AI tool that speeds up patient discharges trialled by NHS
The NHS is testing an artificial intelligence platform at Chelsea and Westminster Hospital to automate discharge paperwork, aiming to reduce delays and free up beds. Backed by health secretary Wes Streeting, the tool extracts data from medical records to draft summaries, fitting into a broader gove…
The National Health Service has launched a pilot of an artificial intelligence system designed to accelerate the discharge of patients from hospital. The trial, taking place at Chelsea and Westminster NHS Trust in London, uses the AI platform to automatically generate the discharge summaries that clinicians must complete before a patient can leave the hospital or be referred to community services.
How the AI tool works
The software scans a patient’s electronic health record, pulls out key information such as diagnoses, test results, medication lists and follow‑up recommendations, and then drafts a discharge summary. Clinicians review the draft, make any necessary adjustments, and sign off. By handling the bulk of the data extraction and document formatting, the system can shave hours off the traditional manual process, which often stalls when doctors are pressed for time.
Government backing and policy context
Health secretary Wes Streeting highlighted the trial as a concrete example of the UK government’s ten‑year health plan, which seeks to digitise routine processes and free clinicians for direct patient care. Streeting said the AI tool could "cut waiting times" and "free up beds for those who need them most," echoing Prime Minister Keir Starmer’s broader pledge to use AI to "turn around" public services and boost the economy.
The platform is hosted on the NHS Federated Data Platform (FDP), a secure data‑sharing environment intended to allow health and care organisations to collaborate while keeping patient information safe. The FDP is part of a wider strategy to create a more interoperable digital NHS, enabling new tools to be rolled out across trusts without each having to build separate infrastructure.
Potential impact on hospitals and patients
Current discharge processes can create bottlenecks, especially in busy urban hospitals where clinicians juggle multiple responsibilities. Delays in completing paperwork can keep patients in beds for hours after they are medically fit to leave, reducing capacity for new admissions and increasing overall waiting times. By automating the paperwork, the AI tool promises to free up bed space, reduce the length of stay, and allow doctors to devote more time to clinical decision‑making and patient interaction.
Early internal testing suggests the system can generate a draft summary in minutes, compared with the typical 30‑60 minutes of manual entry. If the pilot proves successful, the NHS could scale the technology to other trusts, potentially standardising discharge documentation across England.
Other AI initiatives in the public sector
The discharge pilot sits alongside a suite of AI experiments across the UK public sector. The Department for Science, Innovation and Technology announced a tool that halves the time probation officers spend on note‑taking by transcribing meeting recordings. In the NHS, AI is being used to monitor hospital databases for safety signals, run a physiotherapy app that cut waiting lists for back‑pain patients in Cambridgeshire and Peterborough, and predict patient risk of disease and early mortality through a "superhuman" analytics engine.
Technology secretary Peter Kyle, speaking at Chelsea and Westminster during the discharge trial, framed these projects as part of a "smarter, more efficient state" that could unlock billions in productivity gains. He argued that modernising the public sector with AI is essential to address years of under‑investment and to meet rising demand for health services.
What comes next?
The pilot is scheduled to run for several months, during which clinicians will collect data on speed, accuracy, and user satisfaction. The Department for Health and Social Care will assess whether the AI‑generated summaries meet clinical safety standards and whether any unintended biases emerge from the algorithmic extraction of data.
Should the evaluation show measurable reductions in discharge times and no compromise to patient safety, the NHS plans to roll the tool out to additional trusts. Parallel investments in the FDP and other digital infrastructure will be required to support wider adoption.
Meanwhile, the broader AI agenda continues to evolve. The government has earmarked funding for further AI pilots in areas such as radiology, pathology and mental health triage, signalling a long‑term commitment to embedding intelligent automation throughout the health and care system.
Why it matters
Accelerating patient discharge can free critical hospital capacity, reduce waiting times and allow clinicians to focus on direct care, addressing chronic NHS backlogs.
Key points
- AI platform drafts discharge summaries by extracting data from electronic health records
- Pilot at Chelsea and Westminster aims to cut hours of paperwork and free up beds
- Hosted on the NHS Federated Data Platform for secure, interoperable use
- Part of a wider UK government push to embed AI across public services
- Success could lead to national rollout and significant efficiency gains
Frequently asked questions
What is the AI discharge tool being tested at Chelsea and Westminster?
It is a software platform that reads a patient’s electronic record, pulls out key clinical information and automatically generates a draft discharge summary for clinicians to review and sign.
Who is championing the AI discharge pilot?
Health secretary Wes Streeting has promoted the project as part of the government’s ten‑year health plan, with support from technology secretary Peter Kyle.
How could the tool affect hospital capacity?
By reducing the time doctors spend on paperwork, the tool can speed up patient releases, freeing up beds for new admissions and easing pressure on busy wards.
Is patient data safe in this system?
The tool runs on the NHS Federated Data Platform, which is designed to keep health data secure while allowing safe sharing across trusted NHS organisations.
When might the AI discharge system be available more widely?
If the pilot demonstrates safety and efficiency gains, the NHS plans to expand the tool to other trusts after the evaluation period, likely within the next year.
