In an era where billions of people use social media across the globe, the COVID-19 pandemic brought with it an unprecedented infodemic: abundant misinformation spreading faster, one could argue, than the disease itself. To respond to the infodemic, the World Health Organization (WHO), the United Nations Children’s Fund (UNICEF), the International Federation of the Red Cross and Red Crescent Societies (IFRC), and other agencies began to regularly share and publish reports that monitor digital trends through social listening and distill community feedback via regular community engagement activities. While the reports were lauded as excellent tools to inform programming, it also meant that there were numerous reports for key staff to review, creating yet another form of information overwhelm.
“Before the Joint Report we had several platforms we had to keep up with. This makes things much easier to see what is important” — Wilson Kipkoech, World Vision
The Risk Communication and Community Engagement (RCCE) Collective Service, working with several partners, identified a need to streamline the content of social listening reports and community perceptions data into one Joint Report. The Joint Report triangulates data from UNICEF Social Listening Reports, the African Information Response Alliance (AIRA) Infodemic Reports, and the Community Feedback Report, each compiled using different methodologies.
- Social Listening Report: Using artificial intelligence (AI) powered tools and analyst-driven review, this report compiled by UNICEF analyzes digital news, online search trends and social media discourse related to immunization, education, and health services in Eastern and Southern Africa.
- AIRA Infodemic Reports: This report, developed by WHO AFRO, analyzes trends observed on social media platforms, which it compiles using a variety of social listening dashboards specific to the COVID-19 response. When possible, circulation of rumors and misinformation on closed channels (e.g. WhatsApp) are confirmed in the report.
- Community Feedback Report: This report relies on partners on the ground, including Ministries of Health, to input data collected from community members. Based on the IFRC’s community feedback methodology, partners report on prevalent conversations and community insights in the communities where they work.
Collective Service partners, led by AIRA, work together on a monthly basis to discuss the data from these reports, which are published throughout the previous month, to determine a theme, highlight the most important information, and produce critical recommendations for RCCE practitioners at the regional and country levels.
- Information gathering:
- Online survey tool for partners contributing to the community feedback report digitizes information in real time so that it can be analyzed and accessed more readily than paper reports
- Social media platforms (e.g. Twitter, Facebook) are analyzed for trends and misinformation and rumors circulating on closed channels (e.g. Whatsapp) are confirmed when applicable
- High-demand information provided to key staff and decision-makers
- Trending conversations and misinformation flagged for community outreach
- Increased communication and good debate between partners creates a coherent data narrative
- Reduced message confusion by harmonizing information from different sources
- Reduced burden on decision- and policy-makers to synthesize data from different reports and dashboards
- Offers regional- and country-level recommendations
- Development of content to address rumors and trends.
- Established a technical infrastructure and network of partners to inform future emergency response actions
- Creates awareness about the importance of RCCE as part of the COVID-19 response
- Serves as a model for multi-country, multi-organizational collaboration
- Favorite tool for Collective Service partners to make COVID-19 program decisions
- Helps community programs to stop rumors before they start
- Allows organizational partners to speak with one voice to combat misinformation
SPOTLIGHT: THE COMMUNITY FEEDBACK REPORT
“The Community Feedback Report has helped us tremendously. We used to have a closet full of papers [M&E forms] that no one had the time to read or analyze. Now the data we collect is put to use immediately.”—Tendayi Gudo, National Director Apostolic Women Empowerment Trust
In an era of suspicion and fear magnified by the power of the internet, it is crucial to gain community trust, understand community-level perceptions, and respond in real-time to dispel myths and correct information. The Community Feedback report is a real-time dashboard built by the Collective Service Information management pillar using methodology developed by the IFRC, with extensive input from partners.
IFRC methodology to collect community feedback focuses on questions, beliefs, rumors and concerns to advocate for community needs. Feedback is often collected during household visits, interactive radio shows, social media, or WhatsApp. While findings are not statistically representative of the communities where data is collected, the data is nonetheless useful to point out potential red flags for triangulation with other data sources.
