Analysis of impacts of climate variability on malaria transmission in Sri Lanka and the development of an early warning system
Background, Rationale and Problem Statement
Malaria is endemic in 101 countries and about 40% of the world’s population is at risk. In 1998, there were 273 million cases and 1.1 million deaths worldwide. In WHO's Southeast Asia region (which includes Sri Lanka), the case load was 16 million, with 73,000 deaths. Sri Lanka spends approximately 60% of its public health budget on malaria control. Malaria incidence in Sri Lanka has increased during the past 7 years. Plasmodium falciparum, which historically has been of low prevalence in Sri Lanka, has increased from 5% to about 25% of cases over the past decade and is increasingly resistant to the main anti-malarial drug, chloroquine. With an incidence rate of almost 12 per 1000 population Sri Lanka presently ranks as one of the most severely affected countries in Asia.
Sri Lanka has a history of malaria control dating to the 1920's, but is presently struggling to contain the disease because of population increase, large-scale human settlement in disease-endemic areas, rapid agro-ecological change, and altered patterns of population mobility. Malaria in Sri Lanka is unstable and fluctuates in intensity both spatially and temporally. Thus resources have to be spread to cover all potential risk areas, regardless of whether an outbreak will occur or not at a given point in time. This results in financial and manpower waste. Geographic and seasonal specificity of impending malaria risk will be particularly useful in communicating with environmental managers such as irrigation engineers who can use water management techniques to reduce mosquito breeding in pools in river beds. A major constraint to a more focussed approach to malaria control is the lack of a forecasting system.
While many factors play a role in the distribution of malaria and occurrence of malaria epidemics, climate is considered a major determinant. Temperature, rainfall, and humidity affect breeding and survival of vector mosquitoes and development of malaria parasites within the mosquitoes. Historically many epidemics have occurred during drought, as river margins retreat leaving numerous pools suitable for vector breeding, or in the season following a drought when rains return to normal. This second scenario, the post-drought’ epidemic often poses a major public health problem among populations whose vulnerability is heightened due to a period of poor nutrition associated with drought and lowered agricultural output. Sri Lanka has operated very effective malaria control in the past, however it has also suffered several major epidemics which have been triggered by climatic and hydrological anomalies. Recent evidence suggests that ENSO-associated climate variability influences vector borne diseases such as malaria. It has been known since the work of Rasmussen and Carpenter that there is distinct modulation of rainfall and temperature during ENSO events in Sri Lanka. An association between ENSO and malaria for Sri Lanka has been reported by Bouma and van der Kaay based on an analysis of monthly ENSO indices and fatalities from malaria. However, studies at finer temporal and spatial resolutions are needed to establish the mechanisms by which ENSO and other causes of climate variability may influence the transmission of malaria. The availability of an integrated GIS system which can model and inform on the seasonal and interannual dynamics of Sri Lanka’s climate and hydrology and its implications for malaria will be a very valuable control planning tool for the national/regional malaria control services.
Apart from the importance of malaria to Sri Lanka, there are several other reasons as to why this island is an excellent location to study predictive approaches to malaria: (a) There is an extensive literature on malaria in Sri Lanka starting from 1900; (b) Sri Lanka has a high degree of predictability of climate variability. The IRI forecast coverage and skill for Sri Lanka is among the best in Asia; (c) Sri Lanka has good disease and entomological databases (d) The climate data records for Sri Lanka is among the best in the tropics in terms of spatial resolution and historical extent starting from 1869; (e) Gridded climatic data for Sri Lanka at a resolution of 1-km are already available from IWMI and the Department of Meteorology.
The specific scientific and public health questions that this study will try to answer are the following: (1) What are the pathways by which weather and climate influences malaria in Uva Province? (2) What is the appropriate methodology to incorporate climate forecasts and remotely-sensed environmental data into predictive maps of malaria risk? (3) Are spatiotemporal forecasts of malaria risk sufficiently precise to be useful in a moderately endemic tropical setting? (4) Is malaria risk mapping technology the appropriate means to communicate predictions of malaria risk to public health workers, malaria control officials, irrigation managers and policy-makers and (5) What are the costs and benefits of introducing this technology associated with establishing an early warning system?