Objectives: To compare the BBS scores between fallers and non-fallers in people with MS who use bilateral aids for gait. To investigate the sensitivity and specificity of the BBS to correctly identify fallers in this population.
Methods: BBS scores and self-reported falls were extracted from the baseline data of a RCT. Participants were recruited from the local branch of the MS society and had not had a relapse or commenced steroid treatment in the three months preceding the study. A faller was defined by having one or more falls in the three months prior to assessment. The differences between groups were tested with an independent t-test. ROC curve analysis was used to investigate sensitivity and specificity.
Results: There were 109 participants, 55 of whom (50.5%) reported falling. There was no significant difference between BBS scores in the faller (mean 28.2 SD10) and non faller (29.5 SD11) groups (p=0.879). A score less than 28 had only 59% sensitivity and 56% specificity for detecting fallers. A score less than 21 on BBS had 76% sensitivity and 80% specificity in detecting fallers. The area under the ROC curve was 0.496 (p=0.937).
Conclusions: These data confirm that the BBS score does not differ between fallers and non-fallers in people with MS who use a walking aid. It should be noted that this population had considerably lower mean scores than those in previous studies. Sensitivity and specificity of the BBS for identifying falls status only improve to above 75% accuracy when patients score less than 21 out of a possible 56. The BBS should not routinely be used to identify people with MS at risk of falls.