COVID-19 hotspots projected with new website

Owl has just come across this Covid-19 hotspot predictive web site produced by Imperial College. 

Exeter and to a lesser extent Mid-Devon are shown as lingering concentrations of local infection as we move into March

by Hayley Dunning, Dr Sabine L. van Elsland 03 September 2020 www.imperial.ac.uk

Screen shot

A new website uses reported cases and deaths to estimate the probability regions in England, Scotland and Wales will become COVID-19 ‘hotspots’.

The team behind the website, from Imperial College London, define a hotspot as a local authority where there are more than 50 cases of COVID-19 per 100,000 of the population per week.

COVID-19 is, unfortunately, very much still with us, but we hope this will be a useful tool for local and national governments trying to bring hotspots under control. Professor Axel Gandy

Using data on daily reported cases and weekly reported deaths and mathematical modelling, the team report the probability (%) that a local authority will become a hotspot in the following week.

The site also provides estimates for each local authority in England, Scotland and Wales on whether cases are likely to be increasing or decreasing in the following week and the probability of R(t) being greater than 1 in the following week.

The reproduction number R(t) indicates the number of people each infected person will pass the virus onto. An R(t) larger than 1 indicates the outbreak is not under control and cases will continue to increase.

Current hotspots

The site will be updated daily. 

The website was produced by the Department of Mathematics in collaboration with the WHO Collaborating Centre for Infectious Disease Modelling within the MRC Centre for Global Infectious Disease Analysis (GIDA), and Abdul Latif Jameel Institute for Disease and Emergency Analytics (J-IDEA) at Imperial.

Lead researcher Professor Axel Gandy, from the Department of Mathematics at Imperial, said: “The model allows us to project where local hotspots of COVID-19 are likely to develop in England and Wales based on the trends that we’re seeing in those areas. COVID-19 is, unfortunately, very much still with us, but we hope this will be a useful tool for local and national governments trying to bring hotspots under control.”

Enabling swift action

The predictions assume no change in current interventions (lockdowns, school closures, and others) in a local authority beyond those already taken about a week before the end of observations. Each local authority is also treated independently of its neighbours in the modelling, i.e. the epidemic in one local authority does not affect or is not affected by the situation in any adjacent a local authority.

The team also note that an increase in cases in a local authority can be due to an increase in testing, which the model does not currently account for. The model also assumes all individuals within each local authority are equally likely to be infected, so demographic factors such as the age structure of the population are not considered.

Dr Swapnil Mishra, from the MRC Centre for Global Infectious Disease Analysis, said: “We provide weekly predictions of the evolution of COVID-19 at the local authority level in England and Wales. Our model helps to identify hotspots – probable local areas of concern. We hope that our estimates will enable swift action at the local level to control the spread of the epidemic.”