Table 1.

Best practices and ambitions associated with defining reliable, accurate and representative service catchment areas for public services such as healthcare, education and social services

Category	Ambitions and best practices
Data collection	Improving collection and geocoding of residential address (village/estate) data from service users by healthcare providers, educational institutions, local governments and national statistical agencies. This will enhance the definition of service catchment areas for effective planning. High-resolution population density maps and databases of road network, land use/cover, travel barriers, care-seeking behaviour, modes of transport and travel speeds also need a careful curation.
Data and software sharing	Important data sets to define service catchment areas should no longer be kept in silos. The culture of making open-access data analytical models and software should improve across researchers and organisations in SSA. With increasing model sophistication, there is a need for software that can easily be used to define realistic service areas especially for planners.
Community	Building a community of researchers, sharing best practices, identifying difference between services, different diseases, service interruptions (eg, COVID-19 or natural disasters), ecological contexts and demography will be useful.
Service use	With a growth in availability of geocoded data and spatial epidemiologists across SSA, there is a need for an increased investment in research aimed at mapping higher resolution data on service use. Studies should also consider different forms of service access such as vaccination, healthcare, education and social care.
Disease mapping	The use of spatial statistics to map diseases, health outcomes, and demographic and socioeconomic indicators has witnessed huge advancements. However, the use of data from routine services (such as disease registries) together with reliably defined catchment areas requires more attention and quantifying the role played by different approaches and their impact on the mapped quantities.
Sensitivity	Where modelling must be conducted due to inadequate data, authors should test the sensitivity and uncertainty of several models that are used to define a service area. Comparisons will tease out if there any gains in using complex geospatial approaches in lieu of simpler approaches (more accessible to non-experts) to define service areas. There is also a need to recognise limitations such as bypassing the nearest provider due to personal preferences, quality and capacity when results are being interpreted.

SSA, sub-Saharan Africa.