Aspiration of the knee joint before revision arthroplasty.

Sixty-four operative procedures were performed on 55 patients with knee arthroplasties for pain, loosening, instability, or suspicion of infection. Forty-three knees had a preoperative aspiration. In 19 knees, the aspiration showed growth on solid media, and in 18 of these knees the diagnosis of infection was confirmed by the intraoperative cultures. In 1 knee with an infected total knee replacement, the patient was receiving an intravenous antibiotic at the time of arthroscopic irrigation and debridement and the cultures showed no growth. In 23 of 24 knees with a negative preoperative aspiration, the intraoperative cultures showed no growth on solid media. In 1 knee with a preoperative aspiration that had negative results, a single intraoperative culture grew Staphylococcus epidermidis. However, the presenting symptoms, examination, preoperative radiographs, and intraoperative evaluation were consistent with aseptic loosening of a cemented total knee arthroplasty. Thus, the preoperative aspiration of the prosthetic knee joint had a sensitivity of 100%, specificity of 100%, and accuracy of 100%. The Westergren erythrocyte sedimentation rate, peripheral leukocyte count, and presenting symptoms correlated poorly with infection. Radiographs were also not helpful in the diagnosis of infection, with loosening of components, periostitis, focal osteolysis, and radiolucent lines frequently seen in infected and noninfected knees. Preoperative aspiration of the knee is the most helpful study for the diagnosis or exclusion of infection in a prosthetic knee joint.