Skip to main content

Go local: the key to COVID-19 lockdown release

Home > New and events > Go Local Key Covid 19 Lockdown Release

Pandemics, such as COVID-19, are usually assumed to spread rapidly within the population. In reality, the population is more heterogeneous with regard to risk, and there will be large variation on the basis of geography, workplace and other key factors.

Analysing data on a national level therefore risks hiding this heterogeneity and compromises the most effective public health response. New analysis from i-sense researchers at University College London suggests COVID-19 has such diverse effects on the different local authorities in the UK.

The piece, published as a pre-print on arXiv, draws on geospatial data highlighting local variation in mobility across the country, as well as case rate data, which shows potential hotspots of infection.


Cumulative reported COVID-19 case rate by local authority in England from 30 January to 29 June 2020. Cases normalised by population in each local authority (per 100,000 people). Data from Public Health England. 


This type of anonymised aggregated population data, made available from project collaborators O2 (Telefonica UK) and Public Health England (PHE), is key to understanding population patterns in near real-time, allowing for earlier identification of hotspots.

UCL Prof of Biomedical Nanotechnology and i-sense Director, Prof Rachel McKendry, said: “We have already seen in the UK, most prominently through the recent outbreak in Leicester, that a locality-based approach to easing lockdown is needed to control the pandemic.”

UCL Prof of Virology and i-sense Deputy Director, Prof Deenan Pillay, added: “We need to work with and empower local public health and associated health and social care services to rapidly respond to emerging hotspots of infection in their area at a local level. This requires appropriate data integration so that real-time data can be provided to public health practitioners in the most informative way.”

i-sense researchers are suggesting a more nuanced and strategic approach in the UK, similar to that taken in France and Germany. Important to this is access to real-time or near real-time local level data.

What does analysis of mobility data tell us about the easing of lockdown?

Anonymised and aggregated local authority level mobility data helps highlight the varying adherence to lockdown and the response to the easing of restrictions by understanding how people are moving to, from, or within an area.

Co-lead author and i-sense PhD student, Jobie Budd, said: “Access to anonymised and aggregated near real-time data sets, for example from telecoms companies, is really important in helping to better understand the impacts of easing lockdown on a more local level.”

“Our analysis shows that some areas across the UK returned to near pre-lockdown levels of mobility after the second wave of lockdown easing took place on 1 June.”

Other cities continued to show lower levels of mobility, highlighting the need for differing interventions across the country.

“To highlight the effect of policy on human behaviour, it is worth noting that there were much higher levels of mobility across England, where policies were most relaxed, compared to Scotland and Wales that maintained stricter policies.”

Understanding how mobility may affect case rates within an area is hard to determine because of delays in both onset of symptoms and testing, as well as increased testing capacity in a local area.


Percentage increase in non-commute trips (see Data section for definition) between two weeks prior to easing on 10 May and first week of June by local authority. Data shows trips ending in local authority. Data for trips started and ending by local authority was found to be not significantly different. Anonymised and aggregated crowd level mobility data provided by O2 (Telefonica UK).


What is needed for a more local level easing?

Working with local authorities to co-develop policies fit for their area will help build communication and trust with the public, motivating people to adhere to non-pharmaceutical interventions, such as lockdown.

Prof Pillay said: “If we work with local authorities, we can better respond to situations like that in Leicester, with locally-relevant and real-time data, so that local public health authorities are empowered to make decisions in a timely manner.”

An Independent SAGE’s Statement of Leicester and Local Lockdowns, published on 2 July 2020, suggests that the situation arose because of centralisation and unavailability of data, fragmented testing, and lack of coordination of with local authorities.

Co-lead author and i-sense policy lead, Dr Isabel Bennett, suggests: “Looking at datasets such as case rate and mobility data on a local level can give us a better understanding of what is going on in a particular area, helping to develop policies that suit that local population."

Local level policy requires a collaborative approach to build a more responsive system that can ultimately control the COVID-19 pandemic. This approach will require access to anonymised and aggregated data sets from industry and government to understand populations on a more granular level, councils working together to share plans and insights, and making resources, such as secondments, available from central government.

Acknowledgements

This review was conducted in collaboration with experts across University College London and Department for International Trade.

The review uses anonymised and aggregated, UK population representative crowd level mast data from O2 (Telefonica UK) detailing the aggregate number of trips starting within each UK Lower Tier Local Authority (LTLA) for 1 February - 7 June 2020 inclusive. Individuals cannot be separately identified or located from this data set. 

Cases data for this review was collected from NHS and PHE laboratories ('pillar 1') and the mass testing programmes ('pillar 2').

Authors

Isabel Bennett, Jobie Budd, Erin M. Manning, Ed Manley, Mengdie Zhuang, Ingemar J. Cox, Michael Short, Anne M. Johnson, Deenan Pillay, and Rachel A. McKendry

Related links

Figures