Automated counting of nucleated cells in equine synovial fluid without and with hyaluronidase pretreatment

Background Microscopy is usually used to obtain manual total and differential cell counts in equine synovial fluid. A faster, more precise method is desirable. Objectives The objectives were to compare an automated impedance method with a manual method for obtaining total and differential cell counts in equine synovial fluid and to evaluate the effect of pretreatment with hyaluronidase on automated results. Methods Synovial fluid samples (n=48) were collected into EDTA and analyzed within 48 hours. Automated total and differential cell counts were evaluated using a Medonic CA620-VET hematology analyzer before and after pretreatment for 5-30 minutes with hyaluronidase (final concentration 0.01 mg/mL). A hemacytometer count and microscopic evaluation of a direct smear were used as the reference method. Intra-assay coefficients of variation (CV) were determined. Results Thirty-one of 46 untreated samples and 0/46 hyaluronidase-treated samples were error-flagged by the analyzer. Correlation between automated (ANCC) and manual (MNCC) nucleated cell counts in untreated samples (n=15; R2=0.93) and pretreated samples (n=46; R2=0.94) was high, and pseudomedian difference was low. Intra-assay CVs for samples with medium and high cellularity were significantly lower for ANCC (1.5-2.7%) compared with MNCC (6.1-15.7%) (P <.01). Valid automated differential cell counts were not obtained. Conclusions Automated total cell counts obtained on the Medonic analyzer correlate well with manual counts in equine synovial fluid; however, pretreatment with hyaluronidase is required to minimize error flags. Automated differential counts are not accurate for synovial fluid.