Abstract
Background and Aims:
Falls are common in the elderly population. The Berg Balance Scale (BBS), Timed Up and Go Test (TUG), and Five Times Sit-to-Stand Test (5xSTS) are evidence-based clinical performance-based measurements of the risk of falls among elderlies. This study aimed to compare the validity of these assessment tools in predicting falls in community-dwelling elders in Saudi Arabia. Moreover, the cut-off point, sensitivity, specificity, and post-test accuracy percentage for these assessment tools were determined.
Methods:
One hundred fourteen ambulatory community-dwelling individuals aged 65 or older were recruited for this study. Demographic information and a history of falling were collected before all participants were assessed with BBS, TUG, and 5xSTS.
Results:
The BBS was sensitive to group differences between fallers and non-fallers (P = 0.0001). Results from the receiver operating characteristic curve analyses showed that the accuracies of the BBS and TUG at classifying participants with and without a fall history were high-to-moderate (area under the curve [AUC], 0.73 and 0.71, respectively). In contrast, the accuracy of the 5xSTS for the same was low (AUC, 0.67). The cut-off points for the BBS, TUG, and 5xSTS were 42, 20, and 34 seconds, respectively. The highest post-test accuracy was that of the BBS at 84.39%, followed by that of the TUG at 68.05%, and 5xSTS had the lowest accuracy at 25.29%.
Conclusions:
The BBS was a sensitive tool to distinguish fallers among community dwelling elderly as compared to TUG and 5xSTS tests.

Key words: Falls, Berg Balance Scale, Risk assessment tools, Timed Up and Go test, Five Times Sit-to-Stand test, Predictive fall