Climate variability and the Dengue outbreak in Townsville, Queensland, 1992-93

Abstract

Although Australia is free of local dengue transmission, there have been several outbreaks of dengue in Queensland since 1981 because of importation of dengue cases. This study was conducted on monthly time-series data of the dengue outbreak in Townsville in 1992 and 1993. It used graphic assessment, Spearman correlation, AutoRegressive Integrated Moving Average and Generalised Least Square regression analyses. The results showed that the annual mean minimum temperature in the outbreak years was 0.6°C higher than that in non-outbreak years. Monthly mean maximum and minimum temperatures, total amounts of precipitation and relative humidity four months earlier were significantly correlated with the monthly attack rate of the disease. The regression analysis suggested that monthly mean minimum temperature with a four-month lagged effect was the strongest climatic predictor of the dengue fever outbreak and indicated that it was important in dengue transmission. However, the relationship between climatic variables and dengue fever needs to be viewed within a wider context of other social and environmental factors, such as population growth, human behaviour, house conditions and vector control programs. Vigilance in control and prevention of dengue fever may need to be increased in Australia because of possible climate change with temperature increases.