Diagnostic test accuracy of self-reported and self-administered screening instruments to identify frailty among older, community-dwelling populations: a systematic review

Aims: We aimed to determine the diagnostic test accuracy (DTA) of self- reported and/or self- administered screening instruments to identify frailty within community- dwelling older adult populations.
Methods: We included observational studies within community settings comparing a self- reported screening in-strument against a frailty reference standard (the Frailty Phenotype or Frailty Index). The diagnosis of interest was frailty. We required that at least half of the study partici-pants were aged 65 years or over, or have a minimum mean age of 65 years. Studies were excluded if conducted within acute settings, participants had specific diagnoses or were living in residential care. We conducted a systematic search of the literature from the inception of the databases to July 2018. Included studies were assessed for quality against the JBI Critical Appraisal Checklist for Diagnostic Test Accuracy Studies. We specified a minimum sensitivity threshold of 80% and specificity threshold of 60% as an ap-propriate threshold to identify highly accurate instruments.Results: Of an initial 7,455 unique studies, we identified 24 that met our selection criteria. Four self- reported screening instruments met minimum sensitivity and specificity thresh-olds, but study design impacted on the generalisability of the results. Meta- analysis was not conducted due to insufficient numbers of studies analysing individual index tests against the two reference standards.
Conclusions: The current evidence for the DTA of many frailty screening instruments currently in wide use is inad- equate. Further well-designed DTA studies of self- reported screening instruments to identify frailty are required to inform health policy and practice decision- making within community settings.