Development of a Set of Community-Informed Ebola Messages for Sierra Leone

"[S]ystematic, iterative qualitative research should be fully embedded into the message development process from the outset of the response to any future public health emergency, and the communication of those messages should also be conducted on a two-way basis."

This paper describes a project conducted in Sierra Leone in early 2015 that was designed to receive direct community input into the development of messages aimed at preventing the spread of Ebola and at promoting Ebola treatment-seeking behaviours. The work was conducted by a consortium including individuals affiliated with three institutions (two from Sierra Leone, one from Sweden), working in collaboration with relevant national and international stakeholders. Through this, the messaging strategy aimed at being relevant, actionable, and acceptable to all the key actors. In this paper, members of the study team present the methods by which they developed the messages, the format of the messages, and some of the main lessons learned in the process of developing them.

The paper presents a set of evidence-based, culturally contextualised Ebola messages that are based on the findings of qualitative interviews and focus group discussions (FDGs) conducted in "hotspot" areas of rural Bombali District and urban Freetown in Sierra Leone, between January and March 2015. Numerous misperceptions about Ebola had developed in the community over the course of 2014, and it was becoming increasingly clear that the largely top-down messaging strategy in use was not as effective as it could have been. Once the World Health Organization (WHO) declared Ebola to be a Public Health Emergency of International Concern and the pace of the response began to pick up, there was an increasing recognition that local knowledge and full engagement with the affected communities needed to be at the heart of Ebola control efforts. The project thus aimed to actively include the community in the development of a set of actionable Ebola messages that responded directly to their needs and concerns.

An iterative approach was taken in the three-month-long message development process, which began with (i) a formative research phase that included 16 FGDs comprised of 118 people without any sort of leadership position in their respective communities. Researchers also held a total of 24 individual in-depth interviews with community leaders of various sorts (12 in each study area; 13 male, 11 female). These included religious leaders, traditional leaders, traditional healers, women's and youth leaders, and medical staff engaged in the Ebola response. Next, (ii) data from the formative research was subjected to thematic analysis to identify areas of community concern about Ebola and the national response; (iii) 26 draft messages to address these concerns were produced; (iv) the messages were field-tested; (v) the messages were refined; and (vi) a final set of messages on 14 topics was disseminated to relevant national and international stakeholders. Each message included details of its rationale, audience, dissemination channels, messengers, content, and associated operational issues that need to be taken into account.

The results section of the paper is presented in four sub-sections:

1. Four cross-cutting principles of message development emerged through the data, that the study team took into account in their work:
   • The respondents during the formative research and the field-testing said they wanted to hear messages from people they trust. This can present a challenge when Chiefs or other political leaders are used as the messengers: While some people trust them, others may view them - and any message they disseminate - with suspicion.
   • Respondents made clear that in addition to including one-way messaging strategies such as radio discussions, jingles, and billboards, they wanted the messages to be disseminated through community meetings, house-to-house visits, and other such face-to-face dissemination strategies.
   • Messages to promote treatment-seeking behaviour must be matched by good services; otherwise, the credibility of those messages will be undermined.
   • Two locally salient issues emerged, the first of which concerned messaging colours. (If a given message was associated in people's minds with any particular political party, it risked undermining the validity of that message in the minds of people who support another party.) The second local issue concerned language; the study team considered it necessary to provide both English and Krio versions of all the messages in their final messaging document.
2. An explanation of the different components was given for each of the messages, including: (i) the rationale for each message; (ii) the audience to whom the message is directed; (iii) the channel/s through which the message would be disseminated, including, among others: radio discussions involving pastors, imams, and/or youth and community leaders; participatory theatre/drama in the community followed by questions and answers sessions; jingles disseminated via the radio or mobile public announcement (PA) systems; and social media; (iv) the messenger/s; (v) the message content: for example, to combat stigma, the messages were "Our Ebola workers are our heroes!" and "Do not laugh at or avoid Ebola survivors"; and (vi) associated operational issues that would need to be taken into account to make it effective.
3. An example of one of the messages was developed as an illustration of the project output. Because fear of chlorine in ambulances and by burial teams had arisen as a major concern in the formative research, a cartoon character with a friendly face, "Mr Chlorine" (see above) was created by one of the project artists with the accompanying text: "I am Mr. Chlorine, your friend in the fight against Ebola." / "Mi na yu paddi Mr. Chlorine, ar go hep yu fo fet Ebola."
4. The dissemination process was described, which highlighted some important lessons for future public health emergencies. For example, the strategy entailed close networking and engagement with key national- and international-level stakeholders as a means of raising awareness of the work, receiving relevant input, and thereby enhancing the likelihood of the product being adopted into practice.

Through the project experience, a set of recommendations emerged that the authors argue could be incorporated into future international and national public health emergency preparedness and response plans. This "toolbox" includes: the importance of embedding systematic, iterative qualitative research fully into the message development process; communication of the subsequent messages through a two-way dialogue with communities, using trusted messengers, and not only through a one-way, top-down communication process; provision of good, parallel operational services; and engagement with senior policymakers and managers as well as people in key operational positions to ensure national ownership of the messages, and to maximise the chance of their being utilised.

In discussing their experience and its implications for others developing messaging in public health emergencies, the researchers lament that lessons learned in public health emergencies are frequently not taken on board and institutionalised for use in future situations. They cite an example of an outbreak of Marburg Haemorrhagic Fever (MHF) in Angola in 2005, during which the initial messaging, that "There is no cure for this disease", was reportedly understood by community members to mean that even if they accepted hospitalisation, death would be inevitable; this discouraged people from going to hospital, so they stayed in the community, thereby sustaining the cycle of infection and prolonging the outbreak. This same mistake was made during the early Ebola epidemic in Sierra Leone, even when the consequences could quite easily have been foreseen. "[I]n many respects," the authors argue, "it appears not to be a question of an overall lack of knowledge about what to do, but rather of ensuring that the right people have (a) access to the right knowledge and (b) the ability to operationalise it."

For future Ebola outbreaks, the universal messages from WHO on, for example, signs and symptoms and modes of transmission are instantly available; while the culturally contextualised materials from this study, combined with other messages that were produced during the West African outbreak, could be utilised as a starting point in the rapid production of an initial round of messages. These would need to be subjected to field-testing so as to ensure relevance and to minimise the risk of unanticipated consequences.