Sensitivity of a modified version of the ‘timed get up and go’ test to predict fall risk in the elderly: a pilot study

The purpose of this study was to assess the sensitivity of a modified version of the ‘Timed Get Up and Go’ (TGUG) test in predicting fall risk in elderly individuals, using both a quantitative and qualitative approach in individuals older than 65 years. Ten subjects (83.4 ± 4.5 years) undertook the test twice. To assess inter-rater reliability, three investigators timed the two trials using a stopwatch (quantitative). The reproducibility of a qualitative evaluation of the trials was accomplished by the completion of an assessment questionnaire (AQ) at each trial by three investigators. To assess the agreement between the three investigators, the coefficients of reliability (CR), intra-class correlation coefficients (ICC) and limits of agreement were determined for the total time to do the test (TT). The weighted Kappa K of Cohen and ICC was calculated for the AQ. Inter-group comparison: 60 subjects (74.2 ± 4.9 years) were divided equally into four groups: (1) sedentary with previous history of falls, (2) sedentary without history of falls, (3) active with history of falls, and (4) active without history of falls. All of them undertook the modified TGUG test once. One investigator undertook the timing and completed the AQ. CR values for the TT were above 98% and with ICC of TT = 0.999. The differences in TT between the three investigators’ measures ranged from 0.19–0.55 s S.D. of the mean difference. Weighted Kappa K of Cohen ranged 0.835–0.976, with ICC of AQ = 0.954. Inter-group comparison study. Significant differences (p < 0.05) were noted between the mean score of TT of Group 4 and the rest of the groups, and between Groups 2 and 1. Significant differences (p < 0.05) were noted between the mean score of points obtained in the AQ of Group 1 and the rest of the groups. The modified version of the TGUG test demonstrated good sensitivity for detecting fall risk in elderly individuals, and good inter-tester reliability from both a quantitative and qualitative perspective.