Epidemic modelling and statistical support for policy: sub-populations, forecasting, and long-term planning

Lead Research Organisation: University of Manchester
Department Name: Mathematics

Abstract

The ongoing COVID-19 epidemic requires careful monitoring as a variety of measures such as lockdown and social distancing are introduced and subsequently relaxed, leading to varying levels of demand for and capacity within the healthcare system. The disease has varying expected outcomes depending on the age, sex, and underlying comorbidities of cases. Epidemic dynamics, particularly in the presence of changing control policies, will shift the dominant modes of transmission and hence the distribution of disease. We will develop models to integrate the diverse but often noisy and incomplete datasets available, providing real-time policy support together with quantification of uncertainty. We will address three particular challenges. (1) Understanding spread in closely connected sub-populations in which there are close, repeated contacts capable of spreading disease such as households, hospitals, prisons, and care homes. Data from these contexts allow epidemiological parameters relating to infection risk conditional on contact to be identified in statistical work, and they are also important foci for policies. (2) Making short- and medium-term predictions of the epidemic trajectory and healthcare demand with appropriate uncertainty quantification. (3) Modelling long-term prospects for the epidemic, including the likelihood of eventual endemicity, the consequences of different virological assumptions about SARS-CoV-2, and how the different scenarios in this context will interact with long-term societal and health consequences of the pandemic. The project will use mathematical methodology, integrated with interdisciplinary expertise from social science, biology, clinical medicine, and epidemiology.