Interrater reliability and validity of the stair ascend/descend test in subjects with total knee arthroplasty.

Abstract

(1) To determine the interrater reliability and measurement error of an 11-step stair ascend/descend test (STTotal-11) and stair up (ascend) test (STUp-11); (2) to seek evidence for the STTotal-11 and STUp-11 as valid measures of physical function by determining if they relate to measures of physical function and do not relate to measures not of physical function; and (3) to explore if the STTotal-11 and STUp-11 scores relate to lower-extremity muscle weakness and knee range of motion (ROM) in subjects with total knee arthroplasty (TKA). Cross-sectional study. Academic center. Subjects (N=43, 30 women; mean age, 68+/-8y) with unilateral TKA. Not applicable. STTotal-11 and STUp-11 were performed twice, and scores were compared with scores on 4 lower extremity performance-based tasks, 2 patient-reported questionnaires of physical function, 3 psychologic factors, knee ROM, and strength of quadriceps, hip extensors, and abductors. Intraclass correlation coefficient was .94 for both the STTotal-11 and STUp-11, standard error of measurements were 1.14 seconds and .82 seconds, and minimum detectable change associated with 90% confidence interval was 2.6 seconds and 1.9 seconds, respectively. Correlations between stair tests and performance-based measures and knee and hip muscle strength ranged from Pearson correlation coefficient (r)=.40 to .78. STTotal-11 and STUp-11 had a small correlation with one of the patient-reported measures of physical function. Stair tests were not associated with psychologic factors and knee extension ROM and were associated with knee flexion ROM. STTotal-11 and STUp-11 have good interrater reliability and minimum detectable changes adequate for clinical use. The pattern of associations supports the validity of the stair tests in TKA.