Neutrophil–Lymphocyte Ratio and Lymphocyte–Monocyte Ratio correlate with Chronic Prosthetic Joint Infection but are not useful markers for diagnosis

Purpose To investigate reported correlations between Neutrophil-to-Lymphocyte (NLR) and Lymphocyte-to-Monocyte (LMR) ratios and their value in diagnosis of chronic prosthetic joint infection (PJI) in a large cohort of patients from a single specialist hospital. Methods Diagnostic aspirations of 362 patients under investigation for PJI were identified. Of the included patients 185 patients received a final diagnosis of PJI and 177 were classed as aseptic. Established criteria (ICM 2018) were employed to define PJI. Included in the analysis are differential white cell counts, C-Reactive Protein (CRP), Synovial Leukocyte Count, Synovial Alpha-defensin ELISA and Synovial Leukocyte esterase activity. Receiver-operator characteristic (ROC) curves were calculated for each of the available diagnostic tests together with the corresponding area under the curve values (AUC). Youden’s index was utilized to identify the optimal diagnostic threshold point for the NLR and LMR. Other diagnostic tests were evaluated as per the threshold values previously defined in the literature and specified in the ICM criteria. Results Using Youden’s Index to identify the optimal NLR cut-off within our cohort we established a value of 2.93. This yielded a sensitivity of 0.60 and specificity of 0.64. The area under the curve (AUC) of a receiving operator characteristics (ROC) curve was 0.625. Regarding the LMR the results demonstrate similar findings; a positive correlation with a diagnosis of infection but poor sensitivity and specificity. The AUC for LMR was 0.633 and was not superior to NLR ( P = 0.753). Conclusions There is a significant correlation between higher Neutrophil–Lymphocyte and Lymphocyte–Monocyte ratios, and a diagnosis of PJI. The sensitivity and specificity of this calculation is poor and the does not add value to the diagnostic algorithm for PJI. Level of evidence Level III Retrospective Cohort analysis.