Development and national implementation of eFI2, CY P12 20 01

To develop and evaluate an improved prognostic tool, the eFI2, to support identification of older people with different levels of frailty. UK and international guidelines support routine frailty identification in primary care to allow timely and targeted proactive care [1-3]. To enable routine frailty identification, Clegg et al. developed, validated and nationally implemented the electronic frailty index (eFI) using The eFI has some limitations, including the allocation of equal weighting to deficits, and the cumulative addition of deficits that are assumed not to improve or resolve. Additionally, the original validation did not investigate key frailty outcomes, such as loss of independence or falls. Nineteen potentially relevant deficit variables were identified but not included in the eFI, either because prevalence did not increase from age 65 to 95, or because overall population prevalence was low. However, evidence suggests that some of these may be important predictors, e.g. we know that mental health disorders increase the risk of frailty progression, mortality and primary and secondary healthcare use [7, 8]. routine primary care electronic health record (EHR) data [4]. The eFI incorporates 36 deficits (clinical signs, symptoms, diseases, disabilities, impairments), constructed using around 2000 primary care Read codes. The eFI score is calculated by the presence or absence of each individual deficit as an equally weighted proportion of the total possible and can be used to identify frailty categories (not frail, mild frailty, moderate frailty, severe frailty), which predict risk of nursing home admission, hospitalisation and mortality. The eFI is supported in the 2016 NICE multimorbidity guideline and 2014 British Geriatrics Society & Royal College of General Practitioners Fit for Frailty Guideline [5, 6].

01/09/2020

5 years