headline picture of project

Local Mapping for Ebola Emergency Preparedness

Paras Valeh




A method of gathering, cleaning and validating data for city-wide health centre locations and health administration boundaries to inform maps and surveillance dashboards within a two week period.


Goma, North Kivu, DRC


  • Smartphones (with cameras and GPS functionality)
  • Transportation
  • A printer
  • Access to local networks and knowledge


  • KoBo Collect tool from Missing Maps


  • Mapper familiar with smartphone use and GPS
  • Driver with local knowledge

Avoiding duplication

Part 1: Finding what already existed

Considering the timeframe, it was important for us to build on the experiences of our partners. The team gathered as much data as possible from partners and contacts. Some geographical health data did exist, but when we shared it with Ministry of Health (MoH) colleagues, we found that much of it was out of date, including, importantly, the administrative boundaries. The most important document we were given was a spreadsheet with the names of health centres in Goma (though without coordinates).

All location information is important in this first data scramble; from hand-drawn maps to spreadsheets with distances between places and lists of neighbourhoods without GPS coordinates.

Minimum viable data

Part 2: Working out what data needed to be collected

The more data you try to collect, the more time consuming the mapping becomes, so you need to be selective. We decided that at a minimum we needed to know where people lived and where they would go if they got sick. This meant the locations of health centres and health administration boundaries, known in DRC as Zones de Santé and Aires de Santé (health zones and health areas).

Part 3: Going analogue!

Most health zone facilities have maps on their walls, often hand drawn, depicting their catchment areas. We visited the health zone medical supervisors and got permission to take photos of all their maps. These we sent to the Global Information Systems (GIS) team in the MSF OCA Public Health Department, where they were transposed into the OpenStreetMap platform and digitised.

Part 4: Setting up data collection

I knew from previous missions that KoboCollect would be a good tool for the data collection (it works even when offline and is really easy to use) and Jorieke, the Missing Maps project lead, sent me a link to the field mapping forms (which were available in French). We didn’t have long, so we used the data we already had to make a rough map of Goma, which we divided into sections we could cover in a day. Using this, I worked with our local drivers to plan the best routes to cover the city.

Part 5: Mapping with the team

I started mapping on my own, but soon realised I needed support. Through MoH colleagues, I found Gaston, a local guy with previous experience collecting data. We did a half-day of training and then mapped together for another half-day to make sure he understood the process. As Gaston didn’t own a phone, I borrowed one from a member of the MSF team, which he would pick up in the morning and give back each evening because as he didn’t have a reliable enough electrical supply to charge it. We met each morning for a briefing and in the evening for a debriefing. Gaston was paid a daily wage and transport allowance, and made his own way around the city.

When there were areas we couldn’t go to because of security, I downloaded KoboCollect to the phones of the local nurse supervisors and showed them how to use it. Getting them involved was also good because it created ownership and buy-in for the project locally. Each evening I uploaded the day’s data to the MSF server.

Find local mappers. It creates buy in and ownership, and improves the quality of the data.

Part 6: Data cleaning and validation

We worked closely with the GIS team to clean and validate the health centre and boundary data, removing duplications and errors. They then sent us back draft maps, which we verified or corrected with the help of our MoH colleagues.

The software that was used to clean and create the boundary data is called QGIS. You can learn QGIS yourself on our ‘GIS at MSF’ YouTube channel or through one of our MSF GIS courses offered by our L&D department.

Part 7: Conclusion

Unfortunately, my mission ended before the maps were professionally printed, so I never saw them in the flesh. Doing the mapping was a great way to build networks and engage local communities. It also felt good to do something for MSF that would have an ongoing benefit for the local community, as all the data was uploaded to OpenStreetMap by Missing Maps volunteers, meaning it is now freely accessible to anyone that needs it.

If we had had more time, I would have expanded the data collection to include private health centres and traditional healers as these both would play an important role if the outbreak hit Goma.

Since I left Goma, the data we collected has been used by MSF teams and included in maps made by the MSF GIS Unit and other partners responding to the outbreak. In August 2019, the first cases of Ebola were reported in Goma.

Share the results the data collectors and community. Having a printed (or digital) map in every health facility will serve more people in the community.

Would local geographical information help with emergency preparedness planning?

Do you know where local medical facilities are?

Is there good map data already available? A good place to check is openstreetmap.org.

Implementing this design requires review and coordination with HQ. Note that, at the time of going to print, OCA (Jorieke Vyncke, Missing Maps Coordinator & GIS Focal Point) has validated the contents of this feature. If you have any technical questions, please get in touch with your GIS Focal Point.