Abstract
BACKGROUND: Blood biomarkers are first-line tools for identifying periprosthetic joint infection (PJI). C-reactive protein (CRP) is currently recognized as the standard biomarker for PJI diagnosis. Other recently reported novel biomarkers, including plasma fibrinogen, platelet count, monocyte/lymphocyte ratio (MLR), neutrophil/lymphocyte ratio (NLR), and platelet count/lymphocyte ratio (PLR), have also shown promise in diagnosing PJI. This study aimed to evaluate whether these biomarkers were superior to CRP for identifying PJI. METHODS: Patients who underwent revision hip or knee arthroplasty at our hospital from January 2008 to September 2020 were included consecutively and divided into infected and non-infected groups according to the 2013 International Consensus Meeting Criteria. Blood samples were collected preoperatively, and erythrocyte sedimentation rate (ESR), CRP, interleukin-6, fibrinogen, platelet count, MLR, NLR, and PLR were analyzed. The diagnostic values of the tested biomarkers and their combinations were compared with CRP based on the area under the receiver operating characteristic curve (AUC) using the z-test. Classification trees were constructed to explore more accurate combinations of the tested markers for identifying PJI. RESULTS: A total of 543 patients were included, of whom 245 had PJI. Among the tested biomarkers, CRP with a cutoff of 7.39 mg/L showed the highest AUC, which gave a sensitivity of 79.1% and specificity of 86.0%. The AUCs of pairwise combinations of tested markers including CRP also were inferior to CRP itself, as were combinations derived from classification trees. CONCLUSIONS: Preoperative serum CRP with a low cutoff may be the best reliable blood biomarker for identifying PJI, and those traditional or novel available blood biomarkers could not further improve the diagnostic ability on the basis of CRP.