Abstract

Background

This study aimed to compare the diagnostic yield of the FRAIL scale with respect to the physical frailty phenotype measure and their association with mortality in non-dialysis-dependent patients.

Methods

In this prospective cohort study, non-dialysis dependent patients with chronic kidney disease (CKD) stages 3b–5 seen in the nephrology outpatient clinics of two university hospitals were included. The presence of frailty was evaluated by physical frailty phenotype measure and the FRAIL scale. Patients were evaluated for six months, and mortality was recorded. The Kappa test was used to evaluate the diagnostic properties between the methods, and logistic regression to test the association between frailty and mortality.

Results

One hundred fifty-three patients were evaluated; average age was 65 (56–70) years, 50.9% were women, and the all-cause mortality rate was 2.6%. Forty-six patients were classified as living with frailty according to the physical frailty phenotype while 36 patients were rated frail by the FRAIL scale. In adults < 60 years of age, the FRAIL scale showed good accuracy (84.9%) and specificity (93.2%) but had low sensitivity (41.3%) and moderate agreement (Kappa = 0.41; p < 0.001) compared to the definition of the physical frailty phenotype. The adjusted logistic regression model showed that the patients with frailty assessed by the FRAIL scale had a greater chance of mortality than the non-frail patients (OR: 6.8; CI95%:1.477–31.513; p = 0.014).

Conclusion

Physical frailty phenotype identifies more patients as having pre-frailty and frailty in non-dialysis dependent patients as compared to the FRAIL scale. However, the FRAIL scale is a simple bedside tool that can be useful for screening for frailty and whose results were associated with mortality.

Graphical abstract