Substitution of Fried's performance-based physical frailty criteria with self-report questions

Objective: To identify self-report questions that can substitute Fried's performance-based frailty measures for use in large-scale studies and daily practice. Methods: A cross-sectional study was conducted among community dwelling older people (65 +). Based on a literature search and interviews with older people and experts, 11 questions concerning walk time and 10 on handgrip strength were selected. All participants completed these sets of self-report questions as well as the original Fried criteria (including performance-based tests). Regression analyses were performed to find the questions that best substituted the performance-based tests. Results: In total, 135 individuals (mean age 73.8 +/- 7.0, 58.5% female) in different stages of frailty (non-frail 38.5%, pre-frail 40.7%, frail 20.7%) were included. Regression analyses revealed four questions for walk time and two for handgrip strength. Cut-off values of three for walk time (range 0-5) and one for handgrip strength (range 0-3) seem most optimal. This resulted in a sensitivity of 69.2%, 86.1% specificity and 79.4% agreement for walk time and a sensitivity of 73.2%, 71.3% specificity and 71.9% agreement for handgrip strength. The comparison of frailty stages using frailty criteria including the performance-based measures and scores based solely on self-report questions, resulted in an observed agreement of 71.1% (kappa value = 0.55). Conclusions: Considering the agreement between the questions and the performance-based tests, these two sets of questions might be used in settings where the performance-based tests of walk time and handgrip strength are unfeasible, such as in daily practice and large-scale research.