Selimović Amela, Skokić Fahrija, Selimović Zijad, Bazardzanović Mustafa
Odjeljenje za novorodencad Klinike za ginekologiju i akuserstvo, Univerzitetski klinicki centar Tuzla.
Med Arh. 2008;62(4):205-10.
Early-onset neonatal sepsis (EONS) is one of the most common diagnostic challenges in neonatal population. Aims of this study was to analyse values of total white blood count (WBC) and differential count in neonates with EONS and to determine cut-off values with the highest accuracy in diagnosis of infection.
In the retrospective-prospective research we include 341 neonates born from 37th to 42nd week of gestational age, with one or more obstetric risk factors for EONS, in single pregnancy, both genders and without visible anomalies. Capillary blood samples for analysed parameters levels were obtained in the first 72 hours of life. A diagnosis of EONS was based on microbiological findings, clinical signs and radiography of chest.
Obstetric risk for EONS at delivery had 341/12 298 (2.8%) neonates, and EONS occurred in 199/341 (58.4%) neonates. The mean total WBC count was 22.5 +/- 9.5 x 109/L, the mean absolute mature neutrophil count was 13.9 +/- 6.8 x 109/, while the mean total neutrophil count was 15.6 +/- 7.8 x 109/L. Neonates with EONS had medial of total immature neutrophil count of 1.2 x 109/L, immature to total neutrophil ratio (I:T ratio) of 0.09 and medial of immature to mature neutrophil ratio (I:M ratio) was 0.10. ROC curve analysis cut-off values of I:T (> 0.08) and I:M (> 0.09) ratio had acceptable accuracy in the diagnosis of EONS. The value of total WBC > 26.4 x 109/L, with accuracy of 88% predict EONS. High predictive value (82.9%) in EONS had total neutrophil count value of more than 17.6, and total immature neutrophil count value of more than 1.1 x l09/L predict presence of EONS with accuracy of 65.3%. Cut-off values of I:T (66.8%) and I:M (67.3%) ratio almost in the same percent predict EONS. In univariable predictive model, only cut-off value of mature neutrophil count was not significant in the detection of EONS, while cut-off values of other analysed parameters had significant predictive value (p < 0.05).
Total white blood count and differential count are changed in neonates with early-onset neonatal sepsis. The predictive value of analysed parameters cut-off values is important in everyday work of neonatology's to make difference among infected and non-infected neonates.
早发型新生儿败血症（EONS）是新生儿群体中最常见的诊断难题之一。本研究的目的是分析EONS新生儿的白细胞总数（WBC）和分类计数的值，并确定诊断感染时准确性最高的临界值。
在这项回顾性-前瞻性研究中，我们纳入了341例孕37至42周出生的新生儿，这些新生儿单胎妊娠，有一项或多项EONS产科危险因素，不限性别且无明显异常。在出生后的72小时内采集用于分析参数水平的毛细血管血样。EONS的诊断基于微生物学检查结果、临床体征和胸部X光检查。
分娩时EONS的产科危险因素在341/12298（2.8%）的新生儿中存在，EONS发生在199/341（58.4%）的新生儿中。白细胞总数的平均值为22.5±9.5×10⁹/L，绝对成熟中性粒细胞计数的平均值为13.9±6.8×10⁹/，而中性粒细胞总数的平均值为15.6±7.8×10⁹/L。患有EONS的新生儿未成熟中性粒细胞总数的中位数为1.2×10⁹/L，未成熟与总中性粒细胞比值（I:T比值）为0.09，未成熟与成熟中性粒细胞比值（I:M比值）的中位数为0.10。ROC曲线分析显示，I:T（>0.08）和I:M（>0.09）比值的临界值在EONS诊断中具有可接受的准确性。白细胞总数>26.4×10⁹/L，预测EONS的准确性为88%。EONS中总中性粒细胞计数>17.6的预测价值高（82.9%），未成熟中性粒细胞总数>1.1×10⁹/L预测EONS存在的准确性为65.3%。I:T（66.8%）和I:M（67.3%）比值的临界值预测EONS的比例几乎相同。在单变量预测模型中，只有成熟中性粒细胞计数的临界值在检测EONS时不显著，而其他分析参数的临界值具有显著的预测价值（p<0.05）。
早发型新生儿败血症新生儿的白细胞总数和分类计数会发生变化。分析参数临界值的预测价值在新生儿科的日常工作中对于区分感染和未感染新生儿很重要。
