The results in nineteen patients (twenty-one knees) who had a failed
unicompartmental knee replacement followed by a revision total knee
arthroplasty were evaluated. There were twelve excellent, four good, one
fair, and two poor results. The interval between the unicompartmental
replacement and the revision total knee arthroplasty ranged from eight
months to eight years. At the time of the revision, a major osseous defect
was found in sixteen knees (76 per cent). The duration of follow-up after
the revision ranged from two to ten years. At the most recent follow-up
examination, radiographs revealed at least one radiolucent line in thirteen
knees (62 per cent). The technical difficulties associated with the
revision operation are evidence that unicondylar arthroplasty is not a
conservative procedure that allows a total knee arthroplasty to be done
easily later. The results also do not support the argument that a revision
performed after failure of a unicondylar arthroplasty is less technically
demanding than one performed after a failed primary total knee
arthroplasty.