Digital contact tracing uses a proximity-detecting phone app to alert people at risk of being infected. It was implemented for COVID-19 in the UK by the NHS, and also in many other countries.
In a new study published in Science, researchers from the Pandemic Sciences Institute, the University of Warwick and the UK Health Security Agency (UKHSA) analysed data from these apps. They found that digital contact tracing can provide unprecedented insights into epidemic dynamics, allowing public health bodies to better monitor and analyse evolving epidemics.
The authors analysed anonymised data that was collected by the NHS COVID-19 app for England and Wales to ensure its correct function. During the COVID-19 pandemic the authors provided updates of many of the results presented here to the UK Government and public health authorities with weekly frequency, and at peak times daily, for situational awareness.
These results are now for the first time being presented for scientific publication, showing how the analyses performed over the whole pandemic period, along with detailed analyses focused on robustness and generating methods of wider applicability for use in pandemic preparedness.
Researchers used data from the app to calculate the dynamics of the reproduction number R – the average number of times each person with the virus passed it on to someone else – seeing changes five days earlier than other methods, providing an early warning system when the epidemic suddenly changed.
A unique feature of this data allowed the authors to determine when R changed because of changing contact patterns between people, and when it changed because of a higher or lower probability of transmission to each person.
The researchers observed regular variability in transmission events detected by the app by day of the week and by context – for example, they noticed that there were twice as many transmissions associated with brief encounters (less than half an hour) on Saturdays as on Mondays.
Professor Christophe Fraser, Professor of Infectious Disease Epidemiology at the Pandemic Sciences Institute and Principal Investigator of the study, said: “Our work has shown that digital contact tracing frameworks, as well as reducing the spread of respiratory infections like COVID-19, can be of great use in providing real-time information on the state of the epidemic and the nature of transmission. Ensuring that digital systems are in place before new pathogens begin rapidly spreading is critical to preparing for future epidemics. The current worrying spread of highly-pathogenic avian influenza amongst multiple mammal species in the Americas should serve as a warning. We must build systems such as contact tracing in advance to prepare the world for future pandemics, making sure lessons from COVID-19 are implemented rather than forgotten.”
Dr Michelle Kendall of the University of Warwick, who co-led the analysis, said: “Public health interventions which restrict population movement inevitably have socio-economic costs. Measuring which types of human contact are – and are not – driving transmission is important for balancing these costs against the harms caused by the disease. We have shown that privacy-preserving data from digital contact tracing can reveal valuable information about transmission very quickly and in remarkable detail. We are grateful to everyone in England and Wales who engaged with the NHS COVID-19 app. Not only did they help limit the spread, reduce pressure on the NHS and save lives, but their anonymised data also provided important real-time updates on the evolving epidemic and unprecedented insights into how a respiratory virus was transmitted.”
The study builds on previous work by this team on digital contact tracing (collected here): proposing its use to accelerate contact tracing for COVID-19 (Science 2020), evaluating how many infections it prevented and lives it saved in England and Wales (Nature 2021 and Nature Communications 2023), and understanding how each individual’s risk of infection depended on the duration and proximity of their exposure (Nature 2023).