GeoHealthAccess : making health services accessibility analysis accessible
In 2019, it was estimated that 50% of the world’s population lived without full access to essential health services. A key issue driving inaccessibility of health services is the time it takes to reach a health facility. The longer the travel time to health services, the lower the probability of seeking treatment.
To tackle this issue, it is essential to gather insights on how long it takes to get to health facilities. While this may seem like an easy question in the days of ubiquitous geolocation and navigation services, in many countries the travel time to health services is hard to estimate due to lack of data on road and health infrastructure. Computational approaches to estimating access to care have progressed significantly during the last decade. However, the amount of human intervention needed to tweak and run these models prevents up-to-date analyses from reaching decision makers on a regular basis.
Bringing insights on accessibility to health facilities through geospatial data
In 2021, Bluesquare and the Spatial Epidemiology Lab (SpELL), an academic research group from Université libre de Bruxelles with funding from Innoviris, collaborated to develop an open source solution that brings advanced insights on accessibility to health facilities in a regularly updated and interactive format: GeoHealthAccess (GHA).
GeoHealthAccess enables healthcare decision makers such as governments, NGOs, global donors, community-based organisations and local institutions to access up-to-date estimates regarding the accessibility of health facilities. Through interactive maps, users can look at the percentage of the population having access to a given health service in less than a certain amount of time.
The estimation of travel time is made using widely disseminated models. The innovation of GHA lies in its automated data acquisition through direct connection to routine health information systems and big earth data catalogs (see below). This means that accessibility calculations can be updated on a regular basis to reflect the availability of certain drugs, the seasonal variations of water levels or the opening of new facilities. Moreover, GHA provides interactive and easy to use visualisation to help non-specialists understand its results.
Making use of recent advancement on Combining geospatial modelling and Big Earth Data
GHA relies on recent advancement in geospatial modelling (digital simulation of the real-world using spatial relationships of geographic features) that can now be combined with Big Earth Data (big data associated with the Earth sciences).
For example, in recent years, we have seen the development of several open-access resources providing essential, high quality and detailed data including :
- Crowd-sourced geographic information from OpenStreetMap
- Spatial demographic/population data maps developed by WolrdPop or Facebook
- High-resolution land cover maps provided by Sentinel mission and topography maps
Combining this information with routine health data, Bluesquare and SpELL used spatial modeling to compute near-continuous estimates of accessibility. The GHA model is automatized, efficient, and relies mostly on open source data. This makes it not only free to use, but allows it to remain continuously up to date.
Enabling decision-makers to make informed healthcare decisions
As this tool is developed for healthcare decision-makers, its features were specifically defined to make GHA relevant and easy to use for them.
Up-to-date description of access to healthcare facilities
Estimated travel times can quickly change, due to an update in the road networks data, the creation and evolution of roads or the construction of additional health centers. GHA takes all of this into account by offering the most up to date view of access to health services thanks to its usage of the regularly-updated open source data like OpenStreetMap.
This way, the tool can support decision making, by performing fast routine analyses with up-to-date data.
GHA is preconfigured but adaptable to individual situations. It can be integrated with existing healthcare data systems such as DHIS2, which stores information on which health centers deliver which services on a monthly basis.
For instance, on the maps below, you can see the travel time to health centers when all health facilities have sufficient resources to provide basic malaria services and after the pandemic disrupted supply chains and facilities did not have the equipment or personnel to effectively treat uncomplicated malaria . Combining the modifications to travel caused by a lack of medical staff or medication with high resolution population data shows where the population has been most affected by those shortages.
Generation of valuable information in a user friendly interface
When designing the tool, Bluesquare and SpELL knew it would be essential that the results coming from complex statistical models run at country scale were coherent with the experience of users in the field. Therefore, we are working to ensure the results of GHA are validated with ground level data, by gathering insights on the patients who effectively visited the health centers.
And because this tool is designed with the end user in mind, GHA offers the results of its analyses in simple formats through an user-friendly interface.
How can you use GHA for your health programs?
GHA is a collaborative and open-source project, meaning it is accessible to everyone, for limited cost. After almost 2 years of work on the software, the modelling & analysis and the visualization features, SpELL and Bluesquare with funding from Innoviris are happy to announce that GHA is ready for routine use by health ministries and health actors. If you want to learn how you can install and use GHA to assess accessibility of health services, get in touch !
Our goal with this project is to provide governments and global health decision makers with a reliable tool to improve access to health services, a crucial step on the journey towards Universal Healthcare Coverage.