Data Sources for Health Facility Lists and Attributes

Health facility location data for countries are available from a number of sources. All sources have weaknesses and it can be difficult to establish which are the best for a particular task.

1. Hosted by country Ministries of Health or equivalent

2. A collated dataset for sub-Saharan Africa (Maina et al., 2019)

3. healthsites.io and OpenStreetMap

4. WHO Health Resources and Services Availability Monitoring System (HeRAMS)

5. Common Geo-Registry

6. Humanitarian Data Exchange (HDX)

7. Google maps

8. Humanitarian organisation internal use (e.g. MSF)

The table below outlines some of the strengths and weaknesses of the different sources.

Data source	Strengths	Weaknesses
Ministries of Health	-definitive source -more likely to be up-to-date	-not available for all countries (only 7 out of 52 in Africa can be downloaded for analysis) -can be hard to find -often few attribute data -different formats -can be out-of-date too -licensing uncertain
Maina et al. 2019 for sub-Saharan Africa	data cleaned and checked -available for 50 countries -consistent format	-out-of-date (collected 2012-18) -not maintained -just SSA -just public facilities -few attribute data -licensing uncertain
healthsites.io and OpenStreetMap	-consistent format and API access -open for improvement -clear roadmap for improving system -potential to store attribute data -good machine readability -clear licensing	-incomplete -few attribute data -data must be licensed odbl to include
WHO HeRAMS	-definitive source -good attribute data -actively maintained -active involvement local staff -good visualisation tools	-data cannot be downloaded except with authorisation -not all countries
Common Geo-Registry	-can improve storage of MFLs for countries -efficient systems for data checking and improvement	-not yet available -targeted at data owners rather than users -will not guarantee availability of data to users
Humanitarian Data Exchange (HDX)	-well known in humanitarian community -good hosting option when others not available -can host data under different licenses	-can be multiple lists per country -dates may represent when data were uploaded rather than collected -unclear data checking -poor machine findability -different formats
Google maps	-likely to be up-to-date -include private and public facilities	-data not available for download and use
Humanitarian organisation internal data (e.g. MSF)	-maintained by data users	-data not available to others

A recent outline and comparison of different sources of health facility location data for countries in Africa can be found in South et al. (2021).

South A, Dicko A, Herringer M et al. A reproducible picture of open access health facility data in Africa and R tools to support improvement [version 2; peer review: 3 approved, 1 approved with reservations]. Wellcome Open Res 2021, 5:157 (https://doi.org/10.12688/wellcomeopenres.16075.2)

1. Google and OSM have been identified as potential reliable sources providing a global scope of health facilities lists (a mixture of public and private facilities).1 These datasets are frequently updated with robust quality controls. OSM is designed to create a free and editable geospatial database of the whole world, built by a large user community that uses GPS devices, aerial imagery, and field maps to verify that the data is accurate and up to date. Studies conducted to assess the overall quality of OSM data agree that dense areas with high numbers of OSM contributors have higher accuracy and completeness compared to rural areas. 1,17

2. The Global Healthsites Mapping project (https://www.healthsites.io/ ) focusses on baseline health facility data, thus providing a domain specific view of OSM. This open data approach invites organisations to share health facility data to OpenStreetMap and collaborate to establish an accessible global baseline of health facility data. 18

3. A health facility master list was published in HDX in February 2020 (although data was collected from 2012 - 2018). This list was developed from both government and non-government sources from 50 countries in Sub-Saharan Africa and provides a comprehensive spatial inventory of 98,745 public health facilities. Each data record represents a health facility and has eight descriptive variables. This geocoded master facility list has been made publicly and freely available through the World Health Organization's Global Malaria Programme. 19

4. The Common Geo-Registry is an IT solution that serves as a single source of reference information for the standardization, management and use of geographic data. It contains a mix of mandated (MoH approved) and non-mandated (from NGOs) lists of health facilities clearly indicated within this database. Additionally, it allows for other databases to utilize its signature domain to create service domains that fit a purpose or context (See Appendix 6)