Background/Purpose

It has been well studied and accepted that the best method for evaluating joint disease is examination of synovial fluid. Synovial fluid analysis is critical to establish a definite diagnosis, whether the patient has a septic joint or a crystal arthropathy. Our study aims at studying the consistency of crystal identification between the Rheumatology department and the hospital laboratories as well as identifying the factors contributing to the misidentification of crystals.

Methods  

A retrospective study of synovial fluid analysis performed by the Rheumatology Department and the Rhode Island Hospital (RIH) laboratories was done over a year. Synovial fluid was gathered by arthrocentesis performed by the Rheumatology faculty for anyone with suspected crystal induced arthritis. Synovial fluid samples were analyzed by an attending physician with a compensated polarized microscope in the office and reviewed by another attending and/or fellow. Synovial fluid sent to the RIH lab was analyzed within the hour. A standard protocol was used for each sample which included cell count and differential. For this analysis the fluid was diluted with normal saline. Afterwards, part of the sample was dry centrifuged to examine for crystals. Both the wet prep of the sample as well as the dry centrifuged slides were examined by laboratory technicians for crystals. Within 24 hours, all dry centrifuged samples were reviewed by a pathologist. Results from the Rheumatology department were compared to the laboratory results on the same sample fluids.

Results  

A total of 64 synovial fluid samples were examined. 18 discrepancies were found between the Rheumatology Department and the hospital laboratories. 14 of the samples reviewed by the faculty were found to have crystals, while the laboratory reported these to be negative. A McNemar’s test was used to evaluate the data set. For each type of crystal, the p-value was not statistically significant (p-value MSU = 0.45, CPP = 0.07) but the p-value for both crystals was statistically significant at 0.02. Sensitivity for each was calculated, with specificity being 100%, and compared between the faculty and the laboratory. The sensitivity for the detection of any crystals by the faculty was found to be 0.92, while for the laboratory it was 0.66. The sensitivity of MSU crystal detection by the faculty was 0.89 and by the laboratory, 0.74. Similar results were seen for the detection of CPP crystals by the Rheumatologists with the sensitivity being 0.90, while the laboratory’s sensitivity was much less, only at 0.57.

Conclusion  

The results of our study are consistent with previous studies showing that there are still discrepancies in synovial fluid analysis. Many factors may be contributing to these variations including observer error, differences in time spent examining the fluid, technician training and crystal concentration. When analyzed separately, not all of our findings reached statistical significance, but when examined together, statistical significance was met. We believe that this study has profound clinical significance. Errors in crystal detection can have serious impacts on disease management and patient care.

---

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/quality-improvement-in-the-identification-of-crystals-from-synovial-fluid-hospital-laboratory-versus-rheumatology-department-evaluation/