Abstract

Background: Patients with chronic kidney disease (CKD) are at high risk of frailty, leading to adverse outcomes. The Taiwanese version of the Tilburg Frailty Indicator (TFI-T) has been validated in communitydwelling older adults, but its psychometric properties have not been assessed in CKD patients. This study evaluates the reliability and validity of the Taiwan revised version of the Tilburg Frailty Indicator (TFI-TR) for this population.

Methods: The cross-sectional design was conducted for outpatients with nephrology in Taiwan. Five stepswere taken to test the TFI-TR. First, the modified Delphi method was performed to revise the item content of TFI-T. Second, item analysis was used to examine item determination. Third, construct validity was examined using confirmatory factor analysis (CFA), Fried frailty phenotype (FP), and Kihon Checklist (KCL) for criterion validity. Fourth, the reliability was explored using the KR-20 reliability coefficient. Finally, the ROC curve was used to determine the cut points of TFI-TR.

Results: A 15-item CKD-specific version of the TFI-T (TFI-TR) with three domains was developed. Item analysis showed good discrimination. CFA demonstrated good construct validity (TLI = 0.88; CFI = 0.90; RMSEA = 0.043), and criterion validity was supported by moderate correlations with FP and KCL (r = 0.45, 0.62). The KR-20 reliability coefficient was 0.70. A cut-off score of 4.5 yielded good sensitivity and specificity (AUC = 0.86, 95% CI = 0.78–0.94).

Conclusions: The TFI-TR demonstrates acceptable validity and reliability for screening frailty among CKD outpatients. A cut-off of 4.5 provides good diagnostic performance.