Segurança e reprodutibilidade do teste timed up and go em idosos hospitalizados

Abstract

Introduction: Safe and reliable physical tests are important to identify risk of falling in the elderly. However, there are no studies that have evaluated the safety and reliability of the test Timed Up and Go (TUG) in hospitalized elderly. Objectives: To evaluate the safety and reproducibility of the TUG in hospitalized elderly. Methods: This is a cross-sectional study where three measurements of the TUG were collected for each elderly, being considered the best performance in seconds (sec). Other variables collected were cognitive function (MMSE), Charlson comorbidity index, admission profile (clinical or surgical), history of falls in the last year and BMI. To assess the relative reliability we used the intraclass correlation coefficient (ICC), and for the absolute reliability we used the Bland-Altman analysis. Results: The study included 68 elderly with a mean age 70.4 ± 7.7 years, Charlson index 5.4 ± 2.0 and prevalence of clinical profile (64.7%). None of the 204 tests was interrupted by the established criteria. There was a gradual reduction between the first and third measurement (1st = 11.6 ± 6.54; 2nd = 10.7 ± 6.22 and 3rd = 10.3 ± 5.54; p = 0.001) and high ICC (1st and 2nd: ICC = 0.98; 1st and 3rd: ICC = 0.98; 2nd and 3rd: ICC = 0.98; p = 0.001), and the highest correlation with the best performance was associated with third measurement (ICC = 0.99; p = 0.001). We identified that the smallest bias (0.29 sec) and limits of concordance (-1.1 to 1.68 sec) also occurred between the third measurement and the best performance. The measure of the method error to evaluate the variability was 0.5 sec, and the clinically significant change was 3.4 sec. Conclusion: The TUG was a safe instrument with good reproducibility for measuring physical performance in hospitalized elderly.