The Reliability Of Diagnosis Of Infection During Revision Arthroplasties

Background: Follow up studies have shown that 0.5 to 4 % of the total joint arthroplasties will be complicated by infection. Distinction between aseptic loosening and infection is important for prediction of the final outcome after revision arhtroplasty but also for the choice of operative treatment. However, diagnosis of low grade chronic infection is extremely demanding. Materials and

Methods: 68 hip and knee revision arthroplasties were reviewed retrospectively in order to evaluate the reliability of pre- and perioperative analysis of infection during total joint revision arthroplasties. The sensitivity and specificity for clinical signs, blood white-cell count, C-reactive protein level, radiographic analysis, bone and leukocyte scans, joint aspirations, and gram staining were determined. Tissue sample were harvested and cultured in all cases. Positive cultures were regarded as a true infection.

Results: We were not able to characterize the infection by clinical signs. Also no single test was able to show the presence of infection in all cases. The best results were obtained from pre- and perioperative joint aspirations. Joint aspiration showed 1.0 specificity and 0.75 sensitivity.

Conclusion: It is clear from this study that no single test is able to show the presence of infection in every case. Classical clinical signs, laboratory tests, special imaging studies and joint aspirations have all yielded a notable rate of false negative results. Therefore, we recommend that, if arthroplasty patients have pain in prosthetic joint without clear radiological evidence of loosening, bone scans and preoperative joint aspirations should be undertaken. Also, if radiological evidence of loosening is accompanied with one or more of following criteria; C-reactive protein level elevated, radiologic evidence of infection, loosening within the first five years after implantation. In case of infection a delayed two-stage reconstruction should be managed.