Psychometric testing of the fall risks for older people in the community screening tool (frop-com screen) for community-dwelling people with stroke

Abstract

Objective: The Falls Risk for Older People in the Community assessment (FROP-Com) was originally developed using 13 risk factors to identify the fall risks of community-dwelling older people. To suit the practical use in busy clinical settings, a brief version adopting 3 most fall predictive risk factors from the original FROP-Com, including the number of falls in the past 12 months, assistance required to perform domestic activities of daily living and observation of balance, was developed for screening purpose (FROP-Com screen). The objectives of this study were to investigate the inter-rater and test-retest reliability, concurrent and convergent validity, and minimum detectable change of the FROP-Com screen in community-dwelling people with stroke. Participants: Community-dwelling people with stroke (n = 48) were recruited from a local self-help group, and community-dwelling older people (n = 40) were recruited as control subjects. Results: The FROP-Com screen exhibited moderate inter-rater (Intraclass correlation coefficient [ICC](2,1) = 0.79, 95% confidence interval [CI]: 0.65-0.87) and test-retest reliability (ICC3,1 = 0.70, 95% CI: 0.46-0.83) and weak associations with two balance measures, the Berg Balance Scale (BBS) (rho = -0.38, p = 0.008) and the Timed "Up & Go" (TUG) test (rho = 0.35, p = 0.016). The screen also exhibited a moderate association with the Chinese version of the Activities-specific Balance Confidence Scale (ABC-C) (ABC-C; rho = -0.65, p < 0.001), a measure of subjective balance confidence. Conclusions: The FROP-Com screen is a reliable clinical tool with convergent validity paralleled with subjective balance confidence measure that can be used in fall risk screening of communitydwelling people with stroke. However, one individual item, the observation of balance, will require additional refinement to improve the potential measurement error.