While organizations have long been gathering monitoring data on paper, without an infrastructure to digitize this data, it is often outdated by the time there is an opportunity to review it manually. With the Community Feedback Report technical infrastructure developed by The Collective Service, NGO/CBO and/or government staff can gather information from the field using a tablet and an online survey tool (KoBoToolbox or Google Forms ) and upload this data to a real-time dashboard. The dashboard offers an array of relevant data and collaboration tools customized to the response. As KoboToolboxs collects qualitative data that is instantly centralized, it can be analyzed much more efficiently than traditional data gathering methods. Moreover, KoBoToolbox provides organizations the opportunity to digitally search, review, and analyze their data in order to update communications plans and make time-sensitive programmatic decisions.
The ESAR Community Feedback Dashboard codes and organizes the date of data collection, origin, summary feedback, example quotes, and recommendations into visualizations that put the data into perspective.
FROM THE FIELD
Investigating vaccine confidence and acceptance among health care workers in Ethiopia, the Community Feedback Report noted that a viral video might be impacting vaccine acceptance. In the video, people were being vaccinated in their right arm, and a chip inserted. This video was tagged #666, #markofthebeast, referring to the Biblical chapter Revelations in which people are marked as followers of Satan on their forehead or right hand.
During a partner meeting, it was noted that the video was also picked up in the UNICEF Social Listening Report. Ethiopian partners were able to share more details around the specificity of the video including, in order to allay fears, that vaccine administrators offered to vaccinate health care workers on their left arm, which health care workers accepted.
The recurrence of the viral video in both sources confirmed the rise of this conspiracy. New content was created at Viral Facts Africa to dispel the myth, and WHO was able to revise its social listening algorithm to continue to monitor for vaccine hesitancy related to religious symbolism.
“What we see is that rumors and misinformation online are mirrored in our communities, even with lower rates of internet penetration.” – Daudi Ochieng, WHO Afro
With so many sources of information and methodological considerations for each data type, it can be difficult to prioritize data into actionable information and recommendations for responders. The key words, or taxonomies, used to tune artificial intelligence (AI) on big data sets, build in biases that must be discussed and justified. Rumors and misinformation in communities must be analyzed to determine how widespread or serious the impacts might be. Recommendations must be realistic and account for differences across the region. At the same time, this process of deliberation and discussion builds solidarity among partners and enhances confidence in the report’s conclusions and recommendations.
10 LESSONS LEARNED
- Too many social listening reports/platforms make it difficult for decisions-makers to focus in on what is important. Initiatives like the Joint Report help programs make confident decisions.
- Creating the Joint Report not only provides critical information, but it also builds solidarity among partners through the process of discussion and deliberation.
- The triangulation of data is especially critical to confirm long-standing issues and generate discussion on potential solutions.
- Often overlooked but essential is to build monitoring and evaluation into the dissemination process of the Joint Report in order to track utilization of the data and recommendations.
- It is important to disseminate the report widely in order to receive feedback from a broad range of actors and geographic locations.
- The COVID-19 Joint Report stimulates discussion and creates awareness concerning the kinds of assistance that the RCCE Collective Service can provide.
- It is useful to designate a small team or one point person to manage the creation of the Joint Report including convening meetings among partners, following up on concerns, tracking improvements, and monitoring utilization.
- Contributing to the Joint Report, especially through the community feedback mechanism, is an organizing activity that connects partners at the country level, and connects country-level partners to the regional level.
- The Joint Report eases collaboration for organizations working across large geographic areas.
- Partners can lose interest in the report over time if they are not consulted for feedback or make direct contributions.
The Joint Report has been widely praised by Collective Service Partners as an instrumental part of COVID-19 programmatic decision-making. The success of the Report has sparked recent calls for wider dissemination beyond RCCE practitioners. The Collective Service is currently looking at ways to make this information more accessible to more people and to support other regions to follow the same process.
