Discriminant accuracy of standing balance tests for level of gait dependency in hospitalized patients with Alzheimer's disease.

INTRODUCTION When determining the level of gait independence in patients with Alzheimer's disease (AD), detailed functional assessment is difficult in some patients. Previous literature has suggested simple standing balance tests for patients with AD due to their ease of implementation in clinical practice and relevance to gait. However, their usefulness for discriminating the level of gait independence remains unclear. This study aimed to investigate the discrimination accuracy of a simple standing balance test in the level of gait independence among hospitalized patients with AD. METHODS This cross-sectional study was a post-hoc analysis of a study conducted on 63 inpatients with AD in a single hospital. Participants were divided into three groups according to their level of gait independence: independent, modified independent (independent walking with walking aids), and dependent groups (supervision). Gait independence was determined using the Functional Independence Measure. Four standing balance tests were used-closed-leg, semi-tandem, tandem, and one-leg standings-and the discrimination accuracy of each test was calculated by receiver operating characteristic analysis. RESULTS One-leg standing was best at discriminating between the independent and modified independent groups (positive predictive value=80.0%, negative predictive value=94.1%). Tandem standing was best at discriminating between the modified independent and dependent groups (positive predictive value=74.1%, negative predictive value=93.3%). CONCLUSION A simple standing balance test may assist in determining level of gait independence in patients with AD when it is difficult to perform a mobility assessment.