Diagnosis of periprosthetic infection following total hip arthroplasty - Evaluation of the diagnostic values of pre- and intraoperative parameters and the associated strategy to preoperatively select patients with a high probability of joint infection

Background. The correct diagnosis of a prosthetic joint infection (PJI) is crucial for adequate surgical treatment. The detection may be a challenge since presentation and preoperative tests are not always obvious and precise. This prospective study was performed to evaluate a variety of pre- and intraoperative investigations. Furthermore a detailed evaluation of concordance of each preoperative diagnosis was performed, together with a final diagnosis to assess the accuracy of the pre-operative assumption of PJI. Methods. Between 01/2005 and 02/2007, a prospective analysis was performed in 50 patients, who had a two stage revision because of assumed PJI. Based on clinical presentation, radiography, haematological screening, or early failure, infection was assumed and a joint aspiration was performed. Depending upon these findings, a two stage revision was performed, with intra-operative samples for culture and histological evaluation obtained. Final diagnosis of infection was based upon the interpretation of the clinical presentation and the pre- and intraoperative findings. Results. In 37 patients a positive diagnosis of PJI could be made definitely. The histopathology yielded the highest accuracy (0.94) in identification of PJI and identified 35 of 37 infections (sensitivity 0.94, specificity 0.94, positive-/negative predictive value 0.97/0.86). Intra-operative cultures revealed sensitivities, specificities, positive-/negative predictive values and accuracy of 0.78, 0.92, 0.96, 0.63 and 0.82. These values for blood screening tests were 0.95, 0.62, 0.88, 0.80, and 0.86 respectively for the level of C-reactive protein, and 0.14, 0.92, 0.83, 0.29 and, 0.34 respectively for the white blood-cell count. The results of aspiration were 0.57, 0.5, 0.78, 0.29, and 0.54. Conclusion. The detection of PJI is still a challenge in clinical practice. The histopathological evaluation emerges as a highly practical diagnostic tool in detection of PJI. Furthermore, we found a discrepancy between the pre-operative suspicion of PJI and the final post-operative diagnosis, resulting in a slight uncertainty in whether loosening is due to bacterial infection or not. The variation in accuracy of the single tests may influence the detection of PJI. Level of Evidence: Diagnostic Level I. © 2008 Müller et al; licensee BioMed Central Ltd.