Kenya develops tool to predict malaria

The Kenya Medical Research Institute has launched a tool aimed at predicting malaria outbreaks in any area of East Africa two to three months before they occur.

A man sprays insecticide under the eves of a house in Mwea district, Kenya. A new tool aims to predict malaria outbreaks and will help target spraying to increase its effectiveness.

This article first appeared in Reuters AlertNet, a humanitarian news network that aims to keep relief professionals and the wider public up-to-date on humanitarian crises around the globe.

By Isaiah Esipisu
---------------------------------------------------------------------------

In collaboration with scientists from the Kenya Meteorological Department and the International Centre for Insect Physiology and Ecology, the institute has designed a scientific model that uses weather predictions, information about the reproductive mechanisms of mosquitoes, and data on geographical formations of particular areas to predict surges in malaria.

"Rainfall (and) temperatures can be used to explain up to 80 percent of statistical variation in malaria incidences. This is because the temperature variations are extremely important in breeding of mosquitoes. That is why involving the weatherman to predict the level of expected temperatures and the expected amount of rainfall is extremely important for this model to work accurately," said Dr Andrew Githeko, a malaria expert and one of the lead researchers on the project.

So far, the model has worked effectively in tests in western areas of Kenya, including Nyanza province, Western province and the Rift Valley province, as well as in Tanzania and Uganda.

"We have been trying the model for the past nine years in the three countries. We used the platform of the 1997 El Nino rains, the 2003 long rains and the 2006 long rains, where the model was able to predict malaria outbreaks in hundreds of sites, where indeed the outbreaks struck," Githeko said.

Accuracy Range Is 86 To 100 Percent

The most impressive test, he said, was in Kakamega district found in western Kenya, where the model worked with 100 percent accuracy in all nine years of the trial period. Areas like Nandi and Kericho, in Kenya's Rift Valley province, were predicted with 86 percent success.

The model worked with 90 percent effectiveness in all three countries overall, he said.

So far, despite ongoing rains in Kenya, the model has predicted there will be no malaria epidemic this season in Kakamega, a normally endemic malaria area, because temperatures are very low and don't favor mass breeding of mosquitoes.

"Though most of the places where the model was tested are malaria endemic areas, we have factored in even the highland areas because research has shown that the disease is slowly infesting highland areas due to the looming climate change," Githeko said.

According to the researchers, heavy rains linked to the El Nino climate phenomenon have caused the appearance of springs in highland areas, which produce clean water that is suitable for mosquito breeding.

In 1990, for example, significant numbers of malaria cases began appearing in the highlands of East Africa.

"During that time, it was not clear what was causing the epidemic. But, as scientists, we believed that climate variability had something to do with it, prompting KEMRI (the Kenya Medical Research Institute) to propose for a study after the repeat of the same (problem) during the 1997-98 El Nino rains," said Dr John Githure, KEMRI's director.

When malaria strikes such highland areas, it can cause severe health problems. Because most of the residents have not been highly exposed to the disease their immunity against it is poorer than that of people living in areas with a high incidence of malaria.

Research shows that this kind of malaria has been on the increase in East Africa and is an emerging climate-related hazard that needs urgent attention. Malaria incidence increased by 337 percent during the 1987 epidemic in Rwanda, studies show. In Tanzania, Uganda and Kenya, records indicate that it increased by 146 percent, 256 percent and 300 percent, respectively, during and after the extreme rains of 1997-98.

Modeling Helps Target Spraying

To avoid malaria outbreaks in highland areas, where it kills more people than in low-land areas, East African governments have been depending on indoor spraying of long-lasting pesticides when long rains that might lead to an outbreak are anticipated.

But with the new model now in place, spraying can be done only when the model suggests an outbreak is imminent. Experts say that spraying at the right time also reduces the chances of mosquitoes building resistance against the insecticides.

The disease prediction tool should also help policymakers and health officials prepare in time to deal with looming outbreaks.

Malaria was selected as the first disease to study because of its severe effects but the model might also be adapted to other diseases, the report's authors said.

KEMRI now has a special unit to carry out research on the interactions between climate change and human health variability, Githure said.

Even before it was proved effective, the new model for malaria prediction was one of the tools selected by the U.N. as an example of practical adaptation to climate change. The U.N. has strongly suggested the model be further developed.

The project was funded by the Climate Change Adaptation in Africa program, Canada’s International Development Research Centre and Britain's Department for International Development.

Isaiah Esipisu is a science writer based in Nairobi.

Published: 31 Aug 2010

Contact details:

Head Office
Mailing address PO Box 8500, Ottawa, Ontario, Canada K1G 3H9

Asia
208 Jor Bagh, New Delhi 110003, India
Phone: (+91-11) 2461-9411

Latin America and the Caribbean:
Juncal 1385, Piso 14
11000 Montevideo, Uruguay
Phone: (+598) 2915-0492

Middle East and North Africa 
Zahran Gate Complex Suite 302,
25 Ismael Haqqi Abdo Street,
Amman, Jordan

Sub-Saharan Africa
PO Box 62084 00200, Nairobi, Kenya

Street address: Eaton Place, 3rd floor
United Nations Crescent, Gigiri, Nairobi, Kenya
Phone: (+254) 709-074000

+1 613 236 6163
Country: 
Journal:
News topics: 
Content type: