Abstract
Aim
Although established serum inflammatory biomarkers, such as serum C-reactive protein (CRP) and serum white blood cell count (WBC), showed low accuracies in the literature, they are still commonly used in diagnosing periprosthetic joint infections (PJI). For a sufficient preoperative diagnosis novel more accurate serum parameters are needed. The aim of our study was to evaluate the performances of the established and novel routinely available serum parameters in diagnosing periprosthetic joint infections when using the proposed European Bone and Joint Infection Society (pEBJIS) criteria.
Method
In this retrospective study, 177 patients with an indicated revision surgery after a total joint replacement were included from 2015 to 2019. The easily accessible and routinely available serum parameters CRP, WBC, the percentage of neutrophils (%N), the neutrophils to lymphocytes ratio (NLR), fibrinogen and the platelet count to mean platelet volume ratio (PC/mPV) were evaluated preoperatively. The performances were examined via receiver operating characteristic (ROC) curve analysis (AUC). The curves were compared using the z-test. Seventy-five cases (42%) showed a PJI based on the pEBJIS-criteria.
Results
The sensitivities of serum CRP (cut-off: ≥10mg/L), WBC (≥10×10^9 cells/L), %N (≥69.3%), NLR(≥ 3.82), fibrinogen (≥ 457 mg/dL), and PC/mPV (≥ 29.4) were calculated with 68% (95% CI: 57–78), 36% (26 – 47), 66% (54 – 76), 63% (51 – 73), 69% (57 – 78), and 43% (32 – 54), respectively. Specificities were 87% (79 – 93), 89% (81 – 94), 67% (57 76), 73% (63 – 81), 89% (80 – 93), and 81% (72 – 88), respectively. Serum CRP and fibrinogen showed better performances than the other evaluated serum parameters (p<0.0001). The median serum CRP (17.6 mg/L) in patients with PJI caused by a low virulence microorganism was lower compared with infections caused by high virulence organisms (49.2 mg/L; p=0.044). Synovial fluid leucocyte count and histology showed better accuracies than serum CRP, serum WBC, %N, NLR, serum fibrinogen, and PC/mPV (p<0.0001).
Conclusions
Although serum CRP and fibrinogen showed the best performances among the evaluated serum inflammatory markers, their results should be interpreted with caution in clinical practice. Serum parameters may remain normal in chronic infections or may be elevated in patients with other inflammatory conditions. In addition, they also correlated poorly with synovial fluid leukocyte count and histology. Therefore, serum parameters are still insufficient to confirm or exclude a periprosthetic joint infection. Hence, they can only be recommended as suggestive criteria in diagnosing PJI.
