Health Facility catchment area refers to geographical area that covers the population that utilizes its services. Knowledge of health facility catchment area helps to calculate population-based rates of diseases and interrogate effects of socio-economic and environmental covariates on disease transmission. Additionally, facility catchment areas provide opportunity for public health authorities to track indicators such as vaccine distributions. In Malawi information on facility catchment areas is outdated, and mapping activities to generate catchment areas is an expensive activity that cannot be done periodically. Our team previously utilized a 1 square kilometer global friction cost surface generated by the Malaria Atlas Project to generate catchment areas for health facilities across Malawi. We used Least Coast Algorithm to assign the pixels of the friction surface to health facilities. While that this approach enabled us to generate catchment areas for health facilities across the country, we observed that this approach was inadequate at generating catchments areas for communities with high density of facilities. This challenge was observed in urban communities. Urban communities offer unique public health challenges particularly because of existence of huge inequitable in health care access and increased environmental hazards. Especially in Africa, the population of urban communities is expected to grow, with the majority living in informal settlements, most of them with limited sanitation and basic health care services.

In this work, we explored the potential of generating better health catchment areas using open data resources. We aimed at generating catchment areas for Blantyre city in Southern Malawi. To achieve this, we mapped all access roads to health facilities in Blantyre City by creating a collaborative mapping task on Humanitarian OpenStreetMap (HOTOSM) Tasking Manager. Using five randomly selected residential communities in Blantyre city, we observed travel speed associated with different means of transportation.

AccessMod version 5 was used for further processing of the generated data. Here the mapped roads were were merged with other open data resources including a 10-meter resolution land cover map for Blantyre city and 30-meter resolution elevation data. The merged layer was used to generate a 30-meter resolution friction cost surface. To generate catchment areas, a Least cost Algorithm was applied to generate facility catchment areas. We intend to validate the catchment areas using health records in three selected facilities within the city and government records from Health Surveillance Assistants. Furthermore, based on road distances to facilities, we quantified time the travel time to health facilities, a step that enabled us to identify locations that are relatively far from existing health facilities.

With the use of HOTOSM tasking manager, we have been able to map all all the possible access roads to health facilities in Blantyre (including footpaths). Also from the travel speed observations, we have noted that that travel speed varies with road surface, means of transportation and location. While we have observed that road users generally travel faster on tarmac road, than earth road, people who walk, travel faster on footpaths. Furthermore, our observations reveal that motorcycles exhibit similar speeds to vehicles on tarmac roads. However, on earth roads, motorcycles surpass other modes of transportation in terms of speed. Regarding the use of AccessMod, a 30-meter resolution friction cost surface was generated. This friction surface enabled generation of catchment areas of health facilities, and quantification of travel time to health facilities.

The study has demonstrated the practical usage of Open Data in understanding health care access in urban communities. The friction surface that has been generated from the study has potential to be used for studying physical accessibility of other basic services (i.e., security). The generated catchment areas, once they are validated they can be used by public health officials for planning and disease burden estimation.