Linking frailty instruments to the universal language of the International Classification of Functioning, Disability and Health (ICF)

Background: To date the major challenge underlying frailty is that there is no standardised language regarding the concept and operationalisation of frailty. Although the number of published articles containing the Medline MeSH term frail elderly has increased exponentially over the last 30 years, many papers still propose new frailty definitions and assessment tools. The International Classification of Functioning, Disability and Health (ICF) provides a universal language and framework for the characterisation of health and health-related states.
Objective: To identify and compile an extensive list of the available frailty instruments tested in populations of 65 years and older and to link their items to the codes of the ICF. To analyse the overlap and gaps present amongst the current frailty instruments and the domains of the ICF.
Method: PubMed, Web of Knowledge and PsycINFO were screened for relevant papers on frailty instruments using the keywords frailty, frail elderly, aged, aged 80 and over, assessment, risk assessment, classification, and diagnosis. The items of the identified frailty instruments were linked to the ICF codes via the ICF linking rules of Cieza ea. (2005) by 3 independent researchers.
Results: 1,984 potential articles were retrieved from which 67 relevant papers were retained and included in the systematic review. This resulted in the identification of 79 original or adapted frailty instruments, of which 27 use self-reported items to identify frailty, 6 use objective items and 44 use a combination of both. Only 25 frailty instruments clearly identify the pre-frail state. 24 frailty instruments are based on Fried Frailty Phenotype and consist solely of physical components, 1 frailty instrument is based solely on co-morbidities, and all other instruments consist of two or more domains (i.e. combination of medical, functional, cognitive, psychological and/or social components). 75 frailty instruments were linked to the ICF domain Body Functions, 19 frailty instruments were linked to Body Structure and to Capacity, 72 frailty instruments were linked to Performance, 14 frailty instruments were linked to Environmental factors and 32 frailty instruments were linked to Personal factors. The items of 2 frailty instruments were not specified and so could not be linked to the ICF codes.
Discussion: A huge diversity in the existing frailty instruments was observed. While the ICF-domains Body Function and Performance are well represented within the items of the frailty instruments, Body Structure, Capacity, Environmental and Personal factors are underrepresented. Strikingly, the majority of the frailty instruments distinguish robust from frail older persons, without taking into consideration the pre-frail state (subclinical intermediate state).

Conclusion: More than 70 different frailty instruments were identified, potentially creating a high diversity in clinical approach. The ICF-domains on Body Structure, Capacity and Environmental factors were poorly represented and might need more attention in a future consensus of frailty assessment.