Is there any value in looking for frailty on a population or practice basis?
Systematic screening for frailty would be an expensive venture and there is currently no evidence for improved outcomes despite it being a recommendation in earlier international guidance14. Like systematic screening for dementia, there would be a degree of “public unacceptability” (for example; people may be fearful of being diagnosed with dementia and therefore be reluctant to submit to a test for dementia unless it was specifically indicated by their life circumstances). Age UK research 4 has shown that in a series of filmed case studies of ‘frailty’, none of the participants classified themselves as “frail”. Some of them mentioned finite periods where they “had been frail”, but they did not see it as a lifetime condition or as defining them.
A current approach seeks to break down a practice population according to risk of using future health care resources including hospital admission. It uses computer based tools, for example Advanced Clinical Groupings (ACG), Prediction of individuals At Risk of Readmission (PARR) or Scottish Prevention of Admission and Readmission (SPARRA). These tools interrogate a primary care practice computer to identify high risk individuals based on past use of resources, drug prescriptions or particular diagnoses. Unfortunately there is no evidence that focussing resources on these individuals improves outcomes. Additionally, these tools, which were not designed to look for frailty, often highlight individuals who have high cost conditions not amenable to modification (such as immunosuppression after organ transplant).
Some areas and practices have adopted a localised approach to identify frailty, e.g. in Warwickshire, Age UK has trained volunteers to administer the Easy care tool23 which starts the process of identifying needs and developing an individualised care plan. This is similar to an approach in Gnosall, Staffordshire (winner of an NHS innovation award) where everyone receives a questionnaire on their 75th birthday, seeking to identify those who might have, or be developing, frailty. They have achieved a response rate of over 85% and those who respond are then visited at home by an elder care facilitator before undergoing a comprehensive geriatric assessment at the surgery by a GP.
Thus the BGS does not currently support routine population screening for frailty because of the likely considerable cost of completing assessments and the low specificity of available tools. A suitably validated electronic frailty index constructed using existing primary care health record data may enable future routine identification and severity grading of frailty, but requires additional research.