Prediction Model Study In The Application Of Vaginal Birth After Cesarean Delivery

Objective:The purpose of this study is to analyze the prediction models reported in foreign countries,and combine the maternal demographic characteristics and obstetric clinical data.Multivariate Logistic regression analysis was used to calculate the meaningful variables for the outcome of labor.And develop a model that predicts the outcome of vaginal birth after cesarean delivery?VBAC?for a Chinese population.In order to provide the midwife and obstetric medical stuff with accurate assessment of maternal with prior cesarean delivery.It can also provides a basis for the pregnant woman to determine the appropriate delivery mode,thus reducing the secondary cesarean delivery rates.Methods:In 2014,the annual report of the national women and children's annual report showed that the rate of cesarean section was 35%.According to the counting data,the sample size was calculated as 350 cases based on the estimation of the total rate.This prospective study in which included 570 maternal with a history of previous cesarean delivery as well as a delivery at term from July to September,2017.There were 62 cases of trial of labor after cesarean?TOLAC?and 508 cases of elective repeat cesarean delivery?ERCD?.1.The prospective study method was used to select 570 pregnant women according to inclusion exclusion criteria.,and compared the effects of different delivery methods on complications after admission and childbirth.Combined with the main variables and clinical practice of obstetrics in foreign countries,the data were included in prenatal,obstetric and postpartum.Prenatal data including age,height,urban and rural,admitted at hospital BMI?kg/m²?,gestational age?at term?,prior vaginal delivery,prior VBAC,recurrent indication for a cesarean delivery,prior vaginal delivery before cesarean,induct of labor,gestational diabetes mellitus?GDM?,hypertension and premature rupture of membrane?PROM?.Intrapartum data includes:the first,second,third time of labor.Postpartum data include:postpartum hemorrhagic volume,Apgar score in 1 min,blood transfusion,fetal weight,uterine rupture.Influence factors were obtain by analyzing that above data in relation to the outcome of different childbirth for VBAC and ERCD groups.2.Using the univariate analysis method,the significant variables of the VBAC and ERCD were substituted into the Logistic multiple regression analysis.The variables with statistical significance?P<0.05?were calculated by Logistic Forward method?P<0.05?,and the regression coefficient and 95%confidence interval were calculated.3.According to the Odds Ratio?OR?,the weight of each variable was determined,and the regression coefficient was substituted into the multivariate regression model,and the prediction model of VBAC was established.The efficiency of the model was tested by the Hosmer-lemeshow,sensitivity and specificity,and the area under curve?AUC?in the receiver operating characteristic?ROC?.Result1.During the study period,570 maternal had a history of cesarean delivery.A total of 62maternal had a trial of labor,508 women had a ERCD outcome.In these 62 maternal,54 had a successful VBAC?87.1%?.Eight cases of emergency cesarean delivery.No uterine rupture was observed in the study.2.Two significant variables were admitted at hospital BMI?kg/m²?and birth weight?P<0.05?.It was found that the VBAC outcome was significantly more likely among maternal who had a low admitted at hospital BMI and a low birth weight.The VBAC group had low birth weight and low fetal weight compared with ERCD group.There was no statistically significant difference in age,height,and gestational age?P>0.05?.There was also no statistically significant difference between the two groups of Apgar score,blood transfusion rate,postpartum hemorrhage and uterine rupture.3.It was statistically significant that that GDM variable had statistical significance in Chi-square test?P<0.05?.In the obstetric history data analysis and comparison,the prior vaginal delivery,prior VBAC,recurrent indication for a cesarean delivery and prior vaginal delivery before cesarean,this four variables have significant statistical significance?P<0.001?.There was no statistically significant difference between urban and rural,prior induct of labor and premature rupture of membranes?PROM?.4.Single factor analysis of seven meaningful variables were obtained?admission at hospital BMI,birth weight,GDM,prior vaginal delivery,prior VBAC,recurrent indication for a cesarean delivery,prior vaginal delivery before cesarean?.The above variables were analyzed by Logistic regression and identified the four variables that fitted the selection criteria.After the adjustment of regression coefficients,according to the weights of odds ratio?OR value?,and from high to low is arranged in the order,the four variables were:prior vaginal delivery?OR=21.74,95%CI:8.68-54.49?,GDM?OR=0.16,95%CI:0.04-0.71?,birth weight?OR=0.99,95%CI:0.98-1.00?and admission at hospital?OR=0.83,95%CI:0.70-0.98?.5.Based on the above single factors and Logistic multiple regression results,a delivery prediction model of VBAC was established.Variables entered into the model include prior vaginal delivery?X6?,birth weight?X11?,admission at hospital BMI?X5?and GDM?X13?.The final logistic equation used to calculate the individual probability of the VBAC in maternal is:P=1/{1+exp^{[-?5.93+3.08X6+0.84X11-0.19X5-1.85X13?]}}.That is P?success?=100%×1/{1+exp[-?5.93+3.08 prior vaginal delivery+0.84 birth weight-0.19 admission at hospital BMI-1.85 GDM?]}.Model test of effectiveness?Hosmer-Lemeshow?was no statistically significant difference?P>0.05?.It was preferable that that fit of the model was well.It was found that that overall prediction rate was 91.9%,the sensitivity was 98.6%,the specificity was 27.8%respectively.The receiver operating characteristic curve,with an area under the curve of 0.785?95%CI:0.725-0.854?,indicated a good discriminative ability.Conclusion1.The results of single factor analysis showed that the statistical results of VBAC group and ERCD group are statistically significant.In which that value of both BMI and birth weight are negative correlation to the rate of vaginal birth after cesarean delivery.2.This study concluded that the prior vaginal delivery,GDM,birth weight and admission BMI of four clinical variables have a certain effect on VBAC outcome,and the Logistic multiple regression prediction model is constructed.3.Based on the prediction model established by Logistic multiple regression,the performance test of the model fitness,sensitivity,specificity and ROC curve shows that the new VBAC prediction model indicated a good discriminative ability.