Using Population Mobility Patterns to Adapt COVID-19 Response Strategies in 3 East Africa Countries

The COVID-19 pandemic spread between neighboring countries through land, water, and air travel. Since May 2020, ministries of health for the Democratic Republic of the Congo, Tanzania, and Uganda have sought to clarify population movement patterns to improve their disease surveillance and pandemic response efforts. Ministry of Health–led teams completed focus group discussions with participatory mapping using country-adapted Population Connectivity Across Borders toolkits. They analyzed the qualitative and spatial data to prioritize locations for enhanced COVID-19 surveillance, community outreach, and cross-border collaboration. Each country employed varying toolkit strategies, but all countries applied the results to adapt their national and binational communicable disease response strategies during the pandemic, although the Democratic Republic of the Congo used only the raw data rather than generating datasets and digitized products. This 3-country comparison highlights how governments create preparedness and response strategies adapted to their unique sociocultural and cross-border dynamics to strengthen global health security.

air travel and movement over land and water helped drive the international spread of SARS-CoV-2. National and local government agencies, global public health partners, and private sector stakeholders implemented various border health mitigation measures, which included screening at international point of entry (POE) locations and at domestic point of control (POC) locations in communities and priority locations along travel routes to limit the spread of COVID-19 (2).
Although data such as volume and destination are available for formally documented travel by plane or cruise ship, informal traveler movement (i.e., by private conveyance or across porous borders) provides less data for analysis and decision making. Scientists and public health practitioners continue to advance the use of social media and mobile phone data to understand mobility (3). The products from these analyses are very informative, but the capacity to create them is often not available in the areas affected by the movement. This deficiency of data or access to advanced analytic methods on international mobility limits the capacity of public health authorities to build strategies adaptable to the unique risks of disease translocation within and between countries.

The US Centers for Disease Control and Prevention (CDC) developed the Population Connectivity
The COVID-19 pandemic spread between neighboring countries through land, water, and air travel. Since May 2020, ministries of health for the Democratic Republic of the Congo, Tanzania, and Uganda have sought to clarify population movement patterns to improve their disease surveillance and pandemic response efforts. Ministry of Health-led teams completed focus group discussions with participatory mapping using country-adapted Population Connectivity Across Borders toolkits. They analyzed the qualitative and spatial data to prioritize locations for enhanced COVID-19 surveillance, community outreach, and cross-border collaboration. Each country employed varying toolkit strategies, but all countries applied the results to adapt their national and binational communicable disease response strategies during the pandemic, although the Democratic Republic of the Congo used only the raw data rather than generating datasets and digitized products. This 3-country comparison highlights how governments create preparedness and response strategies adapted to their unique sociocultural and cross-border dynamics to strengthen global health security.
Across Borders (PopCAB) toolkit as a resource for governments and other stakeholders to gather and analyze data about population mobility to inform public health interventions (4). In brief, the toolkit provides template guides for focus group discussions (FGDs) and key informant interviews (KIIs), considerations for developing the base maps for participatory mapping, template materials, and techniques for managing and analyzing the data, and training materials on methods for preparing for, implementing, and applying the data from PopCAB activities. CDC can provide the toolkit to countries and partners along with technical assistance, as interested, throughout the process.
The Democratic Republic of the Congo (DRC), Tanzania, and Uganda ministry of health (MOH) offices sought to develop COVID-19 border health surveillance and mitigation measures better adapted to their unique community connectivity and population mobility patterns (5,6). To address its goal, each MOH, in collaboration with partners, implemented the Pop-CAB toolkit after adapting the template materials for FGDs and KIIs with participatory mapping to their country context (7). These community engagement activities and their associated analyses provided the implementers with a better understanding of population movement patterns. The countries applied the information to improve COVID-19 surveillance, testing, and border health policies. These COVID-19 response-focused PopCAB activities built on previous PopCAB efforts implemented in all 3 countries during the 2018-2020 Ebola virus disease epidemic in eastern DRC (8,9). We identified lessons learned and best practices by comparing the PopCAB initiatives the 3 MOHs implemented during the COVID-19 pandemic and how they applied the results.

