Optimal surveillance design for parameter inference in hidden-Markov epidemic models

Lead Research Organisation: Lancaster University
Department Name: Mathematics and Statistics


Bayesian inference on the parameters of dynamical infectious disease models is an established methodology for
outbreak forecasting and decision support. These approaches propose a dynamical state-transition model for disease
progression, in which an individual's infection hazard is a function of the number of infected individuals at any given
time [e.g. Jewell et al. (2009)]. These models assume that all infected individuals are eventually detected, and that
case detection is intrinsically linked to the development of clinical signs of disease.

A common situation, however, is where samples of individuals from a population are subject to a regular disease screening
programme which is independent of the infection process -- so-called 'panel data' -- with the aim of estimating disease
transmission parameters to inform both further disease surveillance and also disease intervention policy. Though methods
for state-transitions models given panel data are well-established for static models [e.g. Jackson (2011)], the time-
inhomogeneous nature of transition rates for epidemic models presents a significant inferential challenge. Moreover,
optimal design of a disease surveillance programme to inform epidemic models is currently under-developed [Herzog et al.

This PhD project will focus on developing statistical methodology to fit epidemic models in a hidden-Markov framework, and
develop principles to ensure that the spatiotemporal design of sample-based disease surveillance is optimal for both
parameter inference and epidemic forecasting.

* Jewell CP, Kypraios T, Neal P, Roberts GO (2009) Bayesian analysis for emerging infectious diseases. Bayesian Analysis.
* Jackson CH (2011) Multi-state models for panel data: The msm package for R. 38(8):1-28
* Herzog, SA, Blaizot S, Hens N (2017) Mathematical models used to inform study design or surveillance systems in
infectious diseases: a systematic review. BMC Infectious Diseases. 17:775-785.


10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/S515127/1 01/10/2018 30/09/2022
2118868 Studentship EP/S515127/1 01/10/2018 30/09/2022 Joshua MacDonald