Validity of frozen sections for analysis of periprosthetic loosening membranes.

Abstract

Clinical findings and blood parameters often are inconclusive in patients with periprosthetic joint infections. Among the accepted criteria for diagnosis, histologic analysis of debrided tissue can detect infection in most cases but does not allow intraoperative decision making. We evaluated the validity of intraoperative frozen sections for detection of prosthetic infections. The results from frozen and permanent sections of periprosthetic membranes of 64 consecutive patients who underwent exchange procedures after hip arthroplasty were compared using the histopathologic consensus classification of Morawietz et al. Blood parameters (erythrocyte sedimentation rate, leukocyte count, C-reactive protein) and culture results of preoperatively aspirated joint fluid and intraoperative tissue samples were correlated with the histologic results. In 50 patients (78.1%), agreement was found between the frozen and permanent sections. Two patients (3.1%) revealed a discrepancy between the two histologic methods. In 12 patients (18.8%), a diagnosis was not possible based on the frozen sections because the tissue samples were not representative enough for definite classification. For the analyzable cases (n = 52), the sensitivity of frozen-section histologic analysis was 86.6%, specificity 100%, and accuracy 96.2%. Our data support a recommendation for use of intraoperative frozen sections for diagnosis of septic versus aseptic loosening in revision hip surgery. Level II, diagnostic study. See the Guidelines for Authors for a complete description of levels of evidence.