Reliability of Virtual Physical Performance Assessments in Veterans During the COVID-19 Pandemic
Elisa F. Ogawa, Rebekah Harris, Alyssa B. Dufour, Miriam C. Morey, Jonathan F. Bean
Abstract
OBJECTIVE: To determine the reliability of 3 physical performance tests performed via a telehealth visit (30-s arm curls test, 30-s chair stand test, 2-min step test) among community-dwelling older veterans. DESIGN: Cross sectional study. SETTING: Virtual. PARTICIPANTS: Veterans (N=55; mean age 75y) who enrolled in Gerofit, a virtual group exercise program. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Participants were tested by 2 different assessors at 1 time point. The intraclass correlation coefficient (ICC) with 95% confidence intervals and Bland-Altman plots were used as measures of reliability. To assess generalizability, ICCs were further evaluated by health conditions (type 2 diabetes, arthritis, obesity, depression). RESULTS: Assessments were conducted among 55 participants. The ICC was above 0.98 for all 3 tests across health conditions and Bland-Altman plots indicated that there were no significant systematic errors in the measurement. CONCLUSIONS: The virtual physical performance measures appear to have high reliability and the findings are generalizable across health conditions among veterans. Thus, they are reliable for evaluating physical performance in older veterans in virtual settings.