“Reporting and entering information to the dashboard is an organizing activity that brings partners together.” – Blessings Mtuwa Nkhata, UNICEF Malawi
At the start of the COVID-19 pandemic, there were several organizations and government departments working independently on risk communication and community engagement (RCCE) in Malawi. As the focus on public health interventions increased, it became clear that RCCE activities were missing in some areas of need, being duplicated in other locations, and/or overlapping with the same target populations. Organizational partners of the Collective Service requested the assistance of the Collective Service’s Information Management team to find a solution that would help partners make COVID-19 programming more efficient in Malawi.
“On the coordination side, the dashboard brought renewed energy to partners from COVID fatigue to rally around a common agenda.” – Mr. Chikumbutso Mtumodzi, Director of Information, Ministry of Information Malawi
The Collective Service gathered feedback from implementing partners and government counterparts at the national and district level, deciding on a “Who is doing, What, When, Where” or “4Ws” dashboard that would map out the organization, location, type, duration, and target population of COVID-19 RCCE activities on the ground. Although there were some existing data platforms before the dashboard, it was hard to see how the data between platforms could be synthesized. Stakeholders therefore expressed the desire for one harmonized platform that could be decentralized and managed collectively. In addition, Collective Service partners and participating organizations wanted to visualize epidemiological data for COVID-19 alongside partner mapping inputs. This would allow them to clearly identify gaps and quickly deploy resources to areas in need.
The Information Management Team created a Power BI dashboard populated with a simple online survey tool (KOBO) that could be updated in real time. The team carried out an iterative process with stakeholders over the course of several months, ensuring a user-friendly interface and useful functionality in the dashboard. With the online survey tool and dashboard, organizations could input data regularly, read the visualizations in the dashboard without difficulty, and create an evidence-base for decision makers to adjust strategies and collaborate with other actors.
As a powerful tool that facilitates collaboration on-the ground, the dashboard provides a wide range of information including:
- COVID-19 case numbers compared across districts
- COVID-19 case numbers over time by district
- of organizations working in each district by sector
- of organizations implementing RCCE activities by activity type
- of organizations using communication channels by channel type (e.g. TV, radio, SMS, etc.)
- of organizations working with various target populations
- Organizations offering RCCE training
- Organizations requesting RCCE support and identifying present challenges
“The Dashboard became a unified, multi-sectoral approach that allowed us to leverage all our resources. We engaged faith leaders and other kinds of organizations, not just purely health-related organizations to fight COVID-19.” –Kenford Mayere, UNICEF Malawi
- Effectively coordinates 41 organizational and government partners by creating a common activity that unites stakeholders in the fight against COVID-19
- Leverages non-traditional public health resources, such as faith-based or educational organizations to fight COVID-19
- Encourages evidence-informed decision making based on epidemiological data
- Serves as a model for replication in other countries
- Works as an excellent use case to engage organizations to work with the Collective Service to identify custom-fit Information Management solutions
- Provided a platform to discuss and invest in the COVID-19 vaccine rollout
“We are committed to exploring sustainable solutions to the challenges in linking the dashboard to existing systems at the Ministry of Health.”– Dr. Annie Chauma Mwale, Head of Epidemiology and Surveillance, Malawi Government
The biggest challenge with the COVID-19 RCCE dashboard is an ironic outcome of its greatest strength: identifying gaps and duplications. Once a gap or duplication is identified, it requires that everyone work together to find a common solution, negotiate the redirection of funds, and use one voice to move forward in speaking with communities. This can sometimes be a painful process and requires patience and diplomacy. Fortunately, the dashboard as an intra-organizational activity also generates a culture of solidarity and team spirit. With enough communication and understanding, addressing gaps and duplications can build even stronger bridges between partners.
Another challenge has been consistently linking the data collected by non-governmental partners with government systems, especially government-collected epidemiological data. To support this, The Collective Service has encouraged ownership of the dashboard within government and has collected evidence on effectiveness and utilization of the dashboard to emphasize the value of this activity. This is an important challenge that must be overcome to ensure the sustainability of the dashboard.
