Understanding the dynamics and drivers of the COVID-2019 epidemic using real-time outbreak analytics

Lead Research Organisation: London Sch of Hygiene and Trop Medicine

Abstract

Tracking and predicting an epidemic can help tailor public health measures so that we
respond quickly and appropriately to evolving threats. We will fit statistical models to
emerging data to monitor key epidemiological and clinical parameters related to
transmissibility and severity, and their likely impact on health services. Mathematical
models will be fitted to surveillance data to forecast future cases and hospitalisations and
evaluate the impact of different public health policies. Importantly, we will also measure
people’s behaviour over time to detect changes in response to the epidemic, assessing
what measures they take to reduce risk and where they get their evidence from. This
information will be used to refine our models and improve epidemiological forecasting. It
will also allow public health authorities to tailor their messaging to ensure that people
respond appropriately to changes in risk. The information that we generate, either from
our analyses of the emerging data, epidemic forecasts, or from our panel of respondents,
will be rapidly summarised and shared on easy-to-use websites so that scientists, health
workers and the public can be best informed about the threats posed by COVID-19 and
the efficacy of different measures to reduce risk.

Technical Summary

This COVID-19 Rapid Response award is jointly funded (50:50) between the Medical Research Council and the National Institute for Health Research. The figure displayed is the total award amount of the two funders combined, with each partner contributing equally towards the project.

Efficient response to COVID-19 requires an understanding of the epidemiological and
behavioural drivers of disease transmission. Due to the rapidly evolving outbreak and the
mitigation strategies likely to be put in place at different times, analyses of epidemic
drivers and policy evaluation need constant updating to provide relevant data-driven
evidence to inform evolving public health choices.
We will provide rapid, continually updated estimates of key epidemiological features such
as disease severity and transmissibility measures and lengths of stay. Surveillance,
serological and sequence data (where available) will be analysed accounting for censoring
and reporting delays. Mathematical models will be fit to the emerging data streams using
Bayesian methods to provide regular forecasting updates and assess the impact of
current or potential future interventions. Contact and precautionary behaviours will be
monitored in a representative cohort along with information on risk awareness and
perceived efficacy of interventions to refine transmission models, improve forecasting, and
assess the effectiveness of social distancing measures.
Special attention will be given to sharing results in an open and timely manner.
Epidemiological parameter estimates and forecasts will be shared on a public website,
updated daily. User-friendly web interfaces will allow users to generate model outputs and
investigate the impact of specific model assumptions on different policy findings.
Highlights of essential results will be gathered in short policy briefs updated weekly.
Findings will immediately inform UK policy through participation on UK Government
advisory committees.

Publications

10 25 50