Multiresolution predictive dynamics of COVID-19 risk and intervention effects

SARS-CoV2 is a novel virus, and even as new data improves scientific insight, many uncertainties remain about key aspects of transmission. Throughout the pandemic, mathematical and statistical models of COVID-19 have had an important role in the analysis of epidemiological data, in forecasting incidence trends and in assessing the potential impact of different intervention strategies. Models developed by the Imperial College COVID-19 response team have been particularly influential, but the absence of detailed data on transmission patterns have necessitated important assumptions that limit their predictive performance. This project will (a) extend predictive models of transmission trends to include complex spatiotemporal correlation to better capture new seeding events and improve early identification of hotspots of transmission, (b) understand the causal effect of interventions on transmission and the limits to which this inference is possible, (c) systematically collate and analyse data on transmission in specific contexts (households, schools, workplaces and care homes) to derive specific transmission parameter estimates for those settings to be used to improve the ability of models to predict the impact of targeted non pharmaceutical interventions, (d) Understand how important epidemiological parameters are changing with time and what is driving these changes. This work will directly support the Imperial team's input into the UK COVID-19 response via the SPI-M, NERVTAG and SAGE committees and our partnerships with PHE and the Joint Biosecurity Centre (JBC).

SARS-CoV2 is a novel virus, and even as new data improves scientific insight, many uncertainties
remain about key aspects of transmission. Throughout the pandemic, mathematical and statistical
models of COVID-19 have had an important role in the analysis of epidemiological data, in
forecasting incidence trends and in assessing the potential impact of different intervention
strategies. Models developed by the Imperial College COVID-19 response team have been
particularly influential, but the absence of detailed data on transmission patterns have
necessitated important assumptions that limit their predictive performance. This project will (a)
extend predictive models of transmission trends to include complex spatiotemporal correlation to
better capture new seeding events and improve early identification of hotspots of transmission,
(b) understand the causal effect of interventions on transmission and the limits to which this
inference is possible, (c) systematically collate and analyse data on transmission in specific
contexts (households, schools, workplaces and care homes) to derive specific transmission
parameter estimates for those settings to be used to improve the ability of models to predict the
impact of targeted non pharmaceutical interventions, (d) Understand how important
epidemiological parameters are changing with time and what is driving these changes. This work
will directly support the Imperial team's input into the UK COVID-19 response via the SPI-M,
NERVTAG and SAGE committees and our partnerships with PHE and the Joint Biosecurity Centre
(JBC).