ASSESSMENT OF THE COMMUNITY-DWELLING SENIOR FOR FALL PREVENTION: AN ANALYSIS OF AGREEMENT BETWEEN TWO COMMON PROTOCOLS OF THE TIMED UP AND GO TEST
MetadataShow full item record
Purpose:To determine if agreement exists between two common Timed Up and Go test (TUG) protocols (i.e., TUGP and TUGSC) in assessing community-dwelling seniors referred for home care physiotherapy services with balance and mobility impairments. Rationale:The TUG is recommended to assess balance and mobility as part of the assessment of falls prevention. Different testing protocols have emerged: i) for mobility: one practice trial plus one measured trial (TUGP), ii) for falls risk: one practice trial, plus three measured trials, averaged together (TUGSC). The TUG could be a valuable tool in fall prevention programs but variability in the protocols makes it challenging for researchers and clinicians to interpret and use literature findings. Knowledge of agreement between protocols can assist researchers and therapists in choosing, and interpreting, appropriate tests to include in falls prevention programs for seniors. Materials and Methods:Seniors, referred recruited for a Home Care Exercise Study  for balance problems, who completed baseline screening, including three experimental trials of the TUG.(N=19) Demographic variables were collected to describe the characteristics of the sample. Analysis: TUGP and TUGSC scores were calculated. Descriptive statistical analyses were performed on all variables. Four statistical methods were used to explore relationships between TUGP and TUGSC: Intra Class Correlation Coefficient(ICC(3,1)), Standard Error of Measurement (SEM), Spearman Rank Correlation (rho), and Limits of Agreement (LoA). Results: Study participants, mean age 82.6 years, presented with a median of four co-morbidities, varied ability to use vestibular inputs, varied balance confidence, but no apparent cognitive impairment. Thirteen participants had a fall history. Half of the sample did not use a gait aid; the other half used either a walker or a cane. The range of TUGp scores was 8.4-49.8 seconds (s), while TUGSC ranged from 8.7-56.4 s. The ICC (3, 1) equaled 0.99 (p < 0.05). The correlation analysis revealed rho=0.98 (p < 0.05), and a linear equation with a slope of 0.8, and intercept of 2.8. The LoA between the two protocols was 0.9 -1.2 s, and LoA plots revealed potential differences in TUG protocols for scores over 16 s. The SEM equaled 1.4 s and 1.7 s for TUGP and TUGSC, respectively, however, when outliers were removed, corresponding SEM values of 0.6 s and 0.5 s were observed. Conclusions: The ICC and Spearman Rank correlation indicated strong associations between the protocols. The line-of-best-fit analysis illustrated the two TUG protocols were interchangeable for those performing the test in less than 16 s. According to Bland-Altman analysis the scores obtained with TUGP exceeded those of the TUGSC by approximately one second, which was not a meaningful difference according to the minimal detectable difference. Agreement between protocols was apparent for those performing the TUG quickly, but did not hold true for the whole range of TUG scores.