Abstract
Purpose: The aim of this study was to compare score distributions and fall prediction accuracies between the Balance Evaluation Systems Test (BESTest), Mini-BESTest, Brief-BESTest, and Berg Balance Scale (BBS).
Methods: The BESTest, Mini-BESTest, Brief-BESTest, and BBS were performed for 57 subjects hospitalized with stroke and fracture. Balance tests were measured at discharge, and the incidence of falls was investigated 6 months after discharge. For each test, the ceiling effect and skewness were quantified. The receiver operating characteristic curve was used to identify area under the curve (AUC), sensitivity, and specificity for predicting falls.
Results: A ceiling effect and deviation in bias in the distribution were identified only on BBS. For fall prediction accuracy, AUC and sensitivity were highest with the Mini-BESTest, and specificity was highest with the BBS.
Conclusion: These results suggest the Mini-BESTest allows more detailed evaluation of balance ability than the BBS and offers moderate accuracy for predicting falls.
