早发型新生儿败血症预后不良风险评估模型构建及其应用效果

Establishment of risk assessment model for poor prognosis of early onset neonatal sepsis and the application effect

目的构建早发型新生儿败血症预后不良风险评估模型,以期为临床预防护理提供参考。方法选取医院2020年12月—2021年12月收治的早发型新生儿败血症患儿共118例,根据患儿预后将其分为对照组(预后良好)106例和病例组(预后不良)12例。收集患儿及其母亲病历、实验室指标。采用单因素分析、多因素Logistic回归分析筛选早发型新生儿败血症预后不良的独立危险因素,并构建Logistic回归风险预测模型,对模型预测效果进行检验。结果单因素分析显示,早发型新生儿败血症预后不良的危险因素有出生体质量、低蛋白血症、PLT、WBC、肝功能损害、合并坏死性小肠结肠炎、合并脑损伤、合并病理性黄疸、休克、机械通气(P<0.05);多因素Logistic回归分析显示,早发型新生儿败血症预后不良的独立危险因素有出生体质量<2500 g、肝功能损害、合并坏死性小肠结肠炎、合并脑损伤、休克、机械通气(P<0.05)。根据回归分析结果构建模型:P_(预后不良)=1/[1+e^(-3.114+2.420×出生体质量(<2500 g)+2.955×肝功能损害+2.901×合并坏死性小肠结肠炎+2.661×合并脑损伤+3.476×休克+2.752×机械通气)]。ROC曲线分析显示,模型曲线下面积(AUC)=0.850,95%CI为0.806~0.895。根据约登指数最大值筛选的预测最佳临界值是P=0.410。以P>0.410为预测EOS预后不良的标准,研究对象的预测结果显示,灵敏度=71.43%,特异度=98.08%,Kappa值=0.741。预测结果与实际结果具有较高的真实性和一致性。结论构建的早发型新生儿败血症风险预测模型,可操作性强,具有一定的预测能力,但漏诊率相对较高需要进一步调整。

Objective To establish a risk assessment model for poor prognosis of early-onset neonatal sepsis,in order to provide reference for clinical prevention and nursing.Methods 118 children with early-onset neonatal sepsis admitted in the hospital from December 2020 to December 2021 were selected and divided into control group(good prognosis)106 cases and case group(bad prognosis)12 cases according to their prognosis.Medical records and laboratory indicators of the children and their mothers were collected.Univariate analysis and multivariate Logistic regression analysis were used to screen the independent risk factors for poor prognosis of early-onset neonatal sepsis,and the Logistic regression risk prediction model was constructed to test the prediction effect of the model.Results Univariate analysis showed that the risk factors for poor prognosis of early-onset neonatal sepsis were body mass,hypoproteinemia,PLT,WBC,liver function impairment,necrotizing enterocolitis,brain injury,pathological jaundice shock,mechanical ventilation(P<0.05).Multivariate Logistic regression analysis showed that the independent risk factors for poor prognosis of early-onset neonatal sepsis were body mass<2500 g,liver function impairment,necrotizing enterocolitis,brain injury,shock and mechanical ventilation(P<0.05).According to the results of regression analysis,the model was built:P_(Poor prognosis)=1/[1+e^(-3.114+2.420×body mass(<2500 g)+2.955×liver function impairment+2.901×necrotizing enterocolitis+2.661×brain injury+3.476×shock+2.752×mechanical ventilation)].ROC curve analysis showed that the area under the curve(AUC)of the model was 0.850,and 95%CI was 0.806-0.895.The pre-optimal critical value filtered according to the maximum Yoden exponent is P=0.410.With P>0.410 as the criterion for predicting poor prognosis of EOS,the prediction results of the subjects showed that sensitivity=71.43%,specificity=98.08%,and Kappa value=0.741.The predicted results have high authenticity and consistency with the actual results.Conclusion The established risk prediction model of early-onset neonatal sepsis is operable and had certain predictive ability,but the rate of missed diagnosis was relatively high and needs further adjustment.