- Inputting data into a country-wide dashboard facilitates coordination between partners by creating a shared goal/activity.
- Several agencies within the same government can overlap activities without being aware of each other’s activities.
- All sectors, not just traditionally public health actors/organizations/government agencies, should be included in the dashboard activity. A multi-sectoral approach expedites the mobilization of resources and provides a holistic approach to public health events.
- Mapping activities not only show gaps and duplications, but also reveal when colleagues are not following through on proposed activities.
- If proposed activities are not taking place, government partners can step in to ask for accountability or alert funding organizations that there is a discrepancy between what is happening and what is being reported.
- It is helpful to set up an RCCE committee with a wide range of actors such as faith-based leaders, local municipal leaders, and NGO/CBO staff to convene regularly and discuss the data that has been entered into the dashboard.
- The identification of the duplication of efforts requires patience and diplomacy to negotiate programming changes, especially with regard to funding mechanisms.
- Mapping exercises can bring up power struggles between and within organizations. It is helpful for the RCCE committee (see lesson #6) to help mediate programming changes.
- An effective way to keep partners on the same page is to remind them that the dashboard is about generating evidence for RCCE work. Keeping partners focused on the evidence creates a shared value to align the coordination of activities moving forward.
- Building capacity for all users of the dashboard will ensure wider and more frequent use of the dashboard and continue to create demand.
- It is essential to monitor how the dashboard is being utilized so that there is an evidence base to justify its continued use at the country level.
- To expand the types of data and make data more robust, it is best to create ownership within government for this activity.
The dashboard continues to be showcased to other countries as an accessible tool for data visualization that helps organizations identify gaps and negotiate overlaps to improve the efficiency of programs nationwide. It is also a testament to the kind of innovative support that the Collective Service can provide through Information Management. The success of Malawi’s dashboard has become a model for Eastern and Southern Africa (ESAR), with similar dashboards already being implemented in South Africa and Somalia, and underway in Kenya and Zimbabwe.
At present, the Collective Service is currently working to expand the dashboard to include qualitative community feedback data, and to hand over management of the dashboard to UNICEF. The Collective Service will continue to proactively monitor and document the ways in which RCCE dashboards are changing the landscape of inter-organizational coordination and programming in the ESAR region.
“[During] the last Ebola outbreak, the partners worked in isolation, each with their own agenda. This time, the joint commission was much more organized and we were able to manage rumors and give accurate information, which made us more credible.” – Christophe Milimolo, Chair of radio network in Nzérékoré
On February 14, 2021, an Ebola Virus Disease (EVD) outbreak was declared in Guinea, following the confirmation of 3 cases in the region of Nzérékoré. As a trade center that shares borders with Cote d’Ivoire, Liberia, and Sierra Leone, an outbreak in this region was of particular concern. This was the first outbreak in Guinea since the 2013-2016 outbreak, which claimed more than 2,500 lives. The World Health Organization (WHO), the International Federation of the Red Cross and Red Crescent Societies (IFRC), and the United Nations International Emergency Children’s Fund (UNICEF) worked together on a community-level response for the treatment and prevention of EVD. With infrastructure in place from the COVID-19 response, the Collective Service for Risk Communication and Community Engagement (RCCE) at the global, regional, and country levels facilitated coordination between the three agencies and leveraged existing networks, resources, and information systems to help provide an innovative, community-centered approach to the outbreak.
Just four months later, the Ebola outbreak was declared over, with only 23 recorded cases and 12 deaths. The success of the response is a testament to the immense efforts made by the Guinean people, and the synergy of partners who together, provided collective knowledge, skills, and resources to prevent an epidemic.
As a dedicated collaborative partnership put in place to fight the COVID-19 pandemic, the Collective Service and its core partners began facilitating coordination and gathering resources as soon as the EVD outbreak was declared. Global and regional staff of the Collective Service adapted information management tools for data collection and analysis and assisted with deployment at the country-level in order to share community feedback across the response. The RCCE coordination team at the regional level created a repository of tools, capacity building assets, policy documents and minimum actions built on experiences from the 2018 EVD outbreak in the Democratic Republic of Congo. These documents were made available to all partners engaged at the country level and promoted through dedicated coordination staff of the Collective Service who were able to nurture collaboration, facilitate connections, and enhance awareness of partner activities in order to avoid the duplications of efforts.
