Prediction Model For Hypertensive Disorders Of Pregnancy In Women That Have Undergone In Vitro Fertilization-embryo Transfer

Background and ObjectiveInfertility is a common global reproductive system diseases,according to the epidemiological research,the number of people suffering from infertility in the world is increasing year by year,among which 5%to 8%in developed countries and 30%in developing countries.[1]The etiology of infertility is complex.At present,it is divided into female infertility,male infertility,bilateral factors and unknown factors.China is a traditional country,and family succession is a top priority for every family.Due to the decline in the rate of marriage,the backward age of marriage and other factors,the fertility is gradually weakened,and the problem of infertility is becoming younger.Most of the couples who fail to prepare pregnancy will ultimately hope to rely on assisted reproductive technology(ART).With the development of science and technology and the birth of the world's first test tube baby born in Cambridge University more than 30 years ago,the ART era has officially started.ART refers to the technology of using medical aids to make infertile couples pregnant,which mainly including artificial insemination(AI),in vitro fertilization embryo transfer(IVF-ET)and its derivative technology.There are studies showing that women who received IVF-ET are more likely to have pregnancy-related complications and poor maternal and neonatal outcomes.Among them,hypertensive disorders of pregnancy(HDP)is the most important complication,and its incidence rate is increased by about 30%compared with natural pregnant women.[3-7]HDP is not only the main cause of maternal death,but also the main cause of perinatal death,placental abruption,premature delivery and intrauterine growth restriction.At present,the most effective treatment is delivery,but the risk of adverse pregnancy outcome events such as neonatal death,occupancy rate of neonatal intensive care unit(NICU)and neonatal respiratory system will increase significantly if the pregnancy is terminated immediately after diagnosis.[8]The purpose of this study was to explore the relationship between IVF-ET and HDP incidence,to establish a prediction model that can identify high-risk women of HDP incidence after IVF-ET and evaluate the effectiveness of its prediction.Materials and MethodsThe clinical data of 276 patients who received IVF-ET in the First Affiliated Hospital of Zhengzhou University from January 2013 to August 2016 were collected.The stage of disease development is defined as pre-transplant(final test results within one week before embryo transfer;stage ?),pre-onset(final test results within one week before the first diagnosis of HDP;stage ?)and onset(first test results within one week after the first diagnosis of HDP;stage ?).HDP in this study including gestational hypertension and preeclampsia.SPSS 21.0 and graphpad prism 5.0 were used to analyze blood routine,biochemical indexes and coagulation function of the sample in three stages.T-test,logistic regression curve,receiver operating characteristic curve(ROC)and unsupervised cluster analysis were used to screen and classify the parameters in the stage ? and ?.Five kinds of dynamic change models are established after identifying intersection by Venn diagram analysis.Combined with the data of stage ?,the key parameters of the model were further analyzed by using multiple Cox regression model.Kaplan-Meier(K-M)curve and ROC curve were used to compare the difference of predictive efficacy of different models for HDP onset and prognosis.the optimal prediction model was selected and evaluated by curve fitting method.Results276 patients were enrolled by screening.The cohort was divided into HDP group and non HDP group(NP group),there was no significant difference in age,weight,smoking and scar uterus between the two groups(P>0.05).Compared with NP group,HDP group was more likely to have headache(P<0.01),chest pain or dyspnea(P<0.05),blurred vision(P<0.01).For the adverse pregnancy outcome,the stillbirth rate(P<0.05),NICU occupancy rate(P<0.001)and the risk of neonatal asphyxia(P<0.05)in HDP group were significantly increased,and the birth weight(P<0.001)and the gestational age(P<0.01)were significantly lower and shorter.Statistical analysis showed that there were statistically significant differences in 5 and 16 parameters in Stage ? and Stage ?,respectively.Venn analysis showed that Hemoglobin(Hb)and serum potassium level in HDP group significantly increased in Stage ? and Stage ?,the differences were statistically significant(P<0.05).There are 3 parameters that change only in Stage ?(2 high level,1 low level),and 14 parameters that change only in Stage ?(5 high level,9 low level).Based on the change of parameters,five dynamic models are established.By comparing K-M curve and ROC curve,it can be concluded that in stage ?,model 1(AUROC=0.793,P<0.001)is better than model 2+3(AUROC=0.649,P<0.05).Similarly,the prediction level of model 1 at stage ?(AUROC=0.741,P<0.001)was better than that of model 4+5(AUROC=0.703,P<0.001).In stage I,model 2+3 and model 4+5 had no predictive effect on HDP,while model 1 had a better predictive effect on HDP(AUROC?0.831,P<0.001).The main adverse outcomes were stillbirth,neonatal intensive care unit and Low birth weight(LBW),the prediction effect of model 1 on adverse pregnancy events in stage ? was significantly better than that in other stages and any other model.Conclusion1.This study successfully established a prediction model that can recognize the risk of HDP(including gestational hypertension and preeclampsia)in women after IVF-ET operation to some extent,that is:Cox prediction model score=ln(1.677)×(Hb)+ln(2.209)×(serum potassium)).The score was positively correlated with the risk of HDP,and the prediction efficiency was inverse time-dependent trend.2.The prognosis of high-risk population is poor.It is suggested that close monitoring should be carried out 12 weeks after pregnancy,and delivery should be carried out immediately if necessary to reduce the occurrence of adverse pregnancy outcomes.3.Combined with the evaluation of prognosis ability,it is suggested that the prediction model has a certain prognostic effect.