees were set up by the community members themselves, and proved to be a critical liaison for effective communication between the biomedical and local communities.

“We put into place sub-committees. We chose the able-bodied youth to be trained as community agents and then they were answerable to the head of the health center…who represented the community to the health system. Decisions were coming from the head of the health center to the community health agents. Then with the supervision of the head of the health center, they (community health agents) went out every time to check the state of certain children (screening) because there were some (children) with hot skin and community agents were informed of these cases.”

                                        Informant, Community Leader, Kindia

One participant explained that because community members, “really trust the old traditional ways…” referring to health and nutrition behaviors, as well as benefit from clearer explanations in local dialects, messaging through familiar, interpersonal channels was important for positive behavior changes.

Several participants gave the example of a commonly-held perception that giving a young child diverse foods will cause constipation during complementary feeding. One example of effective communications to help overcome local perceptions was sensitization through community-level trainings. These trainings reportedly were not in place early in the outbreak but did improve greatly over time and focusing on improved behaviors of WASH, appropriate acute malnutrition screening, dietary diversity during complementary feeding, feeding with RUTF, and appropriate breastfeeding practices, for example.

“It took a major investment from [UN agency] and its partners to rebuild trust within the communities, in terms of service quality in giving higher-quality care (during the outbreak), and also in terms of demand generation so that communities would resume (typical health) consultations within the existing health structures.”

                                        Key Informant, Management, Conakry

Some nutrition-related behavior changes persist today, according to interview data. Informants explained that handwashing practices and malnutrition screening are two primary examples of positive behaviors that have been carried over as part of typical community health practices today. On the downside, some interviews indicated that bottle feeding practices have also continued until this day.