As part of the commitment to a consistent approach to community engagement, the Provincial Health Department along with UNICEF, WHO, IFRC, and the Guinea Red Cross, taught a comprehensive set of modules to train key actors and stakeholders on a range of topics including: epidemiology, transmission, vaccination, public health and safety measures (PHSM), how to communicate about the disease within Guinea’s cultural context, how to work with communities to increase buy-in, how to manage resistance, and how to use media to create awareness and dispel myths. Each of the agencies brought forward their expertise to provide comprehensive training to members of the Response Commission, and the RCCE coordination team was consistently available to support complimentary partnerships.
“During the last epidemic some team members were killed by furious populations. The joint training on reluctance and the sub-commission [on reluctance prevention and management] helped us to manage resistance by combining efforts. We utilized mobile radio and the community feedback mechanism to reach out and exchange with communities to overcome resistance” — Diaoune Mamadou, locally elected municipal official
Following the joint training, a General RCCE Commission and several focused sub-commissions were formed, matching key actors to different aspects of the response. The General RCCE Commission was composed of the Prefecture Department of Health (DPS), the Collective Service core agencies and dedicated interagency coordinator, international NGOs, local NGOs, local elected actors, faith-based organizations, women and youth organizations, traditional-influential leaders (Zowos), radio producers, journalists, and social media influencers.
The General RCCE Commission met daily to discuss updates on the evolution of the response’s activities including logistics, next steps, challenges, community feedback, data collected through social mobilisers, and reports from the sub-commissions. RCCE sub-commissions, composed according to key actor profiles and interests, met weekly and focused on reluctance prevention and management, community feedback, media, and social science research. Focal points in RCCE also attended the daily meetings of the other pillars of the response: surveillance, safe and dignified burial, medical assistance, and control points. Information collected from the other pillars was then integrated in the action plan. Finally, social mobilizers went door-to-door over the course of the response to speak with community members, gather feedback and insights, listen to questions and concerns, ad document rumors and myths. They also supported community action, referred ill community members to health centers, and spoke to community members who had a deceased loved one about testing for EVD, as well as safe and dignified burial. These social mobilizers consisted of community members engaged in the training, in consultation with the RCCE General Commission
This coordination model allowed agencies and partners to deliberate on the response and then speak with one voice to communities. Communities could also respond to just one voice, avoiding confusion and chaos on the ground. RCCE was integrated into each of the response pillars, and with the help of the RCCE coordination team, a community feedback mechanism was put in place to create a direct line for communities (via conversations with social mobilizers and others) to express needs, fears, and expectations.
The first cases of EVD in Guinea were confirmed in Pagalay. In this village, female traditional healers play a special and powerful role in village decisions, particularly when it comes to preparation and burial of the dead. It was therefore particularly important to engage the community in a way that honored their role. UNICEF, IFRC, the Guinea Red Cross and WHO worked together to engage and empower the local healers and created a series of community conversations that enhanced transparency and presented the response as a unified effort.
Each of the proposed activities were acted out so that the village could provide feedback about any changes that were unacceptable, and also, to familiarize themselves with what would happen in the coming months. A mock burial was carried out in full view of community members, along with presentations on testing for EVD, vaccination, and what it would look like to talk to social mobilizers if a death in the family took place. These conversations were well received and the village leader publicly signed a written agreement to cooperate with the Response.
The conversations proved to be extremely fruitful in Pagalay, and village members began to take ownership of the effort to eliminate EVD in their village. People were able to receive medical care with either a suspected or confirmed case of EVD. One man who tested negative for EVD was able to receive medical care for a different ailment, which became a source of great praise toward the Response. By enrolling community members and attending to their needs, the Response was able to correct misunderstandings and Pagalay became a nationwide model for excellent EVD response at the community level.
