Commonly used screening instruments to identify frailty among community-dwelling older people in a general practice (primary care) setting: a study of diagnostic test accuracy

Abstract

BACKGROUND:Rapid frailty screening remains problematic in primary care. The diagnostic test accuracy (DTA) of several screening instruments has not been sufficiently established. We evaluated the DTA of several screening instruments against two reference standards: Fried's Frailty Phenotype [FP] and the Adelaide Frailty Index [AFI]), a self-reported questionnaire. METHODS:DTA study within three general practices in South Australia. We randomly recruited 243 general practice patients aged 75+ years. Eligible participants were 75+ years, proficient in English and community-dwelling. We excluded those who were receiving palliative care, hospitalised or living in a residential care facility.We calculated sensitivity, specificity, predictive values, likelihood ratios, Youden Index and AUC for: Edmonton Frail Scale [EFS], FRAIL Scale Questionnaire [FQ], Gait Speed Test [GST], Groningen Frailty Indicator [GFI], Kihon Checklist [KC], Polypharmacy [POLY], PRISMA-7 [P7], Reported Edmonton Frail Scale [REFS], Self-Rated Health [SRH] and Timed Up and Go [TUG]) against FP [3+ criteria] and AFI [> 0.21]. RESULTS:We obtained valid data for 228 participants, with missing scores for index tests multiply imputed. Frailty prevalence was 17.5% frail, 56.6% pre-frail [FP], and 48.7% frail, 29.0% pre-frail [AFI]. Of the index tests KC (Se: 85.0% [70.2 - 94.3]; Sp: 73.4% [66.5 - 79.6]) and REFS (Se: 87.5% [73.2 - 95.8]; Sp: 75.5% [68.8 - 81.5]), both against FP, showed sufficient diagnostic accuracy according to our pre-specified criteria. CONCLUSIONS:Two screening instruments - the KC and REFS, show the most promise for wider implementation within general practice, enabling a personalised approach to care for older people with frailty.