|| Checking for direct PDF access through Ovid
In a retrospective review of the cases of twenty-nine patients who had a revision of a failed unicondylar unicompartmental knee replacement, the cause of failure of the original unicondylar arthroplasty was found to be loosening of one or both components in sixteen patients, progression of disease in the opposite compartment in nine, patellofemoral symptoms in two, ligamentous instability in one, and malposition of the tibial component in one. A technical error had been made in sixteen of the patients, and ten of the errors were associated with the ultimate failure of the original unicondylar procedure. Ninety-three per cent of the failed unicompartmental arthroplasties were revised using a posterior cruciate-sparing total knee replacement. Bone grafts, augmentation with screws and cement, a long-stem component, or a combination of these were used in half of the patients. The average follow-up after the revision arthroplasty was 4.6 years.