Suspicions and notions of conspiracy were a major obstacle to the response. Vaccines, masks, and medical care were all claimed to be ways to contract EVD, rather than to prevent or treat it. Socio-economic disparity between response personnel and communities created misunderstandings and fueled accusations of ulterior motives. In one community, where cars were rarely seen, the sudden influx of visitors by car was obvious proof to the community that EVD was a hoax, designed to use the community to make money. There was also wide speculation that medical staff were harvesting organs from EVD patients for the black market.
Challenges also differed across generational lines. With greater access to disinformation coming from the internet, younger people tended to push back more on information coming from social mobilizers. Sometimes the spread or lack of spread of the disease did not fit within the typical epidemiological model. In instances where a confirmed death from EVD did not spread to close family members, it confirmed conspiracy theories that the EVD outbreak was not real.
||EXAMPLE(S) FROM GUINEA
- Oftentimes it is not the explanation of scientific information or rationale of health interventions that convinces people to comply with public health and safety measures. Responding to community requests and concerns and including communities in the implementation process builds trust in the response and creates ownership. It is based on this trust that people begin to seek medical care and comply with PHSM.
|After social mobilizers were recruited for the response, it was discovered that many of them did not belong to the communities where they worked, or they were the local medical chiefs in the area. Acting on complaints from the community, the Response replaced the social mobilizers with local members of the community and the communities noticed. The trust built from responding appropriately to this request increased compliance and cooperation with Response teams.
- A deep understanding of cultural norms is important, and it is useful to hire or work with social science researchers. They can provide critical insights into how to work with communities.
|In the village where the first cases of EVD were reported, female traditional healers were traditionally responsible for preparation and burial. An anthropologist was able to facilitate discussions with local women to increase cooperation with the Response’s implementation of secure and dignified burial.
- A long-term presence on the ground builds trust with communities and provides associative credibility for emergency partners.
|In communities where certain organizations were not as well-known, local trusted organizations and volunteers were able to facilitate introductions and build trust for all emergency partners.
- It is important to consider how economic differences between response personnel and communities can impact trust and cause suspicion, jealousy, or resentment.
|In one community, where cars were rarely seen, the sudden influx of visitors by car was obvious proof to the community that Ebola was a hoax, designed only to use the community to make money. Once this issue was identified, response personnel made a concerted effort to carpool and limit the number of vehicles brought into the area.
- Collaboration across agencies harmonizes messages coming from emergency partners and creates less confusion in affected communities.
|Organization of the joint training allowed each agency to offer their expertise to the response and ensure that all relevant topics were covered. Key actors who participated in the training therefore harmonized their approach and spoke with one voice, providing information and messages that were disseminated through various communication channels (NGOs, religious leaders, radio, etc.).
- Providing transparent and detailed information about the response plan is critical to creating buy-in and increasing compliance in communities.
|Visual representations of activities that would take place were acted out for communities. This reduced shock/anger, and overall resistance to the Response.
- An effective response should include many different types of actors in the community, each with their own audiences and modes of communication.
|Faith leaders, local municipal leaders, radio producers, NGO, and CSO staff were all included in the general commission and sub-commissions. This allowed everyone to leverage their skills, networks, and resources to get harmonized, accurate messages to affected communities.
- Risk communication and community engagement is just as important as epidemiological surveillance, medical assistance, secure burial, and control points in outbreak response.
|In previous EVD outbreaks, communities were not always engaged in ways that empowered community members, and some violence occurred. Bringing community engagement to the forefront reduced community resistance and paved the way for other pillars of the response to operate smoothly.
- Using local and community-driven media content facilitates better ownership of messages and participation in the response.
|Deploying mobile radio stations in the affected areas strengthened trust in messages disseminated by familiar, local actors. The radio stations also allowed communities to be able to talk to each other through live broadcasts, provide feedback on specific messages, and promote good examples and practices of community members.
