Philippe BarbazanSutee YoksanJean Paul GonzalezIRD Institut de Recherche pour le DeveloppementMahidol University2018-07-242018-07-242002-06-24Microbes and Infection. Vol.4, No.7 (2002), 699-705128645792-s2.0-0036286405https://repository.li.mahidol.ac.th/handle/20.500.14594/20204Despite the use of a variety of control strategies, dengue hemorrhagic fever (DHF) control is a major and permanent challenge for public health services in Thailand and in Southeast Asia. In order to improve the efficiency of DHF control in Thailand, these activities have to concentrate on areas and populations at higher risk, which implies early identification of higher incidence periods. A retrospective study of spatial and temporal variations of DHF incidence in all 73 provinces of Thailand (1983-1995) allowed discrimination between seasonal (endemic) transmission dependent on climatic variations and vector density and non-seasonal (epidemic) transmission, mainly due to the occurrence of a new virus serotype in a population with low immunity. To identify epidemic months, which appear significantly clustered, a significant deviation from the monthly average incidence was defined. The occurrence of two consecutive epidemic months in a given area has a high probability (P = 0.66) of being followed by a cluster of 2-18 epidemic months (average: 7.7 months). This observation is proposed as a warning of epidemic outbreak enabling an early launch of control activities. As an example, when this method is retrospectively applied to the studied period, 11,388 province months (73 provinces × 156 months), 579 epidemic outbreaks (5.1% of the total) are identified. Control activities can thus be improved through early management and prevention of the 308,636 supplementary cases occurring during epidemics (37.0% of the total recorded). © 2002 Éditions scientifiques et médicales Elsevier SAS. All rights reserved.Mahidol UniversityImmunology and MicrobiologyMedicineArticleSCOPUS10.1016/S1286-4579(02)01589-7