Treatment of infection after total knee arthroplasty by debridement with retention of the components

Thirty-one total knee arthroplasties were followed by infection in twenty-seven patients who were subsequently treated with debridement, retention of the components, and intravenous administration of antibiotics. The results were reviewed retrospectively in an effort to evaluate the function of the prostheses that had been salvaged successfully and to identify the causes of failure of those around which an infection had recurred. At the most recent follow-up (average duration, 8.8 years), infection had recurred around twenty-four (77 per cent) of the thirty-one arthroplasties. Seven knees (23 per cent) remained free of infection. Function remained satisfactory, although revision was subsequently needed in two knees for reasons other than infection. One cause of failure was the duration of the infection before debridement. This averaged twenty-one days for the seven knees in which the prosthetic arthroplasty had been salvaged and thirty-six days for the twenty-four knees in which treatment had failed. Another cause of failure was the type of organism: Staphylococcus aureus had caused the infection in fourteen (58 per cent) of the twenty-four knees in which the treatment failed but in only two of the seven knees in which the prosthetic arthroplasty was salvaged. In addition, eight infections in the first group were resistant to penicillin, whereas both infections in the second group were sensitive to penicillin. The four infections with gram-negative organisms in the series were all in knees in which the treatment failed. All six hinged prostheses that were used failed. The average age of the patients who had a successful result was fifty years and for those who had a failure it was sixty-two years (p ≤ 0.01). The low rate of success (23 per cent) was disappointing. It is apparent that certain criteria musr be met before this approach is used. However, it a patient is treated successfully, function can be expected to remain satisfactory over the long term. The duration of the infection, the virulence of the infecting organism, the age of the patient, and the type of prosthesis all influenced the chance for a good long-term result.