- Integrating real-time perceptions data provides life-saving information to communities and response staff.
|Systematic information sharing among response staff provided critical insights to deal with community resistance. Given the security threats during the previous outbreak, this had the effect of building confidence among Response staff that they had the right information to work with communities. At the same time, Response staff were able to respond quickly to suspected cases.
“As an Imam, I benefited greatly from the training on how to prepare dignified burial[s] for people who died from Ebola. I want to expand this training to mosques across Guinea.” — Amadou Soumaoro, President of Muslim Council
The Collective Service at the global, regional, and country levels will continue to maintain its networks and partners to be able to respond to emerging issues in Guinea and elsewhere, such as EVD outbreaks. Lessons learned from the experience in Guinea will be added to the legacy documents and tools from the 2018 EVD outbreak in Democratic Republic of Congo as a resource that can be quickly adapted to future emergencies. The community feedback mechanism, an immense effort of the partners engaged in the Collective Service and the dedicated Information Management team, is currently being expanded across Western and Central Africa and Eastern and Southern Africa as an up-to-date monitoring tool to complement surveillance, and to keep a pulse on community perceptions that impact public health. In Guinea, with the network of trained professionals from the joint training, public health messages and community engagement interventions continue for COVID-19 along with preparedness activities for other potential emergencies.
The success in Guinea is proof that having ready, dedicated RCCE coordination and information management staff, along with a systematic approach to RCCE coordination at all levels, is critical to disease prevention and management. Prioritizing communication and coordination across partners and agencies, the Collective Service will continue to promote expert-driven, localized RCCE support to governments and partners in order to reduce redundancies, save partners resources and time, and ultimately save lives.
“The RCCE Collective Service database creates a public good so we don’t have to use our own resources to get the same information. This saves us [as partners] so much time and money and means we can make more efficient use of funding” — Tom Black, The Gates Foundation.
The risk communication and community engagement (RCCE) Collective Service database is a first-of-its-kind repository of social and behavioral data on COVID-19, covering 198 countries and consisting of over 250,000 data points from 411 different data sources. The database, and the dashboard that visualizes this data, is an initiative borne out of the Collective Service’s (CS) commitment to provide critical social and behavioral information to partners working to end the COVID-19 pandemic. Insights drawn from the dashboard and the database have come to influence how partners design interventions, make investment decisions, and incorporate RCCE into broader public health programming.
Since its inception a little more than a year ago, the database/dashboard has become a key resource informing COVID-19 programs and initiatives among CS partners including WHO, IFRC, UNICEF, The World Bank, The Bill & Melinda Gates Foundation, and many others. It has strengthened existing networks and partnerships by building unique, invaluable datasets that may not have been available otherwise. At the same time, it has generated new collaborations by attracting organizations who want to contribute, expand, or utilize the database/dashboard to improve vaccine uptake, adherence to prevention measures, and other intervention applications. Finally, as an engaging and informative data visualization, the RCCE dashboard generates relatively significant web traffic, showcasing the capacity of the CS to collate, vet, and make publicly available, high-quality social and behavioral data.
This table provides an example of how data is analyzed for representativeness. Only data that meets “Good” or “Acceptable” standards are included in the database. Other limitations include doubts about data quality, including conflicts of interest
PROCESS & FEATURES
To maintain, update, and expand the database, the Information Management team at the CS pulls relevant data from websites, dashboards, field work assessments, surveys, opinion polls, reports, and peer-reviewed publications. Creating a coherent database from such a diverse range of sources is a fundamental challenge. During the design phase of the database, the Information Management team of the CS worked across pillars, engaging the expertise of the Social Science team to create an innovative data selection methodology and structure the data according to the COVID-19 Behavioral Change Framework.
At the outset, there were methodological obstacles to comparing datasets with different sample sizes, time scales, instrument designs, and so on. While a traditional approach to harmonizing the data might have required that data adhere strictly to one universal survey or data collection method, partners quickly realized that such an approach discounts the need for data collection to be context specific and severely limits the amount of data that can be explored, compared, and triangulated. The CS therefore developed transparent selection criterion to inform users regarding representativeness within the data.
