We Measure Everything We Do

Just as patients deserve medications that have been proven to work, so do communities deserve health systems that have proven efficacy. For Muso, rigorous measurement is our moral responsibility. Traditional healthcare systems measure their impact based on how they treat patients who walk through the door. For Muso, health systems should care as much about the patients who never make it to the door. We test our impact at a population level, to assess how the health of entire communities changes. We do this by conducting rigorous scientific studies with academic partners.

Learning in Real Time

In addition to conducting rigorous population research, we’ve built data feedback loops that help us iterate on our design and provide better care for our patients every day. We do this by analyzing data that community health workers, doctors, and other providers collect daily as they care for patients.
3,259,968Cumulative Home Visits

93Percent of Patients Treated Within 72 Hours

379,041Cumulative Clinic Visits

Three-Year Child Mortality Study

In 2013, the Muso model was put to the test in a study by Harvard, the University of California San Francisco, and the Malian Ministry of Health. The study team conducted a randomized household cross- sectional survey in Muso communities at baseline, 12, 24, and 36 months.


10x difference in child mortality

In the communities of Yirimadjo, the rate of under-five child mortality was 155/1000 at baseline. Three years after the rollout of Muso's health system, the rate of under-five child mortality was 17/1000.

Early Treatment Doubled

The percentage of children starting effective malaria treatment within 24 hours of symptom onset increased from 15% to 28%.

Healthier Children

The percentage of children sick with a fever decreased by 15%.

Patient Access to Care Rises 10x

After the launch of Muso's health system in 2008, the number of patient visits in the home and in the clinic increased tenfold.

Study Limitations

First, there was no control group, so no causal conclusions could be drawn. Second, the study tested the entire intervention as an integrated unit, and therefore could not draw conclusions on the relative contributions of each element of the intervention. These are crucial questions for future research.