Elnur - stock.adobe.com

UCLH uses machine learning to cope with emergency beds demand

University College London Hospitals NHS Foundation Trust has deployed a machine learning tool which uses real-time data to predict how many emergency beds will be needed

University College London Hospitals NHS Foundation Trust (UCLH) is using a machine learning tool to predict bed capacity in the hospital.

Hospital staff at the trust have, together with analysts from University College London (UCL), developed the tool, which is being used to forecast how many emergency beds will be needed in the next eight hours across different departments, including pediatrics, oncology, surgical, medical and haematology.

UCLH clinical operations manager Craig Wood said that managing the flow of patients around the hospital “is the continuous, careful balance of patients who are in for elective treatment with people in A&E who need a bed as an emergency”.

“Since implementing the tool at UCLH, we’ve been able to make targeted actions to free up beds in specific areas of the hospital, improving our ability to manage hospital capacity,” he said.

The NHS as a whole is struggling to cope with demand, and many hospitals are running at full, or over, capacity. The trust hopes the tool can help operational staff ensure there are enough beds in the different departments.

When patients present to A&E and have observations and tests done, the machine learning tool uses the data from things like vital signs, blood test results and whether they need to see a specialist doctor, and the forecast updates every 30 minutes.

It also predicts how many people will present to A&E over the next eight hours, using electronic patient records from patients who visited the department between August 2021 and August 2022.

Read more about the NHS and technology

  • The Labour Party wants to create an NHS powered by “groundbreaking technology”, with genomics testing for every baby, modernising appointment booking and overhaul incentives for tech adoption.
  • Welsh ambulances going through major communications upgrade as part of wider NHS programme.
  • Scotland’s five-year strategy aims to make it easier for people to access their own health and social care data, improve data flows between organisations, and transform the way data is used to enhance services.

Co-developer of the technology Zella King, from UCL’s clinical operational research unit, said their research has demonstrated the potential of using machine learning and data to improve day-to-day operations in hospitals.

“Hospital capacity is affected by many systemic factors, and we’re not suggesting machine learning is a magic bullet to fix this complex challenge,” she said. “But it’s fantastic that our tool is making a meaningful difference to the day-to-day running of UCLH.

“Machine learning could be a powerful tool to support hospital operational staff, but its outputs have to be both actionable and aspirational.”

The research builds on previous work done by the team in 2022, when the tool could predict the total number of beds needed, but at the time, the tool couldn’t yet break the number down by department.

In 2020, during the Covid-19 pandemic, artificial intelligence firm Faculty developed a tool used by the government to give the NHS locally advanced warning of any new upsurges in Covid-19 cases. The tool, according to the government, helped give the NHS at a national and local level a clear understanding of bed capacity and availability. 

    Read more on Big data analytics

    CIO
    Security
    Networking
    Data Center
    Data Management
    Close