Methods
During May 2020-March 2022, MOH-led teams in DRC, Tanzania, and Uganda used the PopCAB toolkit to inform COVID-19 response strategies. As part of PopCAB, these teams, in collaboration with CDC and implementing partners in DRC and Uganda, conducted FGDs and KIIs with spatially-accurate participatory mapping to gather information about community-level, domestic, or cross-border population movement and connectivity patterns.
To implement PopCAB, teams completed actions across 4 phases: preparation, characterization, visualization, and application. During the preparation phase, the team identified objectives and priority geographic areas or population groups of focus, adapted the Pop-CAB materials to address the objectives and context, worked with community leadership structures to identify stakeholders to invite to participate in FGDs and KIIs, and defined the timeline of activities. During the characterization phase, the team implemented as many PopCAB FGDs or KIIs as were needed to gather, process, and consolidate qualitative and spatial data. Depending on project objectives, the team could plan sessions around a specific event, such as a religious festival, or an important temporal rhythm, such as weekly during harvest season. To create the spatial dataset from the annotated maps and locations mentioned in the FGDs and KIIs, the teams geocoded each location of interest (LOI). LOIs represented origins, destinations, or locations along domestic and international travel routes, such as markets, health facilities, border towns, or transit towns. In the visualization phase, the team used the data from the characterization phase to identify, analyze, and visualize population movement and connectivity patterns by creating maps and narrative reports that illustrated population movement with respect to the LOI. Finally, the team used the application phase to adapt and improve programs and strategies with the results. Teams repeated these phases to address evolving public health needs, ensuring that they regularly identified opportunities to share experiences and develop plans to improve adapted materials and implementation approaches.

Analyses
We compared the approaches that teams used to design and implement PopCAB. We compared, by country and PopCAB phase, details about the implementation timelines, team compositions, project goals, priority geographic areas or population groups of interest, data collection strategies, analytic approaches, and the application of results. To consolidate information for the comparative analysis, we gathered qualitative, quantitative, and spatial data from project implementation materials and outputs and qualitative data from discussions with the teams. Here, we intend to present a broad overview of PopCAB results from each country, rather than specific details.

Results
During May 2020-March 2022, teams implemented 94 PopCAB events to inform COVID-19 response measures at binational, national, and local government and POE levels (Table 1; Figure 1): 8 in DRC; 24 in Tanzania; 60 in Uganda; 1 binational between DRC and Tanzania in Kigoma, Tanzania; and 1 binational between DRC and Uganda in Kampala, Uganda. Two of the 8 PopCAB events DRC implemented were also binational and conducted during cross-border meetings with Angola (in Luanda, Angola) and the Republic of Congo (in Kinshasa, DRC). Overall, the PopCAB participants in the 3 countries identified >2,000 unique LOIs through those discussions or associated annotations on the base maps.

Preparation Phase
The national MOH port health director (in Tanzania and DRC) or border health focal point (in Uganda) led the PopCAB teams; their overall goal was to gather information about cross-border  (2), business women specifically (1), community leaders (2), dhow (wooden boat) operators (1), fishermen (2), healthcare providers (2), pastoralists (4), pastoralists and cattle traders (1), peasants (1), petty traders (1), salt producers (1), security officers (1), tour guides (1) All FGDs were implemented at the subdistrict or ward level; Discussions focused on population movement into, through, and out of the administrative level 2 unit where the PopCAB event was being held
In contrast, Uganda and Tanzania implemented PopCAB-focused initiatives to conduct many FGDs and KIIs in a short time. For example, Uganda implemented 30 events in Ntungamo, Isingiro, and Masaka districts near the Tanzania border in May 2020 after a disproportionate increase in COVID-19 cases in the region. Tanzania implemented 24 events in Arusha, Pwani, and Tanga Provinces in July 2020, recognizing the need to adapt COVID-19 response measures for continued cross-border movement and travel. Throughout the pandemic, the DRC border health director incorporated PopCAB events, or orientation to PopCAB, into all cross-border meetings the program joined or hosted. The DRC, Tanzania, and Uganda MOHs shared this accomplishment during bilateral meetings in Uganda with DRC in September 2021 and in Tanzania with DRC in March 2022 (Figures 1, 2).
Across these events, the teams in all 3 countries implemented PopCAB to address a few consistent objectives. One focused on a priority to adapt strategies for surveillance and risk communication at POEs and POCs to limit the spread of COVID-19 across international and domestic administrative boundaries. A S108 Emerging with persons representing a variety of occupations and responsibilities relevant to a broader geographic scale, such as a district or region. Only Uganda implemented KIIs to gather more information from leadership or to address challenges with gathering multiple representatives for a group of interest, such as traditional healers, security personnel, and medical staff.

