Agreement Among Multiple Observers on Crystal Identification by Synovial Fluid Microscopy

Area of Science:

• Rheumatology
• Crystallography
• Medical Diagnostics

Background:

• Polarized microscopy is the gold standard for diagnosing crystal arthritis.
• Clinical implementation faces uncertainties regarding observer agreement and reliability.

Purpose of the Study:

• To analyze interobserver agreement in crystal identification using compensated polarized microscopy.
• To assess potential outcome modifiers affecting crystal detection.

Main Methods:

• A cross-sectional, observational study involving synovial fluid samples analyzed by five independent observers.
• Samples were examined using compensated polarized optical microscopy under ordinary, simple polarized, and compensated polarized light.
• Interrater agreement was quantified using Cohen's kappa, with subanalyses for visualization delay and observer expertise.

Main Results:

• Good overall interobserver agreement (κ = 0.74) for crystal identification in synovial fluid.
• High agreement for monosodium urate (MSU) (κ = 0.88) and calcium pyrophosphate dihydrate (CPP) (κ = 0.69) crystals.
• Most crystal identifications (96.8%) were achieved using ordinary light, with no significant impact from visualization delay or observer expertise.

Conclusions:

• Compensated polarized microscopy demonstrates consistent utility for crystal detection and identification in synovial fluid among multiple observers.
• The findings support the high reliability and clinical applicability of this diagnostic method for crystal arthritis.