Validity and Accuracy of the Tilburg Frailty Indicator Part B for Identification of Frailty in Older Adults Consulting a Rural Geriatric Medicine Clinic

• Published in: Canadian journal on aging 2023-09, Vol.42 (3), p.466-474
• Main Authors: Mehta, Saurabh P., Indramohan, Pavithramohan, Dobariya, Varun, Seccurro, Drake, Goebel, Lynne J.
• Format: Article
• Language: English
• Subjects: Accuracy; Exercise; Frail; Frailty; Gait; Genotype & phenotype; Geriatrics; Medical diagnosis; Medicine; Older people; Quality of life; Rural communities; Rural health care; Self report; Validity
• ISSN: 0714-9808; 1710-1107
• DOI: 10.1017/S0714980823000077

The Tilburg Frailty Indicator (TFI) is a validated tool for determining frailty in older adults. This study examined the validity and accuracy of the TFI Part B (TFI-B) in a North American context. Seventy-two individuals ≥ 65 years of age recruited from a rural geriatric medicine clinic completed a set of self-reported and performance-based measures, including TFI-B. Frailty level was determined using modified Fried’s Frailty Phenotype (FFP). Pearson correlation coefficients (r) assessed the concurrent relationships between the TFI-B and other measures. Accuracy of the TFI-B in classifying frailty level was assessed using assessing area under the curve (AUC). The TFI-B scores showed low correlations (r < 0.4) with gait speed and grip, suggesting that the TFI-B did not consider frailty as merely a physical problem. The AUC of 0.82 indicated that the TFI-B scores accurately classified frail versus non-frail individuals. The score of ≥ 5 on the TFI-B scores showed satisfactory sensitivity/specificity (73%/77%) and excellent negative predictive value (91.95%). This indicates that a TFI-B score of < 5 can be used to rule out frailty.