When Muso launched our health care delivery in Yirimadio, a peri-urban area on the outskirts of Mali’s capital city, Bamako, in 2008, we took on a critical challenge to work in one of the fastest growing areas of one of the continent’s fastest growing cities. Vulnerable urban areas make up an under-recognized, quickly growing front line of the struggle toward health care for all. The African continent has been urbanizing more rapidly than other regions, and 56% of the population is on track to live in urban areas by 2050. Rapid population growth and population movement create challenges for planning sufficient clinics, health workers, vaccines, and medicines to serve everyone quickly. Measuring population proves particularly challenging. Even a costly, carefully done census proves out of date within a matter of months, as tens of thousands of people can move in or out of areas like Yirimadio in a single year.
Previously, Muso used population counts from the Ministry of Health bed net distribution campaigns in Yirimadio, which occurred every few years. Based on the rate of growth between multiple time point estimations from these campaigns, we were able to estimate that Yirimadio’s population had been growing at a rapid pace of 10% per year. In years without a bed net distribution campaign, we used the 10% per year growth rate to calculate an updated population estimate for each subsequent year.
However, due to resource constraints, Mali’s Ministry of Health did not conduct a bed net distribution campaign in Yirimadio this past year as previously planned, so it has been multiple years since Yirimadio had a full population count for a campaign. This old method for measuring population also carries important limitations, assuming a constant rate of growth in a dynamic area.
Together with our partners at Medic Mobile, we have come up with a solution for a more dynamic and accurate way to estimate the population served on an ongoing basis: multiplying the number of “active” households registered in the CHW App with the average (mean) household size. Every neighborhood of Yirimadio we’ve mapped into a zone is served by a specific Community Health Worker. That CHW registers every household they serve into their panel list on their CHW App. But, you may ask, how do we account for families that move out of the area? In the past year, we worked with Medic Mobile to create and deploy a new functionality within the CHW App, a household “muting” report that the CHW can submit to indicate when a family has left the area (temporarily or permanently) or when a family declines CHW services. The CHW can also submit a household “unmuting” report when the family has returned to the area. Using the number of active (non-muted) household registrations ensures that families that have moved or declined services are removed from the population count and from our monthly population coverage calculations. We then multiply the number of registered households by the average number of persons per household (which we measure in our annual, repeated cross-sectional survey of a representative sample of Yirimadio households) to arrive at the total population we serve in Yirimadio.
In arriving at this method, we considered and decided against a few other options. A brief review of the methods we did not use, and why:
Count of the number of sick patients we treat as total population served: We do measure and report on the number of sick patients we care for, both in the community and in the clinic. In many other health care systems, this passive way of counting patients served would be reasonable, as many health care systems measure their success based on how they do for the patients who walk through their door.
But there’s a big problem with this approach: the patients who never make it to the door, the patients who never reach a health care provider, the children who die in their homes without care, don’t count. These are the patients we care most about, the patients left behind by passive health care systems.
Muso strives toward a radically redefined moral mandate: we aim to hold ourselves accountable for population-level health indicators, the rates of access to care, disease, and death across the whole population of the areas we serve. In our urban areas, this means conducting an ongoing study that randomly selects and surveys households representative of the population as a whole. By calculating key indicators such as under-five mortality in this way, we hold ourselves accountable for how patients do regardless of whether they reach one of our health care providers. From this ongoing research, we have thus far published two studies.
Muso’s proactive approach spans both prevention and treatment, with the aim of finding and serving everyone. Prevention support and counseling that CHWs provide serve both the healthy and the sick. In addition, when Muso provides care without fees, we provide financial protection to the whole household, protecting household members against food insecurity and the resulting risk of malnutrition. This is why, to calculate the full population we serve, we use the average household size from the annual household survey and multiply by the number of active households.
The average household size for Bamako: The recent 2018 Demographic and Household Survey found that the average household size in Bamako was 6.4. By contrast, our annual household survey found an average household size of 5.08. This difference may seem small, but across 36,921 households, the DHS method would give us a population size of 236,924 for Yirimadio. By contrast, using our estimated average household size, we arrive at a population of 192,349 for 37,864 households. We decided to use the average household size from our ongoing Yirimadio study for two reasons: 1) because DHS’s figure is an average across of Bamako, while our study is restricted to the Yirimadio communities we serve and 2) because this same survey also serves as the source of our population health impact indicators, particularly under-five mortality rate.
It is quite possible that the new method we’ve adopted underestimates the population that Muso serves in Yirimadio. Indeed, our local government partners have indicated to us they believe this estimate to be a considerable underestimate. Nonetheless, we have adopted this conservative approach, because it enables us to track the population we serve dynamically in real time, and holds us accountable in a consistent way to reach every person. We will continue to pursue ways to improve these methods, and will keep you updated as we do.