Supervised versus unsupervised rehabilitation following total knee arthroplasty: A systematic review and meta-analysis

Background: Outcomes after total knee arthroplasty (TKA) are strongly influenced by the adequacy of rehabilitation and the consequent functional recovery. The economic impact of rehabilitation it is not negligible. Inpatient rehabilitation can be 5 to 26 times more expensive than the home-based rehabilitation. This topic is extremely relevant as the COVID-19 pandemic has highlighted the importance of unsupervised rehabilitation in orthopedic surgery. The aim of this review and meta-analysis is to investigate the scientific evidence regarding the comparison between supervised and unsupervised rehabilitation following TKA. Materials and Methods: Following PRISMA guideline, a comprehensive search of PubMed, Cochrane and Scopus databases using combinations of keywords and MeSH descriptors: ``total ``Knee replacement," ``Arthroplasty", ``Rehabilitation" was performed from inception to December 2021. All relevant articles were retrieved, and their bibliographies were searched for further relevant references. Only English written randomized controlled trials comparing supervised and unsupervised rehabilitation following TKA were included in this systematic review. The outcomes considered were long-term pain, physical function, knee flexion and extension ROM, 6 minute walking test (6MWT) and timed up and go test (TUG). Results: 11 studies (2.181 patients in total) were included in this systematic review. The long-term pain outcome showed no significant differences (Std. Mean Difference [SMD] = 0.00, 95 % confidence interval [CI] -0.16, 0.017) between the supervised (n = 397) and unsupervised (n = 255). Physical function showed no significant differences among the two groups (mean difference [MD] = 0.84, 95 % CI = -1.82, 3.50). Non-significant differences were also found for knee ROM flexion (mean difference [MD] = -0.46, 95 % CI = - 2.95, 2.04) and for knee extension (mean difference [MD] = 0.54, 95 % CI = -0.89, 1.97). At the 52-week follow-up, the unsupervised group showed significant better results in 6MWT (mean difference [MD] = -26.10, 95 % CI = -47.62, -4.59) and in Timed up and go test (mean difference [MD] = 1.33, 95 % CI = 0.50, 2.15).