分类回归树分析确定的预测早发性新生儿败血症经验性抗菌治疗失败的临床参数。
Clinical parameters predicting failure of empirical antibacterial therapy in early onset neonatal sepsis, identified by classification and regression tree analysis.
机构信息
Paediatric Intensive Care Unit, Clinic of Anaesthesiology and Intensive Care, Tartu University Clinics, Lunini 6, 51014 Tartu, Estonia.
出版信息
BMC Pediatr. 2009 Nov 24;9:72. doi: 10.1186/1471-2431-9-72.
BACKGROUND
About 10-20% of neonates with suspected or proven early onset sepsis (EOS) fail on the empiric antibiotic regimen of ampicillin or penicillin and gentamicin. We aimed to identify clinical and laboratory markers associated with empiric antibiotic treatment failure in neonates with suspected EOS.
METHODS
Maternal and early neonatal characteristics predicting failure of empiric antibiotic treatment were identified by univariate logistic regression analysis from a prospective database of 283 neonates admitted to neonatal intensive care unit within 72 hours of life and requiring antibiotic therapy with penicillin or ampicillin and gentamicin. Variables, identified as significant by univariate analysis, were entered into stepwise multiple logistic regression (MLR) analysis and classification and regression tree (CRT) analysis to develop a decision algorithm for clinical application. In order to ensure the earliest possible timing separate analysis for 24 and 72 hours of age was performed.
RESULTS
At 24 hours of age neonates with hypoglycaemia < or = 2.55 mmol/L together with CRP values > 1.35 mg/L or those with BW < or = 678 g had more than 30% likelihood of treatment failure. In normoglycaemic neonates with higher BW the best predictors of treatment failure at 24 hours were GA < or = 27 weeks and among those, with higher GA, WBC < or = 8.25 x 10(9) L(-1) together with platelet count < or = 143 x 10(9) L(-1). The algorithm allowed capture of 75% of treatment failure cases with a specificity of 89%. By 72 hours of age minimum platelet count < or = 94.5 x 10(9) L(-1) with need for vasoactive treatment or leukopaenia < or = 3.5 x 10(9) L(-1) or leukocytosis > 39.8 x 10(9) L(-1) or blood glucose < or = 1.65 mmol/L allowed capture of 81% of treatment failure cases with the specificity of 88%. The performance of MLR and CRT models was similar, except for higher specificity of the CRT at 72 h, compared to MLR analysis.
CONCLUSION
There is an identifiable group of neonates with high risk of EOS, likely to fail on conventional antibiotic therapy.
背景
约 10-20%疑似或确诊早发性败血症（EOS）的新生儿在经验性抗生素治疗方案（氨苄青霉素或青霉素和庆大霉素）中失败。我们旨在确定与疑似 EOS 新生儿经验性抗生素治疗失败相关的临床和实验室标志物。
方法
通过对 283 名出生后 72 小时内入住新生儿重症监护病房并需要青霉素或氨苄青霉素和庆大霉素治疗的新生儿的前瞻性数据库进行单变量逻辑回归分析，确定了预测经验性抗生素治疗失败的母体和早期新生儿特征。通过单变量分析确定的变量被纳入逐步多变量逻辑回归（MLR）分析和分类回归树（CRT）分析，以开发用于临床应用的决策算法。为了确保尽可能早地进行分析，对 24 小时和 72 小时龄的新生儿分别进行了分析。
结果
在 24 小时龄时，血糖<或=2.55mmol/L 合并 CRP 值>1.35mg/L 的新生儿或 BW<或=678g 的新生儿发生治疗失败的可能性超过 30%。在血糖正常的 BW 较高的新生儿中，24 小时龄时治疗失败的最佳预测因子是 GA<或=27 周，而在 GA 较高的新生儿中，WBC<或=8.25x10(9)L(-1)合并血小板计数<或=143x10(9)L(-1)。该算法可以捕获 75%的治疗失败病例，特异性为 89%。到 72 小时龄时，血小板计数<或=94.5x10(9)L(-1)需要血管活性治疗或白细胞<或=3.5x10(9)L(-1)或白细胞增多>39.8x10(9)L(-1)或血糖<或=1.65mmol/L 可捕获 81%的治疗失败病例，特异性为 88%。MLR 和 CRT 模型的性能相似，除了 CRT 在 72 小时的特异性高于 MLR 分析。
结论
存在一组具有高 EOS 风险的新生儿，他们可能在常规抗生素治疗中失败。
Zotero 插件