Visualization Phase
The countries' teams developed narratives and reports that listed all the LOIs and routes. The reports also described themes from the informal qualitative analyses completed by those who directly conduct-ed the discussions or recorded notes. For example, these reports provided details about cross-border travel patterns to seek care from traditional healers, travel routes community members took when seeking healthcare to conceal residence status, or routes to avoid lockdown policies. The teams included photos of the annotated base maps and photos of participants during the events. Each team followed a different approach to managing and analyzing the gathered information. The DRC team completed detailed reports rapidly, within 1 week, after each PopCAB event; however, they did not develop qualitative or spatial datasets of LOIs or S110 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 28, No. 13, Supplement to December 2022  routes for future visualizations and analyses. They took this approach because they intended to focus on applying results immediately after the discussions and their staff availability to process and manage the data over time was limited.
The Tanzania team developed detailed FGD transcripts after each event to facilitate their ability to integrate and analyze results across multiple Pop-CAB activities. They also developed, in collaboration with CDC, summary tables listing all annotated LOIs. The Tanzania team worked closely with CDC team members on qualitative data analysis, use of geographic information systems, and cartography to learn new skills in analyzing the qualitative data and geocoding the LOIs. These dedicated data management and analysis efforts led to detailed, formal spatial datasets of the LOIs and routes and initial qualitative thematic analyses. Although a compiled report from each event or group of events took longer to create, the team could use those results to develop improved visualizations for already-defined and future program goals.
The Uganda team focused on developing detailed FGD and KII transcripts after each event, along with detailed summary tables of all mentioned and annotated LOIs and travel routes. This process was supported by dedicated team members who had expertise in qualitative and spatial data management. The team analyzed the qualitative data for themes about reasons and routes for movement. The spatial analysts and cartographers geocoded the results, building robust spatial datasets of >1,000 unique LOIs and routes. The team also combined the results with other data-gathering initiatives completed by partners on the team, including Infectious Diseases Institute and Baylor Uganda, to characterize crossborder healthcare-seeking behaviors.
The data management approaches of Tanzania and Uganda led to the ability to develop more broadly effective reports and presentations. They were able to include visualizations to address various MOH objectives identified during the preparation phase or newly identified during pandemic response initiatives.

Application Phase
All teams used PopCAB results to adapt national, district, and POE-level policies, programs, and resource allocation plans (Table 5). All teams enacted these adjustments using qualitative and spatial information mentioned during the PopCAB events. Despite the different strategies implemented in the visualization phase, all teams continued to use the results from completed events throughout the pandemic. The Tanzania and Uganda teams completed multiple analyses to respond to established and newly identified goals throughout the response, stemming from the approaches they followed to develop robust datasets with the gathered data.
The teams implemented various contextuallyspecific initiatives using the results to address consistent and unique objectives. The DRC MOH used Pop-CAB results from FGDs in and around Kinshasa to identify 3 urban locations for new mobile COVID-19 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 28, No. 13, Supplement to December 2022 S111  surveillance to increase the ability to detect illness at key traveler congregation points. In addition, the DRC MOH maintained a plan to routinely conduct PopCAB in Kinshasa to guide when and where to adjust the locations of those mobile urban surveillance sites. They also rapidly applied results to extend operating hours for certain POEs to accommodate unique movement patterns identified through the FGDs. The Tanzania Port Health unit identified areas for increased engagement with village committees and security authorities to strengthen border health surveillance. As part of that effort, they identified community health workers in areas with increased cross-border connectivity in Dar es Salaam and Tanga Provinces and provided them with additional training on event-based surveillance. The MOH also used the data to select high-traffic locations where they enhanced community outreach and installed handwashing stations.
District-level officials in Uganda on the Tanzania border worked with owners of nighttime bars visited by persons from across the border to increase COVID-19 surveillance. The Uganda team also identified mobile phone market vendors that serve cross-border communities to support them in distributing COVID-19 risk communication materials to high-priority population groups. Along that border and the western border with DRC, the Uganda team applied results to identify schools and markets preferred by cross-border communities for enhanced risk communication in preparation for and during the COVID-19 response lockdown. Like Tanzania, Uganda also applied results to identify village health volunteers who worked in areas with increased cross-border connectivity, including those along routes used by persons fleeing conflict from DRC into Uganda, for prioritized refresher training on community-based surveillance.
All of the teams applied the results from the binational PopCAB events during cross-border meetings to prioritize official and unofficial POEs and other LOIs along travel routes across their shared borders for enhanced risk communication and traveler surveillance. They also prioritized groups of contiguous administrative areas in cross-border surveillance zones for more robust and sustained collaboration and information sharing.

