| Objective Retrospectively analyze the clinical medical records of pregnant women with hepatitis B infection status and hepatitis B non-infected status,explore the relationship between hepatitis B virus infection status and pregnancy outcome,and try to establish relevant predictive models and evaluate its predictive value.Methods The singleton pregnant women over 18 years old who delivered in our hospital from 2009 to 2018 were selected.Inclusion criteria: Singleton pregnant women with 28 weeks of gestation and delivery in our hospital with relatively complete clinical data.Exclusion criteria:(1)patients with severe medical and surgical diseases such as chronic hypertension and heart disease;(2)latrogenic premature delivery;(3)severe mental disorders and communication disorders;(4)lack of data on maternity examination or hospitalization before pregnancy.After screening,there were 82775 cases,with an average age of 30 years old.Among them,31399 cases had one pregnancy and 51376 cases had more than one pregnancy.According to the status of hepatitis B infection,they were divided into 9699 cases of hepatitis B positive group and 73076 cases of normal group.According to the mode of delivery,6447 pregnant women in the HBV positive group and 49206 pregnant women in the negative group underwent vaginal delivery;3252 pregnant women in the positive group and 23870 cases in the negative group underwent cesarean delivery.Retrospective collection of pregnant women’s hospitalization data: general data(age,gestational week of delivery,gravidity,parity),delivery process(method of delivery,length of labor),pregnancy complications(postpartum hemorrhage,intrahepatic cholestasis of pregnancy,hypertension during pregnancy,gestational diabetes,placental abruption,premature delivery,premature rupture of membranes,etc.),perinatal outcome(neonatal weight,neonatal asphyxia,small for gestational age,migration neonatal department,stillbirth).Compare the difference of pregnancy outcome between hepatitis B negative group and hepatitis B positive group.The BP neural network model is one of the most practical neural networks.This study tried to apply the BP artificial neural network,with the hepatitis B virus infection status as one of the main characteristic values,combined with common clinical parameters,to predict related adverse pregnancy outcomes.Results1.General data comparison: the number of cases with gravidity> 1 in the positive group is 6376(65.74%),the number of cases in the negative group is 45000(61.58%),the difference was statistically significant(P<0.001).The number of cases with parity>0 in the positive group was 4849(49.99%),and the number of cases in the negative group was 33309(45.58%),the difference was statistically significant(P<0.001).There was no statistically significant difference in the age of delivery and gestational week between the two groups.2.Comparison of the two groups of pregnant women’s delivery process: the natural vaginal delivery rate of the positive group was lower than that of the negative group,the difference was statistically significant,P<0.05,but there was no statistically significant difference in the vaginal delivery rate,vaginal delivery rate,and cesarean section rate between the two groups.Comparing the length of labor of the pregnant women who gave birth to vagina,the average length of the first labor of the positive group was(388.51±273.82)minutes,and the length of the first labor of the negative group was(456.56±274.23)minutes.It can be considered that the length of the first labor of the positive group is shorter than that of the negative group.The difference was statistically significant.But comparing the duration of the second stage of labor,the difference was not statistically significant.3.Comparison of maternal complications between the two groups: the incidence of ICP and PPH in the positive group was higher than that in the negative group,the difference was statistically significant,P<0.05,(OR values were 3.40,1.16).The incidence of PROM,HDP,PE in the positive group was lower than that in the negative group,the difference was statistically significant,P<0.05(OR values were 0.91,0.83,0.80).The bleeding volume at the time of placenta delivery and 2 hours postpartum in the positive group was higher than that in the negative group,the difference was statistically significant,P<0.05.However,there was no significant difference in the incidence of GDM,PA,PTB,and HELLP between the two groups.4.Among pregnant women undergoing vaginal delivery,the comparison of maternal complications between the two groups: the incidence of PA,PPH,ICP,and PTB in the positive group was higher than that in the negative group,and the difference was statistically significant,P<0.05(OR values were 1.44,1.19,3.73,1.14).The incidence of PROM in the positive group was lower than that in the negative group,the difference was statistically significant,P<0.05,(OR value 0.88).However,there was no significant difference in the incidence of GDM,HDP,PE,HELLP and SGA between the two groups.5.Among pregnant women undergoing cesarean section,the two groups of maternal complications were compared: the incidence of ICP in the positive group was higher than that in the negative group,and the difference was statistically significant,P<0.05(OR value 3.06).The incidence of PE and SGA in the positive group was lower than that in the negative group,and the difference was statistically significant,P<0.05,(OR values were 0.80,0.65).However,there was no significant difference in the incidence of GDM,HDP,HELLP,PROM,PPH,PA,and PTB between the two groups.6.Comparison of newborn conditions between the two groups: the incidence of newborns with a weight of less than 3000 g in the positive group was higher than that in the negative group,and the difference was statistically significant,P<0.05(OR value1.06).However,there were no statistically significant differences in the incidence of neonatal weight between 3000-3499 g,3500-3999 g,≥4000g,small-for-gestational age infants,neonatal sex,neonatal asphyxia(mild and severe),PICU migration,and stillbirth between the two groups.For pregnant women undergoing natural vaginal delivery and cesarean section,there was no significant difference in the rate of neonatal asphyxia and PICU migration between the positive group and the negative group.7.The BP neural network model was built to predict the risk of postpartum hemorrhage,and the trained neural network is tested for accuracy using test data.The accuracy of the neural network is 85.3% and 84.1%.Conclusions1.HBV positive is a risk factor for a variety of adverse pregnancy outcomes in pregnant women.2.HBV-positive pregnant women are highly correlated with intrahepatic cholestasis of pregnancy and postpartum hemorrhage.During vaginal delivery,special attention should be paid to the risk of placental abruption.Whether it is vaginal delivery or cesarean delivery,many measures should be taken to reduce the occurrence of postpartum hemorrhage.3.The risk of premature babies in the HBV-positive group is higher.4.Combined HBV infection reduces the incidence of hypertension in pregnancy,suggesting that HBV virus may affect the immune function of the maternal-fetal interface,and its specific immune mechanism is worthy of further study.5.The BP neural network model can predict the risk of bleeding 2 hours postpartum.According to the status of hepatitis B infection,age,gestational age and the length of the first and second stages of labor,early warning of the risk of postpartum hemorrhage can be carried out.Through the relevant BP neural network model,it can provide reference for clinical decision-making. |
PDF Full Text Request