Quantcast

Psychometric properties of the German version of the Fear of Falling Questionnaire-revised (FFQ-R) in a sample of older adults after hip or pelvic fracture.

Research paper by Anja A Dautel, Michaela M Gross, Bastian B Abel, Rebekka R Pomiersky, Tobias T Eckert, Klaus K Hauer, Martina M Schäufele, Gisela G Büchele, Clemens C Becker, Klaus K Pfeiffer

Indexed on: 28 Jul '20Published on: 25 Jul '20Published in: Aging Clinical and Experimental Research



Abstract

Identifying patients with maladaptive fear of falling (FOF) is important in the rehabilitation phase after serious fall. The 6-item Fear of Falling Questionnaire-revised (FFQ-R) was seen as promising measurement instrument as it evaluates FOF in a broader way than the one-item-question and independent of physical activities. The purpose of the analysis was to evaluate the psychometric properties of the translated German FFQ-R. Back-translation method was applied. Confirmatory factor analysis (CFA) with diagonally weighted least square estimation was used to verify the two-factor structure. Data were collected during inpatient rehabilitation from hip and pelvic fracture patients [age 84.3 ± 6.2, Mini-Mental State Examination (MMSE) scores > 23] participating in an RCT (N = 112) and a cross-sectional survey (N = 40). Internal consistency was 0.78 (Cronbach´s alpha). No floor or ceiling effects were found. Discriminatory power on item level was moderate to good (r = 0.43-0.65). CFA revealed a good model fit and confirmed the two-factor structure. The German FFQ-R was moderately correlated (r = 0.51) with the Short Falls Efficacy Scale-International (Short FES-I) used as a proxy measure for FOF. Missing rates up to 9% for specific items were because some individuals, independent of cognitive level or age, had problems to rate items with conditional statements on possible negative consequences of a fall. Results demonstrated moderate to good psychometric properties similar to the original English version in a comparable sample of fracture patients.

More like this: