Geoinformatics and predictive model for malaria risk in Thailand

Abstract

Malaria is an infectious disease caused by the parasite Plasmodium spp., and transmitted by an anopheles mosquito as a vector. After an infected female Anopheles mosquito bites a person, a parasite from salivary glands is delivered into the human’s blood flow. The incubation period of the disease is between 10-14 days and in the severe cases, it can cause jaundice, seizures, coma, or death. Many researches indicated that the occurrence of malaria requires many factors, including environmental factors, landscape and climate. Therefore, every healthcare organization delivered policies to control these typical factors towards human’s living styles. Though malaria has been cleared for decades, there is one serious problem that there were emergences of malaria parasite that resisted various types of medicines in use. So, it causes the complication in malaria control processes. To solve such complication, this study collected data mainly from the National Statistical Office database and applied statistical method, combining with the machine learning technique and Geographic Information System to determine risk factors, risk model, and risk map for malaria distribution in Thailand. As a result of these methods, this study proposed a new set of monitoring parameters, 28 parameters, a logistic model tree (LMT) risk model, and a risk map for malaria distribution derived by the machine learning mechanism. The precision and recall of this tree are 0.780 and 0.821, respectively, which are higher than using the risk model created from the typical factors. As a result of this finding, the government can utilize these newly identified factors to develop suitable strategies to control malaria considering proper criteria. For example, the government should control the quality of rivers or even increase the amount of river monitoring in some areas to reduce the malaria risk. Once these factors are completely managed with well-developed plans, each community can reduce the risk of malaria without concern regarding drug resistance. Moreover, according to the risk map, each area in Thailand has been defined as high, medium, low distribution of malaria based on the discovery factors. In such result, the government can set suitable the malaria monitoring and protection strategies for each individual area with a proper budget to be provided.