论文部分内容阅读
目的:建立识别新生儿院内真菌感染高危人群的风险评分方法,为预防性使用抗真菌药物提供依据。方法:通过回顾性调查及Logistic回归分析,筛查出新生儿发生院内真菌感染的独立危险因素,建立风险预测模型及评分方法,并利用受试者工作特征曲线(receiver operating characteristic curve,ROC)评估效能。结果:新生儿发生院内真菌感染的独立危险因素为极早产儿(孕龄<32周)、极低出生体重(<1 500 g)、细菌性败血症、使用广谱抗生素>5天、深静脉留置导管、气管插管,模型组和检验组中评分越高的新生儿发生院内真菌感染的几率越大,并且低、高得分组真菌感染发生率具有统计学差异(P<0.001),模型组和检验组ROC曲线下面积分别为0.868和0.859。结论:该风险预测模型能够帮助临床医生较准确地筛查出可能发生院内真菌感染的高危新生儿,指导预防性抗真菌药物使用。
OBJECTIVE: To establish a risk score method for identifying high risk groups of nosocomial fungal infections in newborns and provide evidence for the preventive use of antifungal agents. Methods: Through retrospective investigation and Logistic regression analysis, the independent risk factors of nosocomial fungal infection in neonates were screened out. The risk prediction model and the scoring method were established and evaluated by receiver operating characteristic curve (ROC) efficacy. RESULTS: The independent risk factors for nosocomial fungal infections in newborns were very preterm children (gestational age <32 weeks), very low birth weight (<1,500 g), bacterial sepsis, broad-spectrum antibiotics> 5 days, deep vein retention The higher the incidence of nosocomial fungal infections in catheter, tracheal intubation, model group and test group were, the higher the incidence of nosocomial fungal infection was and the lower incidence of high fungal infection was (P <0.001). The model group and The area under the ROC curve of the test group was 0.868 and 0.859 respectively. Conclusion: The risk prediction model can help clinicians more accurately screen high-risk neonates with nosocomial fungal infections and guide preventive antifungal drug use.