Parametric Approaches to Model the Prevalence and Force of Infection

Abstract

This chapter discusses several parametric models for the seroprevalence and the force of infection. In the first section we give a brief historical overview of the first parametric models used to model the force of infection, including polynomial and nonlinear models. In the second section we introduce the family of fractional polynomials as a natural extension of polynomial models, circumventing some of the limitations inherent to classical polynomials. These models can be easily fitted to seroprevalence data with R or, e.g., SAS. As usual, we focus on the use of R, with the functions glm and mle. SAS offers similar functionality with the GENMOD and NLMIXED procedure.