“[The organization] shared some maps with us, and when we reviewed with our data, we realized some health facilities that were mapped had been closed and no longer in use. This highlights the importance of sharing data among organizations working on the ground.” - Key Informant Interview
When it comes to creating and utilizing health catchment area geospatial databases, data that is not shared is not useful to organizations working within the same communities, and more importantly to the community itself. A review of the methods for sharing data among stakeholders in the health and humanitarian sector raises serious concerns with data sharing policies which seem to encourage restrictions on access to data, further propagating data silos. However, organizations that have shared data with or used data from the OSM community are inherently bound by its open data policy, making it easier to synchronize and integrate OSM data in partnership projects to some extent.22 There is an urgent need for governments and non-governmental organizations to initiate a review of their data sharing policies ahead of their commitment to contribute or benefit from this collaborative on health catchment area databases as this is the linchpin to its success.
Get in contact to join this work stream
https://health-catchment-areas-working-g.gitbook.io/health-catchment-areas/data-sharing
Key Recommendations
At the national level, there is increasing advocacy for the establishment of National Spatial Data Institutes, an institutional framework to facilitate the production, standardization and sharing of geospatial data across sectors of society. This maximises use and reduces redundancy in the creation and use of geospatial information within the health cluster. 13
An improved collaboration between humanitarian and health sector agencies with commercial organizations is necessary to focus more on quality and consistency of data collected.
There is a need to integrate geospatial platforms with demographic data. This should be governed at country level to enhance continuity.
It is important for local communities to take ownership of their own data (through generation, validation, updates and use cases). Engaging local populations is an efficient method to quickly gather critical information where geospatial data is lacking.
There is a need to move to health boundary layers below the health district. Machine learning can be integrated into existing strategies to automate the identification of villages followed by field mapping and validation. This could also improve identification of missed villages and zero-dose populations.29
Methodologies should be flexible to allow for data collection, validation, or updates in both stable and unstable contexts.
Clear definition of data licensing
https://health-catchment-areas-working-g.gitbook.io/health-catchment-areas/key-recommendations