血液分析仪
重复性
医学
核医学
内科学
统计分析
血液学
血小板
标准不确定度
血球计数
协议限制
全血细胞计数
作者
Ana Nikler,Marija Grdić Rajković,Vanja Radišić Biljak
摘要
ABSTRACT Introduction This study provides a comprehensive verification and comparative evaluation of four hematology analyzers (HA) with a five‐part differential blood count (5‐DIFF) (Siemens Advia 2120i, Sysmex XN‐1000, Mindray BC‐5310CRP, Beckman Coulter DxH 900). Methods The parameters of complete blood count (CBC) with 5‐DIFF were examined. The verification protocol assessed precision, accuracy, linearity, limits of blank (LoB), detection (LoD) and quantification (LoQ), and carryover. Advia 2120i served as the reference HA for accuracy assessment. Statistical analysis was performed using MedCalc and MS Excel. Acceptance criteria followed manufacturer specifications and biological variability analytical performance specifications (APS). Results All analyzers showed acceptable precision for most CBC parameters, while XN‐1000 and DxH 900 showed the best repeatability. BC‐5310 CRP showed the highest proportion of parameters exceeding acceptance criteria: low leukocytes in control (2.9%), and low platelets (6.1%) as well as low (16.4%) and high (4.1%) leukocytes in patient samples. Advia 2120i showed suboptimal repeatability at low hemoglobin in control (53 g/L, CV 1.1%) and low leukocytes (4.7%) for patient samples. Platelet repeatability varied most (CV 5.4%–9.7%). Platelet accuracy was significantly different on XN‐1000 ( p = 0.04) and BC‐5310CRP ( p = 0.05). Determined LoQ for leukocytes were: 0.1 (Advia 2120i) and 0.3 × 10 9 /L (XN‐1000, BC‐5310CRP, DxH 900). LoQ for platelets: 3 (Advia 2120i), 4 (XN‐1000), 5 (DxH 900), and 10 × 10 9 /L (BC‐5310CRP). Linearity was confirmed for all, wherein Mindray had slightly wider 95%‐CI. No carryover was observed. Conclusion While all analyzers demonstrated acceptable performance for the majority of evaluated specifications, Sysmex XN‐1000 and Beckman Coulter DxH 900 achieved the most favorable overall performance.
科研通智能强力驱动
Strongly Powered by AbleSci AI