C-Reactive Protein and Erythrocyte Sedimentation Rate Changes After Arthroscopic Anterior Cruciate Ligament Reconstruction: Guideline to Diagnose and Monitor Postoperative Infection

Purpose: The purposes of our study were to determine normative C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) values from a retrospective review of patients with and without infection after anterior cruciate ligament (ACL) reconstruction and to determine CRP and ESR threshold levels that can serve as diagnostic indicators of infection. We also tried to draw a curve of CRP and ESR value changes after treatment of ACL infection to evaluate the response to treatment of the infection. Methods: A retrospective chart review was performed of arthroscopic ACL reconstruction patients from 2007 to 2008 (noninfection group) and all patients with postoperative intra-articular infection from 1997 to 2010 (infection group). We collected the CRP and ESR values on the third and fifth postoperative days in the noninfection group and before infection treatment and on the first, third, fifth, seventh, 10th, 14th, 21st, 28th, and 35th days after infection treatment in the infection group. Sensitivity, specificity, and Youden's index were calculated for different threshold values of CRP and ESR as predictors of infection. Receiver operator curves were obtained for CRP and ESR on the fifth postoperative day. Results: Of 122 patients, 83 had normal joints and 39 had septic joints. The mean CRP and ESR values in patients with septic joints were 101.9 mg/L and 57.1 mm/h, respectively, which were significantly higher than those in the noninfection group (P < .01). A CRP value of 41 mg/L and ESR value of 32 mm/h were the optimal thresholds to predict an infection, which had the highest Youden's index of all calculated values and had sensitivity values of 94.1% and 91.2%, respectively, and specificity values of 97.6% and 80.5%, respectively. The peak CRP level after infection treatment occurred earlier than the peak ESR level (first day v third day) and returned to normal more quickly (21st day v 28th day). Conclusions: Both CRP and ESR were helpful in determining the presence of a normal or septic joint. The threshold values of 41 mg/L for CRP and 32 mm/h for ESR had the most optimal sensitivity and specificity. The peak CRP level occurred earlier than the peak ESR level after treatment of postoperative infection and returned to normal more quickly. In this study CRP was more useful than ESR to evaluate the response of infection to treatment.