Synovial Fluid Aspirates Diluted with Saline or Blood Reduce the Sensitivity of Traditional and Contemporary Synovial Fluid Biomarkers

Background Recent criteria-based diagnostic tools to diagnose periprosthetic infection (PJI), such as the International Consensus Meeting (ICM) definition of PJI, are heavily reliant on synovial fluid laboratory results. Despite the importance of synovial fluid in PJI diagnosis, the effect of the quality of synovial fluid aspirate on testing results has not been studied. Our laboratory has established quality control parameters to identify synovial fluid aspirates that are highly diluted by saline or blood, which appear to degrade the diagnostic performance of synovial fluid laboratory tests. Questions/purposes (1) What proportion of synovial fluid aspirates analyzed at one laboratory are of poor quality (defined as having a red blood count > 1M cells/uL or an optical density at 280 nm < 0.324 or > 1.19)? (2) Does a poor-quality aspirate decrease the sensitivities of International Consensus Meeting-based scores and other synovial fluid biomarker tests in terms of their ability to anticipate a positive culture? Methods From January 2016 to July 2019, a total of 123,549 synovial fluid samples were submitted to one laboratory for the purpose of diagnostic testing. Of these, 14% (16,773 of 123,549) samples were excluded because they were from a site other than a hip, knee, or shoulder arthroplasty, and an additional 33% (35,660 of 106,776) were excluded due to insufficient requested tests, resulting in 58% (71,116 of 123,549) samples included in this study. Specimens diluted with extreme levels of saline or blood were identified (defined as having a red blood count >1 M cells/uL or an optical density at 280 nm < 0.324 or > 1.19) as poor-quality aspirates. The sensitivities of synovial fluid C-reactive protein, alpha defensin, neutrophil elastase, white blood cell count, polymorphonuclear cell percentage, and the 2018 ICM-based tool were assessed in good-quality versus poor-quality synovial fluid samples. To avoid bias from using these evaluated tests within the reference definition of PJI in this study, a positive culture resulting from the synovial fluid served as the reference diagnosis defining a control cohort of PJI-positive samples. Although the low false-positive rate of synovial fluid culture allows for the valid estimation of synovial fluid test sensitivity, the high false-negative rate of synovial fluid culture prevents the valid estimation of test specificity, which was not evaluated in this study. Results Of the samples analyzed, 8% (6025 of 71,116) were found to have poor quality, in that they were substantially compromised by saline and/or blood. The sensitivity of all tests to detect culture-positive synovial fluid was lower in poor-quality than in good-quality samples: 2018 International Consensus Meeting-based tool (83% [95% CI 80 to 86] versus 97% [95% CI 96 to 97]), synovial fluid C-reactive protein (65% [95% CI 61 to 69] versus 88% [95% CI 87 to 89]), alpha defensin (70% [95% CI 66 to 73] versus 93% [95% CI 93 to 94]), neutrophil elastase (80% [95% CI 77 to 83] versus 96% [95% CI 96 to 97]), synovial fluid white blood cell count (69% [95% CI 65 to 73] versus 93% [95% CI 93 to 94]), and the polymorphonuclear cell percentage (88% [95% CI 85 to 91] versus 95% [95% CI 94 to 95]), with all p < 0.001. Conclusions When synovial fluid is substantially diluted with saline or blood, the biomarkers and cells being measured are also diluted, decreasing the sensitivity of laboratory testing. We recommend that future diagnostic studies exclude these samples because an artificial reduction in test sensitivity will be observed.