Automated Cerebrospinal Fluid Cell Counts Using the New Body Fluid Mode of Sysmex UF-1000i

Abstract

Background: We evaluated the new body fluid module on Sysmex UF1000-i (UF1000i-BF) for analysis of white blood cell (WBC) and red blood cell (RBC) in cerebrospinal fluid (CSF). Methods: WBC and RBC counting were compared between UF1000i-BF and Fuchs–Rosenthal counting chamber in 67 CSF samples. This study also included the evaluation of between-day precision, limit of blank (LoB), limit of detection (LoD), functional sensitivity (limit of quantitation, LoQ), carryover and linearity. Diagnostic agreement for differentiation between normal and increased WBC counts (≥5.0 × 106/L) was also assessed. Results: The agreement between UF1000i-BF and manual WBC counts was otpiaml in all CSF samples (r = 0.99; y = 1.05x + 0.09). A modest overestimation was noticed in samples with WBC < 30 × 106/L (r = 0.95; y = 1.21x − 0.15). A good agreement was observed for RBC counts (r = 0.98; y = 1.15x + 0.55), particularly in samples with RBC ≥ 18 × 106/L (r = 0.98; y = 1.01x + 8.90). Between-day precision was good, with coefficient of variations (CVs) lower than 7.2% for both WBC and RBC. The LoBs were 0.1 × 106 WBC/L and 1.2 × 106 RBC/L, the LoDs were 0.7 × 106 WBC/L and 5.5 × 106 RBC/L, the LoQs were 2.4 × 106 WBC/L and 18.0 × 106 RBC/L, respectively. Linearity was excellent (r = 1.00 for both WBC and RBC). Carryover was negligible. Excellent diagnostic agreement was obtained at 4.5 × 106 WBC/L cut-off (sensitivity, 100%; specificity, 97.4%). Conclusion: The UF1000i-BF provides rapid and accurate WBC and RBC counts in clinically relevant values of CSF cells. The use of UF1000i-BF may hence allow to replace routine optical counting, except for samples displaying abnormal WBC counts or abnormal scattergram distribution, for which differential cell counts may still be required.