RESUMEN
OBJECTIVES: To assess the construct validity and responsiveness of the floor sitting-rising test (SRT) in individuals with total knee arthroplasty (TKA). DESIGN: Cohort study with 6-month follow-up. Secondary analysis using data from a randomized controlled trial. SETTING: An outpatient rehabilitation research center. PARTICIPANTS: Participants (N=240) enrolled in the parent study who had unilateral primary TKA. INTERVENTION: Participants in the parent study underwent 12 weeks of exercise programs. MAIN OUTCOME MEASURES: Validity analysis correlated baseline data of participants who completed the SRT and measures of knee motion, muscle strength, performance-based tests, and patient-reported outcomes of physical and psychosocial function. Responsiveness analysis used the 3- and 6-month follow-up data. Effect sizes were calculated using changes from baseline. Areas under the receiving operating characteristics curve were calculated using a global rating of change as the external anchor. RESULTS: Of the 240 participants (148 women; age, 70±7y), 180 (75%) were able to perform the SRT at baseline. Performers scored significantly better in all physical function tests (P<.0001) than nonperformers. SRT scores generally converged with measures of knee impairment and performance-based tests (associations ranged from small [r=0.15; P=.0516] to moderate [r=0.52; P<.0001]). SRT scores associated with self-efficacy for function (r=0.34; P<.0001) and fear of falls (r=-0.25; P=.001). At 3 (n=174) and 6 months (n=160), SRT effect sizes were 0.38 (95% confidence interval, 0.25-0.52) and 0.42 (95% confidence interval, 0.25-0.60), and areas under the curve were 0.59 (95% confidence interval, 0.49-0.69) and 0.62 (95% confidence interval, 0.52-0.73), respectively. CONCLUSIONS: The results add evidence to the validity of the SRT in patients after TKA. The strength of the associations suggests that the SRT measures a physical function construct not captured by the other tests. The magnitude of indices of responsiveness for the SRT were similar to other performance-based tests, indicating comparable responsiveness to more widespread tests of functional performance.
