Abstract
Purpose: The purpose of this study was to examine the predictive accuracy of measures for falls in people with stroke.
Methods: A systematic literature search was performed in the database PubMed and ICYUSHI. Articles were selected according to the following inclusion criteria: (1) study subjects were people with stroke, (2) study design was prospective data collection of falls, (3) sensitivity and specificity were clearly reported in results. Positive likelihood ratio was used to examine the predictive accuracy of measures for falls.
Results: Seven articles matched the inclusion criteria. Measures for falls met the higher positive likelihood ratio (> 5) were the Berg Balance Scale (BBS), the Step Test, the Stops Walking When Talking test (SWWT), the BBS with SWWT, and the Stroke Impact Scale-16. The most accurate predictor of falls was the BBS with SWWT (positive likelihood ratio = 10.7).
Conclusion: These findings suggest that the BBS with SWWT can be the best predictor of falls in people with stroke.
