gms | German Medical Science

Introduction: Diagnosis of infection after total knee replacement (TKR) may be challenging. There is no single criterion with enough diagnostic power. We evaluated the diagnosis power of two new possible indicators: level of C-Reactive Protein (CRP) and level of leucocyte-esterase activity (LEA) in the articular fluid.

Methods: 32 patients were included and divided into three groups: 10 patients (group A) with a degenerative osteoarthritis; 11 patients (group B) with a knee effusion after TKR without septic complication; 11 patients (group C) with a proved infection after TKR. All patients received aspiration of the joint fluid prior to any treatment. The level of CRP and LEA were determined in the articular fluid by nephelometry and colorimetric technique, respectively. The levels were compared in the three groups with a Mann-Whitney test at a 0.05 level of significance. The ROC curves were plotted to define the optimal cut-off values.

Results: The mean CRP level was significantly increased in the group C (24.4 mg/l) in comparison to group A (1.9 mg/l) and B (3.7 mg/l) (p<0.001). The mean LEA level was significantly increased in the group C (1.9) in comparison to group A (0.1) and B (0.3) (p<0.001). A CRP level inferior to 2.8 mg/l allows excluding the diagnosis of infection, with a 100% negative predictive value. A CRP cut-off level of 5.4 mg/l had a sensibility of 91% and a specificity of 91% for the diagnosis of infection. The diagnosis power of any cut-off level of LEA was lower.

Discussion: These results are in line with those of other teams. CRP level is easy and inexpensive to perform with the same technique than serum level. The high sensibility and the high specificity with a cut-off level of 5.4 mg/l may be the most relevant biological factor to set the diagnosis of infected TKR. LEA level is also easy and inexpensive to perform, but the result might be biased by blood contamination of the fluid, and the reading might be observer-dependent and inaccurate. This bias and the lower diagnostic power preclude considering it as a relevant diagnostic tool.

Conclusion: A synovial fluid CRP cut-off level of 5.4 mg/l had a high sensibility and a high specificity for the diagnosis of infection. It might be interesting to add this criterion to the current diagnostic algorithms.