Abstract
Aims: This study evaluates the value of a new graphic representation of cell count data of synovial fluid in the diagnosis of acute periprosthetic joint infection (PJI). Methods: A total of 75 patients with revisions of 48 primary total knee and 27 hip arthroplasties within the first six weeks after surgery were analyzed with cultivation of the synovial fluid and determination of its cell count as well as microbiological and histological analyses of the periprosthetic tissue obtained during the revision surgery using the ICM classification. The synovial fluid was additionally analyzed for graphic representation of the measured cells using LMNE-matrices. Results: A total of 38 patients (50.7%) had an infection. The following types of LMNE matrices could be differentiated: the indeterminate type (IV) in 14.7%, the infection type (II) in 5.3%, the hematoma type (V) in 33.3%, and the mixed type (VI; infection and hematoma) in 46.7%. Differentiation of LMNE types into infection (types II and VI) and non-infection (types IV and V) resulted in a sensitivity of 100%, a specificity of 97.3%, and a positive likelihood ratio of 37.0. The cell count measurement showed a sensitivity of 78.9%, a specificity of 89.2%, and a positive likelihood ratio of 7.3 at a cut-off of 10,000 cells. The percentage of polymorphonuclear leukocytes showed a sensitivity of 34.2%, a specificity of 100%, and a positive likelihood ratio of >200 at a cut-off of 90%. Conclusion: The graphic representation of the cell count analysis of synovial aspirates is a new and helpful method for differentiating between genuine early periprosthetic infections and postoperative hemarthrosis.