Blood D-dimer levels in the diagnosis of peri-prosthetic joint infections

Abstract

Introduction Serum D-Dimer levels has been used for the diagnosis of thromboembolic disorders and have been seen to be elevated in infections. This study aims to assess the use of serum D-Dimer levels as an adjunct in diagnosis of Peri-prosthetic Joint Infections (PJI). Methodology This is a retrospective analysis of cases with revision total joint arthroplasties and suspected PJI during the year of 2016 in a single centre. All patients with pre-operative D-Dimer levels and work up for infection were included in this study. There was a total of 24 patients with 26 revision total joint arthroplasties performed. Diagnosis of PJI was based on the Musculoskeletal Infection Society criteria for PJI. 14 cases were revised due to non-infected causes while 12 cases were due to PJI. D-Dimer levels were analysed using Receiving Operator Characteristic (ROC) and area under curve (AUC) statistics to determine diagnostic accuracy, sensitivity and specificity. Head to head comparisons with currently established blood investigations, namely Erythrocyte sedimentation rate (ESR), C-Reactive protein (CRP) and leucocyte count, were performed. Results Age adjusted D-Dimer levels were significantly elevated in PJI cases. Patients with PJI had an average D-Dimer level of 3.65mg/L compared to non-PJI patients who had a level of 1.60mg/L (p = 0.01). Diagnostic cut-off value for D-Dimer was determined to be >1.56mg/L with a sensitivity and specificity of 83.3% and 74.1% respectively. D-Dimer demonstrated good diagnostic accuracy for PJI with an AUC of 0.8 (p<0.01). ROC analysis of leucocyte count, ESR and CRP found AUC of 0.66, 0.9 and 0.97 respectively. CRP had the highest sensitivity and specificity with 91.7% and 100% for a diagnostic cut-off of >3.07mg/dL ESR had a sensitivity of 83.3% and specificity of 92.9% with a cut-off value of >79mm/Hr. Leucocyte count test was not found to be useful in the diagnosis of PJI with no statistically significant difference between PJI and non-PJI groups. Pair-wise comparisons with ESR, CRP and leucocyte count against D-Dimer found that D-Dimer was superior to leucocyte count for diagnosis. When compared to CRP and ESR, D-Dimer fared less favourably, but overall differences in AUC were not statistically significant (p=0.05, p=0.19). Conclusion This study has proven D-Dimer can be used in the diagnosis of PJI. Although it does not have the diagnostic accuracy of CRP and ESR, D-Dimer has a strong negative predictive value. Combined use of all three blood investigations can improve diagnostic accuracy, sensitivity and specificity in the setting of PJI.