Background/Purpose:

The gold standard for crystal-related arthritis diagnosis remains synovial fluid analysis by a microscope fitted with compensated polarized filters as it has been shown an immediate, valid, easy to learn, and reliable procedure (for the last, only in two studies not designed for this purpose). The objective was to analyze the agreement between multiple observers in crystal identification using a compensated polarized optic microscope. Finding good reliability results between multiple observers would reinforce the value of the technique.

Methods:

Cross-sectional, observational study performed at a single center with consecutive synovial fluid sampling. Samples were immediately analyzed when possible or kept refrigerated at 4ºC. Five different observers analyzed samples under a compensated polarized optical microscope to detect and identify crystals, independently and blinded to clinical data. All observations were performed at 400x fields. It was recorded the presence and type of crystal (No crystals; monosodium urate -MSU-; and calcium pyrophosphate dihydrate -CPPD-). Inter-rater agreement by Cohen’s kappa (k) statistic was measure. Also, sub-analyses were performed on a) time of sample visualization (£24 hours or >24 hours of sampling) and b) expertise on crystal identification (£10 years or > 10 years).

Results:

Synovial fluid samples were obtained mostly from knees (67.3%), ankles (13.2%) and wrist (6.8%) of patients seen at rheumatology clinic or during admissions. Main rheumatic diseases were gout (33.8%), rheumatoid arthritis (18.4%), CPPD related arthritis (15%) and arthritis under study (8.3%). A total of 31 samples were analyzed by all five observers (155 observations). Overall k was 0.76 (CI95% 0.63-0.89) thus indicating good agreement. Agreement for detection was k 0.75 (CI95% 0.61-0,90) and k for MSU was 0.91 (CI95% 0.76-1.05) while for CPPD was 0.66 (CI95% 0.47-0.84). Regarding secondary endpoints, no differences were noted neither between observations made before or after 24 hours (p 0.68 for k comparison), nor in expertise in crystal analysis (p 0.63 for k comparison).

Conclusion:

Even requiring agreement among multiple observers, compensated polarized microscopy remains consistent in the detection and identification of crystals in synovial fluid, confirming its high utility in clinical practice.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/compensated-polarized-microscopy-for-crystal-identification-shows-high-reliability-among-multiple-observers/