基于Logistic模型与CART决策树模型构建新生儿低血糖风险预测效能的比较

打开文本图片集

Comparison of the predictive efficacy of neonatal hypoglycemia risk based on the Logistic model and the CART decision tree model

SuzhouHospitalAffiliatedtoNanjingMedicalUniversity，Jiangsu2151ooChina *Corresponding Author CHEN Qingqing，E-mail：276061163@qq.com

AbstractObjective：Toanalyzetheriskfctorsofneonatalhypoglycemiaandestablishaisk warning modelforneonatalhypoglyemia byusingthedecisiontreeCARTalgoritm.Metods：Theclinicaldataof235newborsandtheirmothersdeliveredinourhospitalfrom March202ltoNovember223wereretrospectivelyanalyzed.Theyweredividedintohypoglycemiagroupandnon-hypoglycemiagroup acording totheresultsofblood glucosetest.Thepredictionmodelforneonatalhypoglycemia wasconstructedbyLogisticregesion modelanddecisiotreCARodel.Te5foldosldationmthodwasuedforintealvalidationndtepredictieefyof themodelwascompared.Results：Amongthe235ewboms，36newbornsdevelopedhypoglycemia，199newborsdidnotdevelop hypoglycemia，and the incidence of hypoglycemia was 15.32% . Logistic regression analysis showed that gestational diabetes，mode of delivery，prematurefant，evelopent，rthightndeastfeingtieredepedetsactosfooatala （ P<0.05 ）. The overall accuracy of the probability prediction model was 81.8% .After5-fold cross validation，the prediction accuracy was （20 72.5% .The decision tree modelshowed thatpremature infants werethe mostimportant influencing factorfor neonatal hypoglycemia， withaninformationgainofO.4O.TeAUCvalueoftheLgisticregressionmodelwasslightlyhighertanthatoftedecisotr.886 vs0.854），andthepredictioneficiencyofbothodelsasoderate.Conclusins：Gestationaldabetes，esareanectionatre infant，small for gestational age infant，low birth weight infant，and breastfeeding time |⩾2h were all influencing factors for neonatal hypoglycemia.Clinically，corresponding preventionand treatment plans could beformulated based ontheabove factors.

Keywordsneonate; hypoglycemia; influencing factors;decision tree;prediction efficacy; nursing

摘要目的：分析新生儿低血糖症的危险因素，并通过分类回归树算法（CART）建立新生儿低血糖症的风险预警模型。（剩余8395字）