Data that has been rigorously evaluated for inclusion in the database is also classified according to how well it matches the COVID-19 Behavioral Change Framework. This Framework, which underpins the RCCE database/dashboard, was developed in consultation with CS partners and is based on the UNICEF behavioural driver model. A comprehensive set of 35 RCCE core indicators, along with six dimensions for each indicator, characterize quantitative data concerning perceptions, knowledge and practices, and social and structural factors that influence health behavior in the context of the COVID-19 pandemic. Having gone through selection, indicator matching, and a representativeness evaluation, the data is then compiled and aggregated to populate the dashboard, or for users taking a deeper dive into the database.
This flow chart provides an overview of how social and behavioral data for the database/dashboard is processed from beginning to end.
Big data approaches are still somewhat new in the field of public health and international development. In the beginning, some partners were concerned about data quality and wanted only their own data in the database. The CS responded by making a concerted effort to help partners understand the value of creating a repository at scale and addressing concerns by offering confidence interval calculations for data sources.
Depending on the context, the initiative also encountered a great deal of hesitancy at the country level by partners who were concerned about privacy and potential negative ramifications of sharing data that could be seen either by the local government or by international organizations like the WHO. Again, the CS worked hard to strengthen these partnerships and build trust so that data could be shared safely and respectfully.
Finally, the Information Management and Social Science teams struggled to capture the complexity of social and behavioral measures in such a brief format. This has meant that the data synthesis is sometimes not rich or detailed enough for use at the country level and there is a recognized need to improve the data for country-level decision making. For example, efforts are currently under way to develop data visualizations that can show relevant qualitative data alongside quantitative metrics.
“The scale of the Collective Service’s networks and partners gives me confidence that the right inputs are in the database. I’m not sure where else I would be able to find this data.” – Jeremy Cooper, The Gates Foundation
- The Gates Foundation has used the RCCE database to perform multi-variate regression analysis on vaccine acceptance to understand potential correlations between vaccine hesitancy and case rates and/or death rates. The World Bank has used the RCCE database to uncover the efficacy of interventions via in-depth analysis of country-level data. This analysis was used to inform funding decisions.
- WHO EURO is working closely with the Collective Service to integrate 19 in-depth behavioral insights surveys conducted across Europe.
- WHO HQ uses the dashboard/database as an invaluable resource for shaping and informing COVID-19 policy and research. The data is incorporated into regular briefings and analysis including WHO member states briefings and international expert consultations with national and regional Centers for Disease Control (CDC), among others. Identifying the latest social and behavioral trends is an essential part of advocating for RCCE responses at the global, regional and national levels.
- The Africa Infodemic Response Alliance regularly uses the RCCE database for understanding social and behavioral trends and triangulating this data with social media listening in Eastern and Southern African countries.
In many ways, the RCCE global database/dashboard has been a point of departure for critical social and behavioral data that can inform RCCE policy and action. Creating and implementing a workable methodology, building a set of meaningful indicators, and getting partners on board to understand the value of the project was just the first phase. As the initiative continues, the RCCE global database/dashboard will be expanded to include the all-important qualitative data that is so essential to building communications campaigns and engaging communities. As part of this effort, the CS plans to incorporate qualitative data being collected through community feedback mechanisms. In the future, the CS hopes to provide automated reports to inform activities and coordination at the country-level.
More focused dashboards may also be developed in the future in order to visualize finer details within the database. Just as important, the CS will soon release a web API (public access to the database through a back-end server) that will allow partners to create their own dashboards/visualize data that is relevant to their own interests and projects.
As the CS moves forward, we will continue to document how partners are utilizing the database and share these stories within the CS community in order to spark new ideas, nurture collaboration, and build solidarity to fight COVID-19 and other social and public health issues as they arise.
For more information about the RCCE dashboard/database please visit: https://www.rcce-collective.net/data/behavioural-indicators/