RELIABILITY AND RESPONSIVENESS OF PHYSICAL ACTIVITY MEASURES IN INDIVIDUALS AFTER TOTAL KNEE ARTHROPLASTY

Few instruments that measure physical activity (PA) can accurately quantify PA performed at both light and moderate intensities, which is particularly relevant in older adults. Evidence of their reliability and responsiveness to change is limited. The purposes of this study were to: 1) determine test-retest reliability of the Actigraph (ACT), SenseWear Armband (SWA) and the Community Healthy Activities Model Program for Seniors (CHAMPS) questionnaire in assessing free-living PA; and 2) determine the responsiveness to change in PA measured by the three instruments after an exercise program in individuals wih knee osteoarthritis who underwent total knee arthroplasty (TKA). \ud Test-retest reliability was determined by asking subjects to wear the activity monitors for two weeks and complete the CHAMPS at the end of each week. Reliability was measured using intra-class correlation coefficient (ICC2,k). For responsiveness, subjects wore the activity monitors for one week at baseline and at 6-months, and completed the CHAMPS at the end of each week. Changes in PA and standardized response mean (SRM) were calculated and compared across instruments. Weighted-Kappa (K) was used to determine agreement between the instruments on identifying changes in PA based on measurement error.\ud Test-retest reliability ranged from moderate to excellent for ACT (ICC=0.75-0.86), and were excellent for SWA (ICC=0.93-0.95) and CHAMPS (ICC=0.86-0.92) in the subjects who reported similar PA during two-weeks. The 95% confidence interval of the ICCs from SWA was the only one within excellent reliability range (0.85-0.98). Results on responsiveness revealed small changes in PA (p>0.05) measured by each instrument after the intervention. SRMs indicated small degree of responsiveness (SRM