Spatial and temporal analysis of the distribution of hantavirus pulmonary syndrome in Buenos Aires Province, and its relation to rodent distribution, agricultural and demographic variables

We studied the spatial and temporal distribution of Hantavirus Pulmonary Syndrome (HPS) cases from 1998 to 2001 in the Buenos Aires Province, Argentina. HPS is a severe viral disease whose natural reservoir are rodents of the subfamily Sigmodontinae (Muridae) and which occurs in many countries of South and North America. We considered two spatial arrangements: cells of 18.5x18.5 km(2); and departments, the political subdivisions of the province, as spatial units. We tested the departure from a Poisson distribution of the number of cases per cell and per month with the Variance/Mean index, while the interaction between spatial and temporal clustering was tested by means of the Knox and Mantel tests. We constructed probability maps in which the HPS rates per department were considered Poisson variates according to population, area and the product of population and area. We analysed the relation between rodent distribution, environmental and demographic variables and HPS cases conducting preliminary univariate analysis from which we selected variables to enter in general linearized models. We found that both the spatial and temporal distribution of cases is strongly aggregated. The spatiotemporal interaction appears to be related to a strong seasonality and the existence of particular ecological conditions rather than epidemic transmission of the disease. The main explanatory variables for the distribution of HPS cases among the departments of the Buenos Aires Province were human population, the distribution of the rodent Oxymycterus rufus and evapotranspiration. The last two variables are probably indicators of favourable ecological conditions for the reservoirs, which encompass other variables not taken into account in this study.