Finding the Optimal Screening Test for Periprosthetic Joint Infection: A Prospective Study

Introduction No single test has demonstrated absolute accuracy in the diagnosis of periprosthetic joint infection (PJI). Serological markers are often used as screening tools to avoid unnecessary joint aspiration in cases with a low probability of infection. This study aimed to determine the utility of standard-of-care serological tests as a screening tool for PJI in patients undergoing revision arthroplasty. Methods This prospective study enrolled 502 patients undergoing revision hip or knee arthroplasty between May 2017 and August 2021. A PJI was defined using a modified definition of the 2018 International Consensus Meeting (ICM) criteria. Plasma D-dimer, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and fibrinogen were measured preoperatively. There were 82 patients undergoing reimplantation who were excluded. Additionally, 8 patients who had an inconclusive ICM score were also excluded. Of the 412 included patients, 317 (76.9%) underwent revision for aseptic failure, and 95 (23.1%) had PJI. Receiver operating characteristic curves were used to assess the diagnostic utility of each serological test. A pairwise comparison with Bonferroni correction was performed in order to determine whether the differences in areas under the curve (AUCs) between the tests were significant. Additional analyses were performed to find the threshold for each test that offered 100% sensitivity, allowing it to be the optimal screening test. Results All four serological markers, D-dimer (AUC 0.860, sensitivity 81.3%, specificity 81.7%), CRP (AUC 0.862, sensitivity 90.4%, specificity 70.0%), ESR (AUC 0.833, sensitivity 73.9%, specificity 85.2%), and fibrinogen (AUC 0.798, sensitivity 74.7%, specificity 75.4%), demonstrated comparable accuracy for the diagnosis of PJI (all P > 0.05). When maximizing sensitivity to 100%, D-dimer demonstrated the highest specificity (AUC 0.860, specificity 40.2%), outperforming ESR (AUC 0.833, specificity 3.3%), fibrinogen (AUC 0.798, specificity 2.3%), and CRP (AUC 0.862, specificity 0%). A plasma D-dimer level of ≥ 244 ng/mL was identified as the optimal cutoff for use as a screening test. Conclusion Although plasma D-dimer demonstrated similar diagnostic accuracy as CRP, ESR, and fibrinogen, it outperformed all three aforementioned serological markers when used as a screening test for PJI.