A Hierarchical Modelling Framework for Correcting Delayed Reporting in Spatio-Temporal Disease Surveillance Data
share
Delayed reporting is a well established problem in the surveillance of infectious diseases such as dengue fever and influenza. For effective monitoring and intervention, it is crucial to be able to detect outbreaks well before they have been fully observed, based on any partial reports which may be available at the time. Here we propose a substantial spatio-temporal extension to an existing hierarchical modelling framework, to allow for structured and unstructured spatio-temporal variability in the incidence of disease cases and in the reporting delay mechanism itself. We apply a specific instance of this framework to severe acute respiratory infection (SARI) data from the state of Paraná, Brazil. We illustrate how between-region similarity in temporal and seasonal trends can be accounted for, and demonstrate the importance of flexibly allowing for dependence between regions.
READ FULL TEXT