Shortlisted for Research Project of the Year: STEM – Times Higher Education Awards 2025


An artificial intelligence (AI) system developed by a team of Reading researchers is being used by the NHS to predict and prevent missed hospital appointments, enabling targeted patient support and substantial cost savings.

“Did-not-attend” (DNA) appointments pose significant challenges for the NHS. In 2024, missed hospital appointments reached 11.8 million, translating to nearly £1.9 billion in lost NHS funds. Importantly, missed appointments widen health inequalities – the patients who need care most (due to chronic illness, mobility issues, or lack of support) are most at risk of missing their appointments. Patients living in the most deprived neighbourhoods are twice as likely to miss appointments as those in affluent areas. Barriers such as transport difficulties, inflexible work or carer commitments, and communication issues often contribute to no-shows among disadvantaged patients.

Reducing missed appointments is vital, but identifying which patients are high-risk – and understanding why – is complex, as these factors are frequently invisible in routine clinical data. The challenge lies not only in accurately predicting non-attendance using messy real-world data but also in going beyond prediction to contextual insight: recognising that each patient’s needs are unique and that effective intervention requires making these barriers visible and actionable.

In partnership with Royal Berkshire NHS Foundation Trust (RBFT) – one of the largest NHS trusts in the UK, serving a population of 1 million – Professor Weizi (Vicky) Li co-developed an explainable artificial intelligence (XAI) model that used over 500,000 past appointments spanning 80+ specialties to predict each patient’s risk of non-attendance with ~92% accuracy. Crucially, this project is the first in the NHS to use XAI that not only identifies high-risk patients but also highlights individual patient risk factors (e.g. previous missed appointments, deprivation level, mental health issues or long travel times).

The XAI model was developed into a decision support system with accompanying guidance material, enabling staff to deliver targeted, compassionate care. Using AI-generated insights to guide conversations, RBFT staff proactively call high-risk patients to understand their individual circumstances and offer tailored support based on specific barriers identified by the model. Examples include travel cost reimbursements for patients facing financial constraints, interpreters to navigate language barriers, or flexible scheduling for those with caring responsibilities.

The decision support system has been operational across all RBFT sites and departments since 2023, supporting a large number of outpatient appointments annually (694,199 in 2023/24 with a rising trend) and reducing high-risk non-attendance appointments by 40%. The reduction in missed appointments translates to estimated cost savings of up to £250,000 during 2023/24, based on a cost of £130 per appointment. Patients also provided positive feedback about how the personalised approach of RBFT staff helped them overcome barriers and attend their appointments.

The team is seeking funding to scale the project to other hospitals and is in discussions with two other NHS trusts and NHS England about implementation. The success of this work, which was partly funded by UKRI ESRC, has also led to a series of AI developments funded by UKRI and NIHR that support the patient pathway, including early disease detection to improve referrals, disease activity forecasting to improve follow-up services, and patient clinical status forecasting to enhance triaging to the virtual ward.

“Not only are we able to help identify high-risk DNA (Did Not Attend) patients but with the right intervention we are able to reduce our rate of missed appointments. helping to improve clinical and operational outcomes for both the hospital and our patients.”

Eghosa Bazuaye, Associate Director of Informatics, Royal Berkshire Hospital

In partnership with Royal Berkshire NHS Foundation Trust

Team: Weizi (Vicky) Li and Nicholas Berin Chan (University of Reading), Eghosa Bazuayeshire, Toluwanimi Akinola and Kiki Kontra (Royal Berkshire Hospital)

Published: September 2025