Trials have begun of a system that will use machine learning to help predict the demand for ICU beds and ventilators during the pandemic.
Using data from Public Health England, the Covid-19 Capacity Planning and Analysis System (CPAS), developed by NHS Digital data scientists and a team of researchers from the University of Cambridge, will support hospitals to plan more accurately and help to ensure that resources are deployed to best effect.
CPAS will support a hospital to predict how many patients may require ICU admission, how many may require ventilators, and how long patients are likely to be in hospital or ICU.
Taken together, these predictions will provide information to support the hospital to plan the resources they will need to look after those patients, make up shortfalls, or share excess capacity with other hospitals.
CPAS is built around a machine-learning engine for medical researchers called Cambridge Adjutorium, developed by Professor Mihaela van der Schaar and her team at the University of Cambridge and already used to develop insights into cardiovascular disease and cystic fibrosis.
The information and capabilities available to hospitals using the CPAS tool in the trial are made up of three key elements: statistics, forecasts, and a “simulation environment”.
If the trial deployments prove successful over the coming days, the CPAS system will be rolled out across the NHS, and several other countries have already expressed an interest.
New studies funded by the 24.6 million rapid research response funded by UK Research and Innovation (UKRI), and by the Department of Health and Social Care through the National Institute for Health Research (NIHR) have been revealed.
Two, led by UCL and UCLH, will see thousands of people across the country being tested for Covid-19 to see how the coronavirus spreads among communities and NHS workers.
Professor Andrew Hayward, from UCL Epidemiology & Health Care, will lead the £3.2m ‘Virus Watch’ study to recruit 25,000 individuals and collect data from April 2020 to March 2021.
He said: "So far most of our information about Covid-19 is from people who are ill enough to need to go to hospital. This means we have a very incomplete understanding of the infection and how we can fight it.
“We will find out how common Covid-19 is over time, what is the full range of symptoms, how easily it spreads within the household, how the virus affects different age groups, those with chronic illness, different ethnic groups and different occupations.”
Virus Watch will also investigate how people’s movements and use of public spaces during periods of lockdown and when the lockdown is eased affect the risk of infection, as well as establish how many people have developed antibodies and how effectively these antibodies help to protect against future infection. The research team will be releasing data throughout the study period.
The University of Manchester is also to undertake a £300,000 study as part of the rapid response funding, looking at the role of gig workers and delivery supply chains in preventing disease transmission.
Professor Martie van Tongeren and Dr Hua Wei will lead a team, which will work with companies in this sector to collect data on deliveries to construct mathematical models.
The models will determine how the delivery sector contributes to minimising the risk of spreading the disease. They will also set out the impact of additional measures to protect the workers and reduce infection risk during the Covid-19 pandemic.
The study will inform the UK’s policy making in social distancing and coordination of supply chains as a key component of national response to pandemics.
The research funding has been co-ordinated with other funders and the World Health Organisation (WHO) to ensure there is not duplication of effort and expertise is applied strategically.
A study released today (20 April 2020) from the University of Oxford highlights the importance of age and demographic science in explaining the difference in fatalities across countries affected by the virus.
The research, from Jennifer Beam Dowd, Melinda Mills and colleagues at the Leverhulme Centre for Demographic Science, University of Oxford and Nuffield College, emphasises the potential for dramatically higher fatality rates in countries and localities with older populations.
Covid-19 mortality risk is highly concentrated on people aged 80 and over but understanding a country’s age demographics can help predict the burden of critical cases and aid in more precise planning of availability of hospital beds, staff and other resources.
The study was motivated by the surprising early severity and number of deaths from Covid-19 in Italy. Compared to South Korea, which also had an early surge in cases, Italy has one of the oldest populations in the world with 23.3% over age 65, compared to 14% in South Korea.
Using the current age-specific case fatality rate in Italy, the researchers illustrate how population age structure interacts with high Covid-19 mortality rates at older ages to generate large differences in numbers of deaths. In Italy, the predicted number of fatalities was 1.7 times greater than for South Korea.
Beam Dowd J, Rotondi V, Andriano L et al. Demographic science aids in understanding the spread and fatality rates of Covid-19.Proceedings National Academy of Sciences 15 April 2020
Leave a Reply