Time series analysis of dengue incidence in Guadeloupe, French West Indies: Forecasting models using climate variables as predictors

摘要: During the last decades, dengue viruses have spread throughout Americas region, with an increase in number of severe forms dengue. The surveillance system Guadeloupe (French West Indies) is currently operational for detection early outbreaks goal study was to improve this by assessing a modelling tool predict occurrence epidemics few months ahead and thus help efficient control. Box-Jenkins approach allowed us fit Seasonal Autoregressive Integrated Moving Average (SARIMA) model incidence from 2000 2006 using clinical suspected cases. Then, used calculating year 2007 compared observed data, three different approaches: 1 year-ahead, 3 months-ahead month-ahead. Finally, we assessed impact meteorological variables (rainfall, temperature relative humidity) on prediction outbreaks, incorporating them fitting best. most appropriate effective public health response, accurate (Root Mean Square Error, RMSE = 0.85). Relative humidity at lag-7 weeks, minimum lag-5 weeks average lag-11 were positively correlated Guadeloupe, meanwhile rainfall not. predictive power SARIMA models enhanced inclusion climatic as external regressors forecast 2007. Temperature significantly affected better forecasting (p-value 0.03 lag-5, p-value 0.02 lag-11) but not humidity. Minimum best variable predicting (RMSE 0.72). improves forecasts than rainfall. data independent could be easily incorporated into (3 months-ahead) reliably monitoring outbreaks. This which practicable has implications helping epidemic therefore timely implementation prevention activities.