THU0414 ORDINARY LIGHT MICROSCOPY IS ABLE TO IDENTIFY MOST CRYSTAL-CONTAINING SYNOVIAL FLUIDS

Background: Optical microscopy remains the gold standard for the diagnosis of crystal arthropathies. The complete protocol consists of three phases. In the first stage, microscopy with simple light provides information on the morphology of the crystal. The second stage, polarized light, allows detecting the intensity of the birefringence. Finally, with the first-order red compensator, the type of elongation is detected, positive for calcium pyrophosphate (CPP) crystals and negative for monosodium urate (MSU) crystals. Finally, with the obtained data, the presence and type of crystals is concluded. Objectives: Analyze the validity and agreement of each stage of microscopy regarding the conclusion, emphasizing ordinary light microscopy. Methods: Fifty consecutive samples of synovial fluid obtained in routine clinical practice were independently analyzed under the compensated polarized microscope by 5 observers blinded to clinical data (250 observations in total). Each observer recorded the presence and type of crystals at each stage and reached a conclusion after gathering all the information. To estimate the diagnostic yield of each microscope stage, sensitivity, specificity and positive and negative predictive values, as well as the accuracy (number of correct observations/number of total observations), were calculated; also, the total weighted kappa was used to assess the degree of agreement with the complete protocol. Results: Main results of the study are shown in Table 1. Regarding diagnostic yield, ordinary light microscopy showed excellent sensitivity, specificity and predictive values, similar to the results noted with simple and compensated polarized microscopy. Table 1. In parentheses, 95% confidence intervals. Accuracy Sensitivity Specificity Positive predictive value Negative predictive value Kappa Ordinary light 96.8% (93.8-98.4) 97.2% (93.1-98.9) 96.2% (90.7-98.5) 97.2% (93.1-98.9) 96.2% (90.7-98.5) 0.954 (0.919-0.989) Simple polarized light 92.0% (88.0-94.8) 84.1% (76.8-89.5) 100% (97.0-100) 100% (96.5-100) 86.1% (79.5-90.8) 0.874 (0.821-0.927) Compensated polarized light 97.6 % (94.9-98.9) 95.5% (89.8-98.0) 99.3% (96.1-99.9) 99.1% (94.8-99.8) 96.5% (92.1-98.5) 0.962 (0.933-0.992) Diagnoses established by ordinary light microscopy matched conclusions (accuracy) in 242/250 (96.8%) observations. Discrepant cases were crystals missed under ordinary light in 4 cases (3 MSU, 1 CPP), and 4 samples with CPP crystals initially seen but later concluded their absence. Interestingly, lowest accuracy was seen with simple polarization; CPP crystals were not detected in 20 out of 93 observations with CPP (21.5%). The accuracy of compensated polarized light was similar to ordinary light. On 5 occasions no crystals were seen but finally they were present (1 MSU, 4 CPP); on the contrary, CPP was registered in one observation but the conclusion indicated no crystals. Regarding agreement with the complete protocol, the kappa with simple light is 0.954, similar to compensated polarized light (0.962), while simple polarized light showed the lowest agreement (0.874). Conclusion: Ordinary light microscopy is enough to correctly reach the majority of diagnoses, with a very high degree of agreement with the complete protocol. Results were comparable to using a compensated polarized microscopy. Thus, if a microscope with polarizer and first-order compensator was not available, using ordinary light would be enough on most occasions. Polarized light microscopy better identifies MSU crystals, but over 20% of CPP crystals were missed at this stage, reinforcing the value of the ordinary light microscopy. Acknowledgments: Thanks to Loreto Carmona for the help with the statistical aspects. Disclosure of Interests: : None declared