Abstract

Aim of this study was to evaluate the role of synovial interleukin-6 and synovial lactate for predicting native septic arthritis. We analyzed retrospectively synovial fluid parameters (interleukin-6, total-protein, glucose, lactate, synovial-fluid-white-blood-cell-count) of 62 patients with culture-verified native septic arthritis and compared them to 57 patients with acute aseptic arthritis. Receiver-Operating-Characteristic-curves were calculated to determine the ‘Area-under-the-curves' (AUC), the best thresholds and the corresponding likelihood-ratios. The best parameter for diagnosing septic arthritis was synovial lactate (AUC = 0.864, sensitivity = 74.5%, specificity = 87.2%), followed by synovial interleukin-6 (AUC = 0.803, sensitivity = 92.5%, specifi­city = 64.1%) and the synovial-fluid-white-blood-cell-count (AUC = 0.782, sensitivity = 71.2%, specificity = 84.9%). Synovial lactate levels above 10 mmol/l almost proofed septic arthritis (interval-Likelihood-Ratio = 20.4), synovial interleukin-6 levels lower than 7000 pg/ml almost ruled out infection (interval-Likelihood-Ratio = 0.12). If none of these thresholds are met, physicians should estimate disease probability by the simultaneous use of the interval-Likelihood-Ratios of synovial lactate, synovial interleukin-6 and synovial-fluid-white-blood-cell-count.