Assessment of minor or moderate balance disorders: A reliability study and comparison with healthy subjects

Abstract

The purpose of the study was to examine the inter-rater, test–retest reliability and internal consistency of a new scale, the BDL Balance Scale, for assessing balance in patients with mild or moderate balance problems. An additional aim was to compare the results of assessed patients with the results of healthy subjects in order to demonstrate the discriminate ability of the scale. Patients (n = 30) aged 25–65 years and with neurological disorder and healthy subjects (n = 35) aged 30–65 years were assessed. The patients were assessed twice, by two different observers with 20 min in between, and the healthy subjects were assessed once. The reliability of the results obtained when testing the patients was calculated as intraclass correlation coefficient (ICC2.1), Cohen's kappa coefficient and Cronbach's alpha coefficient. The score differences were plotted in Bland–Altman graphs. The median total score was 24.0 points for the patients and 38.5 points for the healthy subjects (max 40 points). Cronbach's alpha coefficient was 0.89. The kappa coefficient for inter-rater reliability varied between 0.56 and 1.0, and ICC2.1 for the total score was 0.99. The kappa coefficients for test–retest reliability varied between 0.39 and 0.73, and ICC2.1 for the total scores was 0.96. The score between test occasions must be at least 5 points before the patients balance could be considered improved or deteriorated. The BDL Balance Scale has excellent to good intra-rater reliability and fairly good test–retest reliability and can differ between people with mild to moderate balance problems and healthy people of the same age. The sensitivity of the scales has to be further investigated.