RESUMO
To determine whether leukocyte esterase reagent strip test (LERST) analysis could help distinguish inflammatory arthritis from mechanical joint effusion. We analyzed synovial fluid (SF) from consecutive patients with a non-traumatic joint effusion during a 6-month period. Inflammatory SF was defined by white blood cell (WBC) count ≥ 2000/mm3. The LERST was performed with both semi-quantitative visual analysis (VA) and automated colorimetric reader (ACR) analysis. Leukocytes ≥ 1+ was considered a positive LERST result and WBC count was the reference. We obtained 100 SF samples (87 knees, 7 ankles, 5 hips, and 1 elbow) from 100 patients (mean ± SD age 61 ± 17 years, 59% men). The laboratory analyzed 88 SF samples (37 mechanical and 51 inflammatory). The remaining 12 SF samples were 10 hemarthrosis not allowing LERST analysis and 2 samples with coagulum not allowing WBC count. As compared with the laboratory analysis, the LERST had sensitivity and specificity 55% and 89% with VA and 47% and 92% with ACR analysis. The positive and negative predictive values were 87.5% and 59% with VA and 89% and 55% with ACR analysis. We found almost perfect agreement between VA and ACR results (kappa 0.70 [95% CI 0.50-0.90]). The WBC count increased with number of + observed after VA. Our results confirm that the LERST is able to detect inflammation in SF of native joints, thereby representing a useful and cheap tool in primary care. Its low sensitivity limits its use for ruling out inflammatory disorders. Key Points ⢠Reagent strip tests can detect inflammation in synovial fluid. ⢠In primary care practice, this method is cheap and easy to do.