dottedyeti - stock.adobe.com

NHS to use forecasting tool to manage post-crisis Covid-19 spread

A forecasting tool based on machine learning is to be used by the NHS to identify Covid-19 hot spots and manage post-crisis spread

The UK Joint Biosecurity Centre (JBC) will give the NHS locally advanced warning of any new upsurges in Covid-19 cases, using a forecasting tool developed by artificial intelligence (AI) firm Faculty.

The JBC, announced by the government in June, is part of the NHS Test and Trace service, and is headed by Clare Gardiner, who was most recently director of national resilience and strategy at the National Cyber Security Centre.

According to a statement from digital innovation agency NHSX, Faculty has developed the tool as part of the Covid-19 datastore project, that also involves data analytics supplier Palantir, whose Foundry software is used as a front-end data platform. The forecasting tool is “based on Bayesian hierarchical modelling”.

It is said to work by “learning from the data from previous outbreaks, such as bed use and early warning indicators such as 111 call volumes, to model what might happen in the future”.

Indra Joshi, director of AI at NHSX, said: “This tool helps services plan the bringing back on of services for other patients safely, while flexing capacity locally for Covid-19 care.”

Health secretary Matt Hancock added: “The Covid-19 pandemic has shown the power of technology to improve patient care. [This] will help arm the NHS with the insight it needs to predict hospitalisations, weeks in advance and ensure hospitals are prepared.”

The data store and related technology has, according to the statement, “helped provide the NHS at a national level a clear understanding of bed capacity and availability across the country, and has helped ensure ventilators and oxygen supply has been targeted where it is needed so no hospital has run out.

“Now local hospitals have been given access to the new forecasting tool that will help them to plan how to use their available capacity for both Covid-19 patients and routine care and operations, with the benefit of advance knowledge of how the need to care for more or less patients with the virus might change in the coming one to three weeks.”

Read more about NHS Covid-19 data analytics programmes

The government published the contracts for the NHS datastore in early June ahead of a potential court case brought by privacy organisations that might have forced its hand.

According to its contract, Faculty’s role is, in part, to provide “short- and long-term forecasts about the spread of Covid-19 and the impact that has on resources across the healthcare system”.

“The short-term forecasts will be generated through extrapolations of data that exists, and long-term forecasts will be generated using simulation… The Faculty team will be collaborating with an epidemiological research group, at Oxford BDI [Big Data Institute], to develop their simulation tool to meet the objectives of the Covid response group.”

In announcing the forecasting tool, NHSX also confirmed a new four-month contract with Palantir. It stated that the supplier is “a data processor and is only permitted to use the data as directed by the NHS. All data is de-identified or anonymised prior to loading into Palantir’s platform”.

Palantir's involvement with NHS patient data has been considered controversial due to the company's links to US intelligence services. Faculty has also attracted attention because its CEO Marc Warner's brother works for 10 Downing Street special advisor Dominic Cummings.

Read more on Healthcare and NHS IT