CONTACT US: Contact info
|
Citation |
Hohtari-Kivimäki U, Salminen MJ, Vahlberg T, Kivelä SL. Aging Clin. Exp. Res. 2013; 25(6): 645-650. |
|
Affiliation |
Institute of Clinical Medicine, Family Medicine, University of Turku, Lemminkäisenkatu 1, 20014, Turku, Finland, ulanho@utu.fi. |
|
Copyright |
(Copyright © 2013, Holtzbrinck Springer Nature Publishing Group) |
|
DOI |
|
PMID |
24170328 |
|
Abstract |
BACKGROUND AND AIMS: The aim of this study was to assess the adequacy of the short, 9-item Berg Balance Scale (BBS-9) to predict fall risk among the community-dwelling aged. METHODS: The subjects (n = 519) were derived from the participants in a multifactorial fall prevention intervention conducted in Pori, Finland. Receiver operating characteristic (ROC) analysis was used to determine the cut-off score for BBS-9 (range 0-36) to classify aged people with a fall risk during a 12-month follow-up. Logistic regression was used to analyse the relationship of potential confounders with fall risk. The association between the cut-off score for BBS-9 and fall risk was tested using the Chi-square test. RESULTS: In determining the cut-off score of BBS-9 to classify fall risk, the highest sensitivity (0.51) and specificity (0.57) (when both presumed to be above 0.50) sum score was within the limit range 32 scores or below. The area under curve (AUC) was significantly better in the model adjusted for significant confounders (vision and the number of regularly used drugs) (AUC = 0.64) than in the unadjusted model (AUC = 0.57) (p = 0.045). Among patients who scored 32 or below in BBS-9 the incidence of multiple falls was 20.0 %, whereas among those who scored 33-36 it was 15.7 %. CONCLUSIONS: BBS-9 with the cut-off score of 32/33 together with data on vision and the number of regularly used drugs predicted moderately the risk of falling among the community-dwelling aged. Language: en |