File
Viveiro LAP1, Gomes GCV1, Bacha JMR1, Carvas Junior N1, Kallas ME2, Reis M2, Jacob Filho W2, Pompeu JE1
1University of Sao Paulo, Department of Physical Therapy, Speech and Occupational Therapy, School of Medicine, Sao Paulo, Brazil, 2University of Sao Paulo, Department of Geriatric, School of Medicine, Sao Paulo, Brazil
Background: In nursing home environments, the annual rate of falls is 30-50%. There is no consensus about which assessment tool is the most useful and presents the best sensitivity and specificity to identify the risk of falls and predict falls in older adults.
Purpose: To verify and compare the ability to identify and predict falls of the Berg Balance Scale (BBS), Balance Evaluation Systems Test (BESTest), Mini-BESTest, and Brief-BESTest in older adults who live in nursing home.
Methods: This was a cross-sectional study. Older adults were recruited from a nonprofit nursing home. All participants were assessed through BBS, BESTest, Mini-BESTest, and Brief-BESTest. The history of falls during the previous year was collected. The occurrence of falls during the follow-up of six (6m) and twelve (12m) months after assessment was also collected. To analyze the ability of each test to identify fall status and predict falls, we used receiver operating characteristic (ROC) curves, whose statistical significance we verified using the area under the ROC curve (AUC) and respective 95% confidence intervals (CI). We determined the cut-off points to identify falls status (history of falls) and predict falls (6m and 12m) by the Youden Index. To verify the prediction of falls, we calculated the risk ratio for each balance test after 6m and 12m of follow-up. The 95%CI of risk ratios were obtained from the Wald's method and the significance of the association was assessed by Fischer's exact test. The statistical analysis was done using IBM SPSS Statistics (version 20). We used a level of significance of p 0.05.
Results: Of the 49 participants in our sample, 30 were women (61.2%) and 19 were men (38.8%). Their mean age (SD) was 77.8 (7.2) years. Thirty-one individuals (63.2%) were categorized as non-fallers according to the history of falls, while 18 (36.8%) were fallers. After the follow-up, a total of 20 (40.8%) individuals felt and 8 (16.3%) died. The AUCs of the BBS, BESTest, Mini-BESTest, and Brief-BESTest ranged from 0.712-0.762 (history of falls), 0.787-0.810 (6m), and 0.807-0.819 (12m), indicating an acceptable ability to identify and predict falls. The sensitivity of falls prediction for 12m of the Brief-BESTest (85.0%) and BBS (75.0%) were higher than the BESTest (70.0%) and Mini-BESTest (70.0%). However, the specificity was higher for the Mini-BESTest (85.7%). We found the following cut-offs of falls prediction for 12m for the balance tests: 42 for the BBS, 47 for the BESTest, 10 for the Mini-BESTest, and 11 for the Brief-BESTest. The risk ratios for 12m of follow-up were 3.63, 2.85, 3.29, and 4.01, for the BBS, BESTest, Mini-BESTest, and Brief-BESTest.
Conclusion(s): All balance tests were able to identify and predict falls in older adults who live in nursing home.
Implications: It is extremely important to identify balance issues and fall probabilities among older adults. To know the diagnostic properties of the tests can help physiotherapists to choose the best assessment tool for clinical practice.
Keywords: Postural balance, accidental falls, aged
Funding acknowledgements: We are grateful to Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the Ph.D. scholarship.
Purpose: To verify and compare the ability to identify and predict falls of the Berg Balance Scale (BBS), Balance Evaluation Systems Test (BESTest), Mini-BESTest, and Brief-BESTest in older adults who live in nursing home.
Methods: This was a cross-sectional study. Older adults were recruited from a nonprofit nursing home. All participants were assessed through BBS, BESTest, Mini-BESTest, and Brief-BESTest. The history of falls during the previous year was collected. The occurrence of falls during the follow-up of six (6m) and twelve (12m) months after assessment was also collected. To analyze the ability of each test to identify fall status and predict falls, we used receiver operating characteristic (ROC) curves, whose statistical significance we verified using the area under the ROC curve (AUC) and respective 95% confidence intervals (CI). We determined the cut-off points to identify falls status (history of falls) and predict falls (6m and 12m) by the Youden Index. To verify the prediction of falls, we calculated the risk ratio for each balance test after 6m and 12m of follow-up. The 95%CI of risk ratios were obtained from the Wald's method and the significance of the association was assessed by Fischer's exact test. The statistical analysis was done using IBM SPSS Statistics (version 20). We used a level of significance of p 0.05.
Results: Of the 49 participants in our sample, 30 were women (61.2%) and 19 were men (38.8%). Their mean age (SD) was 77.8 (7.2) years. Thirty-one individuals (63.2%) were categorized as non-fallers according to the history of falls, while 18 (36.8%) were fallers. After the follow-up, a total of 20 (40.8%) individuals felt and 8 (16.3%) died. The AUCs of the BBS, BESTest, Mini-BESTest, and Brief-BESTest ranged from 0.712-0.762 (history of falls), 0.787-0.810 (6m), and 0.807-0.819 (12m), indicating an acceptable ability to identify and predict falls. The sensitivity of falls prediction for 12m of the Brief-BESTest (85.0%) and BBS (75.0%) were higher than the BESTest (70.0%) and Mini-BESTest (70.0%). However, the specificity was higher for the Mini-BESTest (85.7%). We found the following cut-offs of falls prediction for 12m for the balance tests: 42 for the BBS, 47 for the BESTest, 10 for the Mini-BESTest, and 11 for the Brief-BESTest. The risk ratios for 12m of follow-up were 3.63, 2.85, 3.29, and 4.01, for the BBS, BESTest, Mini-BESTest, and Brief-BESTest.
Conclusion(s): All balance tests were able to identify and predict falls in older adults who live in nursing home.
Implications: It is extremely important to identify balance issues and fall probabilities among older adults. To know the diagnostic properties of the tests can help physiotherapists to choose the best assessment tool for clinical practice.
Keywords: Postural balance, accidental falls, aged
Funding acknowledgements: We are grateful to Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the Ph.D. scholarship.
Topic: Older people
Ethics approval required: Yes
Institution: School of Medicine at the University of Sao Paulo
Ethics committee: Research Ethics Committee
Ethics number: 1.818.309/2016
All authors, affiliations and abstracts have been published as submitted.
