Verheul F E, Spitters J M, Bergmans C H
St. Maartens Gasthuis, Venlo, The Netherlands.
Eur J Clin Chem Clin Biochem. 1993 Mar;31(3):179-86.
The Coulter STKS is a fully automated haematology analyser which quantitatively measures and computes haematological quantities including a full "five-part" differential. The aim of this study was to evaluate whether the Coulter STKS can be used for screening purposes in a non-academic clinic. Results from the haemocytometry part of the apparatus correlated very well with those from the Coulter S Plus IV. Reasonable to good correlations were obtained with the standard manual 100-cell differential for neutrophils, lymphocytes, monocytes and eosinophils. Significantly higher percentages of monocytes were found by microscopic observation than by the STKS. Samples from 1080 patients were analysed on the STKS, and manual differentials were also performed. After revision of discrepancies and with the exclusion of neonatal samples, a good numerical agreement for the "five-part" differential was obtained. The STKS gave 1.8% false negatives, and both diagnostic sensitivity (96.3%) and specificity (83.3%) were high. A false negative rate of 1.9% was found for the indication left-shift, with a predictive value of a negative test of 97.4%. The predictive value of a positive test, however, was only 40.3%. An authorization procedure was implemented to reduce false negative samples. The routine laboratory work load for our differentials can be reduced by up to 70%. The Coulter STKS gave a high number of suspect flags for neonatal samples (75%) and is therefore not suitable for screening neonates.
库尔特STKS是一款全自动血液分析仪，可定量测量并计算血液学指标，包括完整的“五分类”白细胞计数。本研究旨在评估库尔特STKS能否用于非学术性诊所的筛查。该仪器血细胞计数部分的结果与库尔特S Plus IV的结果相关性非常好。中性粒细胞、淋巴细胞、单核细胞和嗜酸性粒细胞与标准手动100细胞分类计数的相关性合理至良好。显微镜观察发现的单核细胞百分比显著高于STKS检测结果。对1080例患者的样本进行了STKS分析，并同时进行了手工分类计数。在修正差异并排除新生儿样本后，“五分类”白细胞计数获得了良好的数值一致性。STKS的假阴性率为1.8%，诊断敏感性（96.3%）和特异性（83.3%）均较高。左移指征的假阴性率为1.9%，阴性检测的预测值为97.4%。然而，阳性检测的预测值仅为40.3%。实施了授权程序以减少假阴性样本。我们分类计数的常规实验室工作量可减少多达70%。库尔特STKS对新生儿样本给出了大量可疑标记（75%），因此不适合用于新生儿筛查。
