Interrater agreement between telerehabilitation and face-to-face clinical outcome measurements for total knee arthroplasty.

Outcome measures in physical therapy provide the basis for determining the patient's rehabilitation needs, developing an individual intervention plan, and reassessing the evolution of the condition after therapeutic intervention. Questions surrounding the validity and reliability of outcome measures obtained in the context of telerehabilitation remain. The goal of this study was to explore which outcome measures can be used reliably in the context of telerehabilitation after discharge from an acute care hospital for lower limb orthopedic surgery. Fifteen patients recently discharged after total knee arthroplasty were evaluated by two experienced therapists. Each therapist evaluated under a given condition (face-to-face assessment, telerehabilitation assessment) eight outcome measures taken from standard clinical tests routinely used in the management of orthopedic rehabilitation after total knee arthroplasty. Evaluations were measured at 1-day intervals. Telerehabilitation evaluations were conducted with a videoconference link (H.264 CoDecs with Pan, Tilt, Zoom cameras) between either the participant's home or a clinical environment and a remote clinical station over residential DSL lines at 512 kbps. Interrater agreement between the two measurement modes was analyzed using the Bland and Altman method and Kripendorff's alpha reliability estimate. The 95% confidence interval for mean difference between evaluation methods varied between -20% and 8% for knee range of motion measures, -85% and 55% for scar management, -33% and 29% for functional evaluations. Five out of the eight outcome measures showed reliability estimates of >0.80, with lowest reliability obtained for the scar assessment scale (0.34) and the highest reliability for the evaluation of the range of motion at the knee (0.87 in flexion and 0.85 in extension). Clinical variables typically measured in face-to-face evaluations can be measured successfully under telerehabilitation conditions with moderate reliability.