Super Poisson variance model for the study of aggregation of the human immunodeficiency virus epidemic

Objective To quantitatively analyse the clustering trends of of human immunodeficiency virus(HIV)prevalence in space and populations over time.Methods Chongqing HIV epidemic data were fitted over Poisson variance model and multi-level super-Poisson variance model.The coefficients of super variance of different years were calculated,and curves of the coefficient changes over times were drawn.Results Both super variance coefficients of space and populations increased gradually over time from 1994 to 2009.Conclusion The clustering trend of HIV prevalence in Chongqing gradually increased.