Revision of total knee replacement.

In the London and Toronto areas between 1969 and 1978, 597 patients had 661 cemented knee replacements. Eighty-nine (13.5%) prostheses failed; 73 of them were revised. Nineteen (2.9%) failed primarily because of infection. The revised replacements were examined and rated using a modification of the British rating system. The overall results of revision were excellent or good in 30% and fair or poor in 52%. Fusion was attempted in two patients after revision. In 16 patients with failed replacements that were not revised, fusion was attempted in 9; 4 of the 16 patients had amputation. The results of failure in knee replacement surgery are similar to those of hip replacement, but are helped considerably by lower patient expectation.