目的 探讨新生儿早发型败血症（EOS）预后不良的预测因素。方法 收集371例EOS新生儿的临床资料，根据预后分为预后良好组（264例）和预后不良组（107例）。比较两组患儿的围生期特点、临床表现、实验室指标、合并症、治疗过程等，采用多因素logistic回归模型分析EOS预后不良的预测因素。结果 预后不良组患儿的出生体重、胎龄均低于预后良好组（P < 0.05）；预后不良组的早产、低出生体重、极低出生体重及双胎比例高于预后良好组（P < 0.05）；预后不良组母亲产前使用激素、产前使用抗生素比例高于预后良好组（P < 0.05）；预后不良组反应差和呼吸功能低下的发生率高于预后良好组，皮肤黄染发生率低于预后良好组（P < 0.05）；预后不良组WBC<5×10⁹/L、血小板<100×10⁹/L、贫血、凝血功能障碍、肾功能损害、肝功能损害、低蛋白血症及低血糖的发生率高于预后良好组（P < 0.05）；预后不良组新生儿呼吸窘迫综合征、肺出血、坏死性小肠结肠炎、脑出血、脑损伤、肺动脉高压及休克的发生率高于预后良好组（P < 0.05）；预后不良组住院时间和抗生素疗程长于预后良好组（P < 0.05）；预后不良组使用机械通气和血管活性药的比例高于预后良好组（P < 0.05）。多因素logistic回归分析显示，极低出生体重（OR=41.734）、坏死性小肠结肠炎（OR=12.669）、脑损伤（OR=8.372）、休克（OR=5.889）、机械通气（OR=5.456）及肝功能损害（OR=4.075）是新生儿EOS预后不良的独立预测因素（P < 0.05）。结论 极低出生体重、机械通气及合并坏死性小肠结肠炎、脑损伤、休克或肝功能损害对新生儿EOS预后不良有预测价值。

Objective To study the predictive factors for poor prognosis of neonates with early-onset sepsis (EOS). Methods The clinical data of 371 neonates with EOS were collected. According to prognosis, they were divided into a good prognosis group with 264 neonates and a poor prognosis group with 107 neonates. The two groups were compared in terms of perinatal conditions, clinical manifestations, laboratory markers, comorbidities, and treatment process. Multivariate logistic regression analysis was used to investigate the predictive factors for poor prognosis of EOS. Results The poor prognosis group had significantly lower birth weight and gestational age than the good prognosis group (P < 0.05). Compared with the good prognosis group, the poor prognosis group had significantly higher proportions of preterm neonates, low birth weight neonates, very low birth weight neonates and twins (P < 0.05), as well as a significantly higher proportion of mothers who used hormone or antibiotics before delivery (P < 0.05). Compared with the good prognosis group, the poor prognosis group had significantly higher incidence rates of poor response and respiratory hypofunction (P < 0.05) and a significantly lower incidence rate of jaundice (P < 0.05). Compared with the good prognosis group, the poor prognosis group had significantly higher incidence rates of white blood cell count <5×10⁹/L, platelet count <100×10⁹/L, anemia, coagulation disorder, renal dysfunction, liver impairment, hypoproteinemia, and hypoglycemia (P < 0.05). The poor prognosis group had significantly higher incidence rates of neonatal respiratory distress syndrome, pulmonary hemorrhage, necrotizing enterocolitis, intraventricular hemorrhage, brain injury, pulmonary hypertension, and shock than the good prognosis group (P < 0.05). Compared with the good prognosis group, the poor prognosis group had significantly longer length of hospital stay and course of treatment with antibiotics (P < 0.05) and a significantly higher proportion of neonates receiving mechanical ventilation or vasoactive agents (P < 0.05). The multivariate logistic regression analysis showed that very low birth weight (OR=41.734), necrotizing enterocolitis (OR=12.669), brain injury (OR=8.372), shock (OR=5.889), mechanical ventilation (OR=5.456), and liver impairment (OR=4.075) were independent predictive factors for poor prognosis of neonates with EOS (P < 0.05). Conclusions Very low birth weight, mechanical ventilation, necrotizing enterocolitis, brain injury, shock, and liver impairment have a certain value in predicting the poor prognosis of neonates with EOS.