Statistical methods to estimate infectious disease parameters and individual heterogeneity using multivariate serological data (R-8133)

The PhD proposal aims at developing and applying statistical methods for the analysis of multivariate serological data in the presence of observed and unobserved heterogeneity in the acquisition of infectious diseases. These methods will be used to infer important epidemiological parameters such as the basic reproduction number, the force of infection and the amount of unobserved heterogeneity. More specifically, we aim at developing estimators to quantify the association between two infections in the two-disease model proposed by Rohani et al. (1998) and at extending this model to allow for individual variation in the acquisition of these infections. Furthermore, an extension of the bivariate correlated gamma frailty model to allow for more flexible correlation structures among infection-specific frailty terms will be pursued. On top of that, we aim at developing tools for the analysis of trivariate serological data, thereby assessing identifiability and unbiasedness. Finally, a model for doubly interval-censored data on recurrent infections will be introduced thereby accounting for individual heterogeneity in acquiring these infections, and accommodating dependence between the event and observation time processes.

Keywords:MODELS INFECTIOUS DISEASES

Disciplines:Scientific computing, Bioinformatics and computational biology, Public health care, Public health services