MalarInfo: A machine learning information toolset for predicting malaria risk, rates and trends in intervention planning: A pilot project in Zambia

This project aims to translate the methodology and statistical models developed at Ulster for Malaria risk and rate trends into a practical near real-time (pilot) tool.


  Project Funders

Ulster University Internal GCRF Funding Pump Priming

  Funding Amount

£37,125

About the Project

Malaria causes over 400,000 deaths per annum and in excess of 200 million cases worldwide, with over 90% occurring in African countries. While mortality from malaria continues to decline globally, incidence rates in many countries are rising.

This project aims to translate the methodology and statistical models developed at Ulster for Malaria risk and rate trends into a practical near real-time (pilot) tool that can be used for better prediction of malaria infections to improve the monitoring, planning, targeting and forecasting of intervention needs and health impacts.

The system will provide predictive analytical capabilities of malaria in monthly and quarterly time periods at the smallest spatial scale. As health facility level malaria data improves in quality, the use of fine scale predictions will provide better decision making for policy, resource mobilisation and interventions.

The project will evaluate the potential value of such a system for malaria monitoring, planning and early warning with a view to further development and deployment.

The system will be developed in collaboration with:

  • National Malaria Elimination Centre (NMEC) in the Zambian Ministry of Health
  • Macha Research Trust (MRT)/Malaria Institute At Macha (MIAM) in Choma, Zambia
  • Malaria Research Institute in the Bloomberg School of Public Health
  • Johns Hopkins University, USA
  • National Centre for GeoComputation in Maynooth University in Ireland

The system will be tested in the Choma District of southern Zambia with support from the Macha Research Trust (MRT)/Malaria Institute At Macha (MIAM).

Projected key outcomes

  • Translation of the methodology and statistical models developed at Ulster for Malaria risk and rate trends into a practical near real-time (pilot) tool that can be used for better prediction of malaria infections to improve the monitoring, planning, targeting and forecasting of intervention needs and health impacts
  • Development of new machine learning capabilities, particularly predictive analytics capabilities, to enable better decision making for policy, resource mobilisation and the implementation of interventions in relation to malaria
  • Integration of an enhanced up-to-date national malaria dataset with existing and new models
  • An evaluation of the potential value of such a system for malaria monitoring, planning and early warning