Impact case study database

Frailty is a condition that is common in old age. It develops because as we get older our bodies change, and can lose their inbuilt reserves. These changes mean that older people with frailty can experience sudden dramatic changes in their health when they have an illness. International guidelines recommend frailty should be identified routinely so a more holistic approach to care can be taken, and effective treatments provided. However, the main historical difficulty with identifying frailty routinely is that available clinical tools, such as measuring walking speed or frailty questionnaires, require additional resource, and might be inaccurate.

Professor Clegg and Dr Teale led the underpinning research to develop and validate the eFI in a landmark study funded by the NIHR ( 1). The study used the cumulative deficit model, which identifies frailty on the basis of a range of variables including symptoms, signs, diseases, disabilities and abnormal laboratory values, collectively referred to as deficits. Primary care electronic health record (EHR) systems in the UK use Read Codes to categorise and log multiple patient characteristics including symptoms, signs, laboratory test results, diseases, disabilities and information about social circumstances. Primary care EHR systems therefore provide a potentially simple yet powerful mechanism for identifying cumulative deficits to identify frailty routinely.

Between 2014 and 2016, we followed published guidelines to develop, internally validate and externally validate the eFI in a retrospective cohort study using anonymised primary care EHR data in the ResearchOne and The Health Improvement Network (THIN) databases. The work was led by Clegg and Teale in a collaboration including researchers from the University of Birmingham and the University of Bradford. Following initial development, we constructed eFI categories (fit, mild frailty, moderate frailty, severe frailty). We validated by investigating predictive validity for outcomes of all-cause mortality, unplanned hospitalisation and nursing home admission, using adjusted hazard ratios (HR) at 1, 3 and 5 years, and assessed discrimination and calibration.

We used data from a total of 931,541 patients. The eFI incorporates 36 deficits constructed using 2,171 primary care Read Codes. The eFI calculates a frailty score as a simple proportion of the deficits present in an individual divided by the total possible (e.g. if 9/36 deficits are present, the eFI score = 0.25). One year adjusted HR for mortality was 1.92 (95% confidence interval, CI, 1.81-2.04) for mild frailty, 3.10 (95% CI 2.91-3.31) for moderate frailty and 4.52 (95% CI 4.16-4.91) for severe frailty. Corresponding estimates for hospitalisation were 1.93 (95% CI 1.86-2.01), 3.04 (95% CI 2.90-3.19) and 4.73 (95% CI 4.43-5.06), and for nursing home admission were 1.89 (95% CI 1.63-2.15), 3.19 (95% CI 2.73-3.73) and 4.76 (95% CI 2.73-3.73), with good to moderate discrimination.

These key research insights were globally leading as this work was the first international study to use routinely available, nationally representative data to develop and validate a measure of frailty. Following validation, Professor Clegg and Dr Teale led the national implementation of the eFI into the SystmOne, EMISWeb, Vision and Microtest primary care EHR systems, making it freely available to 100% of general practices in England and 95% of all UK practices. The national implementation work was also globally leading, enabling the UK to become the first country to have a nationally available tool for identifying frailty as part of routine care.

This underpinning research work provided the foundation for the impact described in this case study, and has also become a key reference point for future research, including beyond the original institution and research team. This has included internationally-leading work identifying that:

• Home modifications reduce care home admissions for older people with moderate and severe frailty, in collaboration with researchers at the Universities of Swansea and Liverpool ( 2).

• People with frailty account for an additional £6 billion of NHS expenditure, in collaboration with researchers at the University of York ( 3).

• Frailty is a key predictor of hospitalisation after a diagnosis of heart failure, in collaboration with researchers at Imperial College London ( 4).

• Frailty is an indicator of older people nearing end of life, who may be suitable for advance care planning, in collaboration with researchers from the University of Newcastle ( 5), and

• Hypertension risk is modified in the presence of frailty, indicating need for tailored blood pressure care in frailty, in work led by researchers at King’s College London ( 6).

The underpinning eFI work has led directly to considerable competitive grant capture, including:

Associate Director Clegg A, Co-I Teale E

Health Data Research UK North (HDRUK North) (£3,400,000)

HDRUK

05/2020 – 04/2023

PI Clegg A

Development and evaluation of the electronic frailty index+ (eFI+) tool (£548,507)

NIHR

10/2019 – 04/2022

PI Clegg A, Co-I Teale E

Improving care for older people with frailty (£1,500,000)

NIHR ARC

10/2019 – 09/2024

PI Clegg A

Personalised care planning to improve quality of life for older people with frailty (£2,748,666)

NIHR

10/2017 - 02/2023