Discussion
Throughout the COVID-19 pandemic, the DRC, Tanzania, and Uganda MOHs, in collaboration with CDC and other partners, adapted bilateral, national, and local-level strategies to their complex, crossborder sociocultural contexts by gathering and analyzing community-level information on population movement patterns. Although the countries' MOHs developed implementation goals and plans independently, all 3 followed consistent approaches for developing multisectoral collaboration for participation and applying the results. However, they differed in the intensity of data management and analysis methods, reflecting varying availability of resources including staff time and expertise. Those unique analytic approaches influenced the magnitude of tabulations of locations and routes and qualitative analytical results compiled in formal datasets. Regardless of the depth of analyses, the countries, each with unique COVID-19 epidemiology, border health and public health policies and infrastructure, and cross-border dynamics, provide various examples of ways to incorporate population mobility, a key driver of communicable disease spread, into mitigation measures. Despite differences in data compilation and analysis, their approaches highlight opportunities to achieve impacts across varying staff and financial commitments for creating qualitative and spatial databases.
The teams' experiences revealed challenges with implementing PopCAB overall and during a pandemic. Although the resources needed to implement any one PopCAB event is low, requiring only field travel support and person-time from the implementers and participants to complete the 1.5-hour event in addition to a printed map and supplies such as markers, pens, and paper, leadership had to secure additional funding and persontime to train staff on the toolkit and to process the gathered data. In addition, the teams had to adhere to COVID-19 mitigation measures preventing in-person trainings and field events at different times throughout the pandemic. To address these considerations, the teams designed and employed online training techniques. They also ensured that previously-trained staff participated in subsequent PopCAB implementation over the 2 years with minimal refresher training, introducing a few new staff at a time rather than entirely new field teams. The teams also adjusted field-based protocols to incorporate COVID-19 mitigation measures during FGDs and KIIs, including physical distancing, with only the facilitator annotating the maps, and use of cloth face coverings. Also, highlighted by the DRC MOH's approach, teams adjusted expectations for data processing and analyses to accommodate the available resources.
The teams were not able to attribute trends in COVID-19 epidemiology in their countries to changes they made to mitigation measures and policies using PopCAB data because of the complexity of SARS-CoV-2 translocation among mobile populations and difficulty differentiating travel-associated spread from domestic transmission. However, national MOH leadership overseeing these PopCAB activities expressed confidence that their interventions more appropriately supported their communities because they were adapted to the unique, multisectoral and sociocultural environments and were designed through community-engagement efforts. In addition, district-level leadership participated in the initiatives ensuring continuity of these efforts in local programming and resource allocation decisions.
The 3 governments in East and Central Africa implemented community-engagement efforts using the PopCAB toolkit following various staff-time and data management approaches over 2 years of the pandemic to design COVID-19 mitigation strategies. More specifically, they adjusted public and border health policies and programming to address formal and informal cross-border movement patterns, to enhance surveillance and capacity building at newly identified community-based locations and healthcare facilities, and to strengthen cross-border collaboration between neighboring countries. Those MOH-led, community-based initiatives can complement other analytic methods using existing travel and mobility data to incorporate community dynamics more accurately in border health and other preparedness and response strategies for COVID-19 and other communicable diseases. Furthermore, the MOHs can continue to apply the results to other public health goals, including broader outbreak preparedness strategies and cross-border collaboration priorities. Their experiences reveal options government leadership can follow to integrate population mobility and sociocultural factors into public health preparedness and response strategies.