Addis Ababa — Ethiopian and Norwegian researchers have developed a mathematical model that can identify conditions that increase the likelihood of a malaria outbreak up to two months ahead of its occurrence.
The computer model, Open Malaria Warning (OMaWa), incorporates hydrological, meteorological, mosquito-breeding and land-use data to determine when and where outbreaks are likely to occur.
Torleif Markussen Lunde, one of the model's developers and a researcher at Norway's University of Bergen, told SciDev.Net that the model made direct use of the limited real-time information available in typical rural areas.
"The model also reproduces observed mosquito species composition in Africa. It is the first time this has been done with a biophysical model. We are now looking at which areas in Africa the model can be applied," he said.
Lunde said that past attempts at predicting malaria epidemics have had limited success because "some models [were] oversimplifications of the reality, and might have led to problematically high or low sensitivity to changes in the environment".
Predictions made by the model compared favourably with observations from field trials and health clinics, the researchers said.
However the model needs to be tested during a significant malaria outbreak, and its outputs compared with case studies and field observations, according to Bernt Lindtjørn, professor of international health at the University of Bergen and a co-author of the paper.
"It is [also] specific to African mosquitoes and may require modification before being applied outside Africa," he added.
"Our model is not only a tool for predicting malaria, but can also be used to understand the dynamics of malaria transmission," he added, noting that the tool could be used to better understand the effects on a malaria outbreak of interventions such as residual spraying and bednet use.
Daniel Argaw, of the World Health Organisation in Ethiopia, said that "the development of a model that can predict malaria outbreaks will have a significant role in combating malaria," adding that no other models have been developed for this purpose.
The research was published in Forecasting Malaria in April.