Predictive Performance of the FIF Screening Tool in 2 Cohorts of Community-Living Older Adults

Objectives: The First-time Injurious Fall (FIF) screening tool was created to identify fall risk in community-living older adults who may benefit from primary preventive interventions. The aim of this study was to evaluate the predictive performance of the FIF tool in 2 cohorts of older adults. Design: Longitudinal cohort study. Setting and Participants: The Swedish National Study on Aging and Care in Skåne (SNAC-S) and Blekinge (SNAC-B), Sweden. Community-living people aged ≥60 years (n = 2766). Methods: Nurses and physicians collected data in the 2 cohorts through interviews and testing. Data on injurious falls were collected from register data and were defined as receipt of care after a fall. The FIF tool, consisting of 3 questions and 1 balance test, was examined in relation to injurious falls for up to 5 years of follow-up using Cox proportional hazards models. The predictive performance of the FIF tool was further explored using Harrell C statistic and Youden cut-off for sensitivity and specificity. Results: The hazard ratios (HRs) of an injurious fall in the high-risk group for women and men were 3.80 (95% confidence interval [CI] 2.53, 5.73) and 5.10 (95% CI 2.57, 10.12) in SNAC-S and 4.45 (95% CI 1.86, 10.61) and 32.58 (95% CI 4.30, 247.05) in SNAC-B compared with those in the low risk group. The sensitivity and specificity of the Youden cut-off point (3 or higher for high-risk) were 0.64 and 0.69 for women and 0.68 and 0.69 for men in SNAC-S, and 0.64 and 0.74 for women and 0.94 and 0.68 for men in SNAC-B. The predictive values (Harrell C statistic) for the scores for women and men were 0.73 and 0.74 in SNAC-S and 0.72 and 0.89 in SNAC-B. Conclusions and Implications: Our results suggest that the FIF tool is a valid tool to use for prediction of first-time injurious falls in community-living older adults