Trial of machine learning tool to manage Covid-19 pressure launched

A system that uses machine learning to predict upcoming demand for intensive care beds and ventilators is being trialled by the NHS.

The Covid-19 Capacity Planning and Analysis System (CPAS) aims to support hospitals to more accurately plan and ensure resources are deployed where they’re most needed to treat patients.

It will focus on demand in individual hospitals and across regions in England, as the NHS looks to mitigate the unprecedented pressure placed on services by the coronavirus outbreak.

The tool is designed to help hospitals predict how many patients may need an ICU bed, how many may require ventilators and how long patients are likely to be in hospital. First stage trials have begun in four hospitals.

Professor Jonathan Benger, NHS Digital’s chief medical officer, said: “With the pressure being placed on intensive care by the current coronavirus pandemic it is essential to be able to predict demand for critical care beds, equipment and staff.

“CPAS allows individual hospitals to plan ahead, ensuring they can give the best care to every patient. At the same time, the wider NHS can ensure that the ventilators, other equipment and drugs that each intensive care unit will need are in place at exactly the time they are required.

“In the longer term, it is hoped that CPAS can be used to predict hospital length of hospital stay, discharge planning and wider intensive care demand in the time that will come after the pandemic.”

The machine learning tool was developed by NHS Digital data scientists and a team of researchers from the University of Cambridge, using data from Public Health England on Covid-19 patients admitted to hospital.

CPAS uses a machine learning engine called Cambridge Adjutorium, developed by Professor Mihaela van der Schaar and her team at the University of Cambridge. The highly flexible system has already been used to develop insights into cardiovascular disease and cystic fibrosis.

NHS Digital is working with the team to improve the capability and accuracy of the system by integrating a wider range of data collected by NHS Digital alongside the data from Public Health England.

“Although the system uses data from individuals to build its models, the system does not make treatment decisions about individual patients,” Professor van der Schaar said.

“Rather, by aggregating that data we can make more accurate predictions about larger groups, at the level of a hospital, a trust, a region or nationally. So while we can say with a high level of confidence that 30 out of 40 ITU beds in a hospital will be occupied next week, we are not trying to predict which patients will be in them.”

If the trial proves successful the system will be rolled out across the NHS and “several other countries have already expressed an interest” NHS Digital said.

Capabilities available to hospitals using the tool:

• statistics, which provide a demographic picture of the population of patients being admitted and key information such as additional medical conditions
• forecasts, which will help predict the need for beds, ventilators or other resources over the coming days
• a “simulation environment” which will allow planners to test the effect of alternative scenarios, such as increasing the number of available beds