Researchers develop falls prediction model using health records to predict patients’ risk of falling in the UK

Patients’ risk of falling in the next 12 months could be predicted from their NHS data using a newly developed calculator.

eFalls is a falls prediction model which uses routinely available primary care electronic health record data, the first of its kind in the world.

Developed and tested by researchers from the University of Leeds, the University of Birmingham, and a team of collaborators, it can be used to help identify people at risk of hospitalization or emergency department attendance after a fall over the next 12 months. This means these people can be provided with interventions to prevent falls taking place.

A research paper outlining the findings is published in Age and Aging.

Falls are common among people aged over 65 and can be devastating for people’s personal independence. The risks are multifactorial and include conditions that affect mobility or balance; medications, and home hazards. A history of falls is the strongest risk factor. The incidence of falls is also projected to rise in line with the global aging demographic.

The findings help proactive identification of people who are at risk of experiencing a fall in the next 12 months. eFalls uses existing primary care data, reducing the need for intensive clinical falls assessment, saving doctors and nurses valuable time. Once identified as at risk of falling, people can be referred on to a specialist falls prevention service for assessment and treatment to prevent future falls.

The National Institute for Health and Care Excellence (NICE) estimates that 40% to 60% of falls result in major lacerations, traumatic brain injuries, or fractures. Other complications of falls include distress, pain, loss of self-confidence, reduced quality of life, loss of independence, and mortality.

Principal Investigator Andrew Clegg, Professor of Geriatric Medicine in the University of Leeds School of Medicine, said, “Falls are a global health problem of major importance to health and social care systems. Currently, people’s fall risk is usually only assessed when they have already experienced a fall, which means that they might have already experienced a major injury such as a hip fracture.

“Our eFalls calculator means that, for the first time, it is possible to proactively identify a person’s risk of future falls which means that they can be referred to specialist falls prevention services, reducing the risk of a fall from happening. The ability to put plans in place to protect those at risk is invaluable to the patient and their loved ones.

“The benefit to the health service is that it reduces the need for treatment and care in hospital and in the community, and the associated costs to the NHS of that treatment. We hope that eFalls will be widely adopted across the NHS to prevent falls from taking place.”

Lead author Lucinda Archer, Assistant Professor in Biostatistics at the University of Birmingham, said, “The eFalls calculator can be used to predict a person’s risk of a fall, based on information that is already included in their GP records. The accuracy of the tool has been thoroughly tested in two large datasets, containing routinely recorded information on patients from Wales and England, which has shown promising results.

“If this accuracy is consistent across the wider population, the use of eFalls to target those who would benefit from specialist assessment could vastly improve the way that falls prevention services are provided in the UK.”

Health Minister Andrew Stephenson said, “Suffering from a fall can be traumatic for both the individual and their family but innovations such as eFalls could provide a fantastic solution to prevent such incidents, saving people from a lot of pain, as well as time and resource for the NHS.

“Our ongoing work to ensure people get the right care at the right time includes giving people access to local falls services and rehabilitations services, but I’m proud that the UK is at the forefront of developing further solutions to such a widespread issue, through co-funding the development of this technology.”

The team set out to produce and assess a robust and reliable method to proactively identify people for falls prevention interventions, due to the currently limited availability of such systems.

The team developed the eFalls tool using data from more than 750,000 health care records. Of these almost 35,000 people experienced a fall or a fracture resulting in A&E attendance or hospitalization within 12 months.

The researchers now hope the eFalls prediction model to be successfully integrated into UK primary care electronic patient record systems and are keen to work with UK policymakers to explore how eFalls could be used to inform health policy.