Analysis Of Factors Influencing Neonatal Meconium Aspiration Syndrome And Construction Of Prediction Model

Purpose:1.To explore the influencing factors of neonatal meconium aspiration syndrome(MAS),and to construct a predictive model of MAS occurrence to evaluate its incidence,and to develop a quantitative evaluation tool that can assist clinicians to predict MAS.2.Compare the influencing factors of light,moderate and severe MAS.Objects and methods:The clinical data of 458 children with meconium-stained meconium-stained amniotic fluid were retrospectively analyzed.They were divided into MAS group and MAS group,and the related factors of MAS were analyzed.Modeling 75% and 25%of the original dataset as modeling set and validation set respectively,then establishing a full independent variable prediction model(Full Model)and using stepwise regression to filter the independent variables to establish a simplified prediction model(Stepwise)Model).The ROC curve and the calibration curve are used to evaluate the discrimination and calibration degree of the model respectively.Finally,the optimal model is selected to establish a nomogram model for predicting the risk of MAS occurrence.The children with MAS were divided into three groups:light,medium and heavy.Result:1.By univariate analysis and multiple regression analysis,gestational age,hyperbilirubinemia,amniotic fluid contamination,arterial PH at admission,cesarean section,1 minute and 5 minute Apgar scores were independent of MAS.The influencing factors were statistically significant(P<0.05).2.The area under the ROC curve of Full modell and stepwise model are 0.856 and 0.845,respectively,P=0.126>0.05.The difference between the two is not statistically significant and has good accuracy.We finally choose the stepwise model with less variables.As a predictive model.Then,324 cases were included in the modeling group,and 118 cases were included in the verification group.The areaunder the ROC curve of the two groups was 0.817 and 0.833,respectively,and the calibration curve showed that the calibration degree was also higher.The model had better prediction performance,and finally the R software was used.Make a MAS risk line graph.3.Comparison of influencing factors of mild to moderate MAS showed that there was no direct correlation between amniotic fluid pollution index and severity of MAS;low Apgar score was a high risk factor for severe MAS,but not suitable for light/moderate MAS;Caesarean section and gestational age had no significant effect on the severe classification of MAS.Conclusion:1.The occurrence of MAS was associated with gestational age,presence or absence of hyperbilirubinemia,amniotic fluid contamination,hospital admission,cesarean section,1 minute and 5 minute Apgar score.2.MAS nomogram can quantitatively assess the risk of MAS,and has a good evaluation efficiency,which can help clinicians identify high-risk children with MAS at an early stage,which has clinical application value.3.Heavy amniotic fluid pollution and low Apgar score are high risk factors for heavy MAS,and there is no linear correlation between MAS severity and influencing factors.