Hyperlocal geospatial data to guide COVID-19 vaccination in Senegal

By Parysa Oskouipour, Derek Pollard on June 1, 2022 in COVID-19, GIS, News

Senegal is hardly alone in experiencing challenges with their COVID-19 vaccine roll out. Like many countries, much of the challenge has been related to hesitancy and demand. Recognizing that having ready access to COVID-19 vaccination-related behavior indicators can inform a better understanding of why coverage is not reaching saturation and to whom additional resources and efforts should be directed, Akros (in collaboration with Fraym and GRID3) worked with Senegal Ministry of Health and Social Action (MSAS) departments and in-country partners to build a custom geospatial dashboard that demonstrates these data.

Hesitancy and lack of demand issues for vaccines can stem from a number of causes. For example, rural communities far from health facilities administering vaccines may require significantly extra effort and expenses to travel to get the intervention, resulting in a demand challenge. The response to this demand-driven low-coverage problem will be different than hesitancy-driven challenges and the data required to respond effectively is likewise different. In this example, understanding: 1) Which communities are farther than a reasonable traveling distance to the health facility, 2) where exactly those communities are, and 3) how many people are expected to be found there, is valuable information that can be used to maximize the chances of a successful response. However, access to these types of granular geospatial demographic and health data that promote this level of evaluation to allow progress against such bottlenecks, has not been widely available or accessible to staff needing to make critical resource prioritization decisions.

The solution—hyperlocal geospatial data for COVID-19 vaccinations

With vaccination hesitancy and demand as the major challenges to achieving higher coverage, the data prioritized for this dashboard were proxy indicators for providing more insight into these challenges. Further, in order to decentralize decision making and empower district and health facility staff to make decisions that drive up coverage, this data was made accessible through the dashboard at 1km x 1km cells that can be aggregated up to health facility and district-level indicators. The dashboard interface allows this hyperlocal data to be, quickly and easily, geospatially explored before downloading for further analysis or input into external planning tools.

This work was built upon a history of collaboration among these partners — which has had success in providing detailed microplanning services utilizing granular spatial data to government malaria and neglected tropical disease (NTD) programs, with demonstrated examples for malaria in Zambia, Nigeria, and Senegal and for NTDs in Rwanda and Kenya. Building on that technical capacity, the dashboard in Figure 1 (showing COVID-19 vulnerable populations against health facility catchment areas) was built to display modeled COVID-19 vulnerability data to enable more informed decisions within vaccination planning workflows.

The geospatial dashboard allows users to filter out key COVID-19 planning data at a granular level.

The geospatial dashboard consolidated a wide variety of data and relevant COVID-19 vulnerability and risk models into the visualization to be filtered by region, district, and health facility to inform all levels of health planning. Largely using demographic and health surveys, the data includes statistically sound high-quality, geo-tagged household survey data, satellite imagery-derived data products, health metrics, and health infrastructure. This hyperlocal data, down to 1km grid cells, allows for the visualization of the spatial distribution of priority groups and classifies individuals within priority groups using WHO-guided indicators of vulnerability. These include elderly population groups and groups that receive a high vulnerability score generated within the COVID-19 vulnerability model. Other COVID-19 indicators within this model included vaccine allocation, exposure, co-morbidities, information access, prevention activities, and vaccination likeliness — all of which were able to be filtered, displayed, and extracted for all levels of the health administration hierarchy to inform microplanning.