In partnership with the Mozambique Ministry of Health (Ministério da Saúde), the U.S. Health Services and Resources Administration (HRSA), and the U.S. Centers for Disease Control and Prevention (CDC), ICAP worked with Dimagi to leverage the existence of an existing mobile health application called CommCare to test the use of a predictive algorithm to identify patients at higher risk of default. CommCare, and its Mozambique-specific version, Infómovel, was developed by Dimagi to support community-based HIV and TB services in Mozambique. The main front-line users of the app are facility-based case managers (gestor de casos) and community health workers (activistas).
The OpCon Mozambique project was designed to apply Machine Learning (ML) to predict the risk that individuals receiving HIV treatment will “default” – e.g., fail to return for a scheduled appointment – using Infómovel. Key project activities included: development of the ML algorithm; preliminary design of a user interface (UI) or “flag” within Infómovel; and rapid field testing to elicit user feedback.