Preferred Outcome Measures Used in Randomized Clinical Trials of Total Knee Replacement Rehabilitation: A Systematic Review

• Published in: PM & R Vol. 12; no. 7; pp. 706 - 713
• Main Authors: Blasco, José M., Acosta‐Ballester, Yolanda, Igual‐Camacho, Celedonia, Hernández‐Guillén, David, Gómez, María C., Roig‐Casasús, Sergio, Puigcerver‐Aranda, Pablo
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01-07-2020
• ISSN: 1934-1482; 1934-1563
• DOI: 10.1002/pmrj.12312

Objective
To determine the most frequently used outcome measures in total knee replacement rehabilitation trials.
Literature Survey
Systematic review of randomized trials searched in five databases: Web of Science, MEDical Literature Analysis and Retrieval System, Physiotherapy Evidence Database, Scopus, and Cochrane Library.
Methodology
Trials were included if participants underwent total knee replacement rehabilitation and outcome measures were used to assess rehabilitation outcomes. A descriptive synthesis determined the frequency of using outcome measures and preferred assessment time points. Outcomes were classified into eight categories: patient‐ and clinician‐reported function, performance‐based function, balance, anxiety and depressive symptoms, quality of life, and others.
Synthesis
Eighty‐one trials were included and 102 different outcome measures were classified. The most frequently reported outcome was knee range of motion, used in 54% of trials, followed by a visual analog scale of pain (43%) and Western Ontario and McMaster Universities Arthritis Index (WOMAC; 40%). Patient‐ and clinician‐reported function were the categories most frequently assessed (74%), whereas performance‐based measures were implemented by 56% of trials. The most frequent assessment time points were 1 week presurgery (52%) and 3 months postsurgery (39%).
Conclusions
There is consensus regarding the need to evaluate functional outcomes in total knee replacement rehabilitation trials but none regarding the outcome measure that should be used. These findings suggest that most trials include patient‐ and clinician‐reported functional measures, along with pain and performance‐based measures in trial designs.