**2. Method**

For inclusion in this review, studies had to evaluate the efficacy of a post-arthroplasty regimen for patients who had metacarpophalangeal or joint arthroplasty. Preferred study designs were metanalyses, systematic reviews, and randomised controlled trials, but all published literature except expert opinion was accepted. Patients may have received any type of implant and soft-tissue procedure, due to rheumatoid arthritis, osteoarthritis or trauma.

Electronic databases searched were the Cochrane Musculoskeletal Disease Group Register, The Cochrane Library of Systematic review, Google Scholar, and Scopus. Manual searches included of the Journal of Hand Therapy, Hand Therapy and the Journal of Arthroplasty. Search terms in all combinations included 'joint replacement, hand, wrist metacarpophalangeal, arthroplasty, rehabilitation, post-operative, occupational therapy, physical therapy'. The search included papers from 1990 onward, aiming to find research about currently used prostheses and not prostheses of older designs and materials.

Studies were appraised as described by the Cochrane Collaboration (7) for sources of methodological bias that could decrease the internal validity of a study. The types of methodological bias were in patient selection, equality of treatment, attrition of patients, and detection of all relevant outcomes. If the study could not be fully appraised from the publication, information was sought by writing to the authors.
