We carried out a survival analysis of elbow synovectomy (ES) and excision of the radial head (RHE) performed on 171 rheumatoid elbows. The failure criteria were revision surgery (performed or desired) and/or the presence of significant or severe pain. The cumulative survival was 81% at one year which thereafter decreased by an average of 2.6% per year. The strongest predictor for success was a low preoperative range of supination-pronation when corresponding survival curves were compared. A low range of flexion-extension also predicted failure. Combining both factors gave better prediction (failure: 6.3% v 67%), but a long duration of elbow symptoms before surgery predicted failure (72%, p = 0.04). At review, there was a mean gain of 50° in supination-pronation and 11° in flexion-extension; both correlated with success. Failure correlated with recurrence of synovitis, elbow instability, ulnar neuropathy, poor general mobility and poor upper-limb function. The last was independently affected by the severity of RA in the ipsilateral shoulder. Our findings show that although the short-term result of ES and RHE in rheumatoid arthritis is good, the long-term outcome is poor except in a subgroup with more than 50% limitation of forearm rotation

Six highly unstable elbows with severe bone loss due to rheumatoid arthritis were replaced by a non-constrained, unlinked prosthesis. Bone defects were filled with autogenous bone grafts. The mean follow-up was 4.5 years (2 to 8). The clinical results were excellent in four elbows and good in two, with good varus-valgus stability in all. Radiological follow-up showed no appreciable signs of loosening, and the bone grafts had retained most of their original size, with minimal resorption. There were no major complications such as dislocation, skin necrosis, infection or ulnar neuropathy. The study has shown that the so-called mutilans elbow can be successfully replaced using a properly selected type of non-constrained, unlinked prosthesis with bone grafting of the major defects