Research On The Delivery Mode After Cesarean Section And Prediction Formulas

Objective:To explore the optimal delivery mode for pregnant women with history of cesareansection,to evaluate the applicability of the existing prediction model used in vaginal birthafter cesarean section in China, and to establish suitable prediction model for Chinese.Method：1. The survey was conducted from1439cases in No.1Affiliated Hospital of SuzhouUniversity, Suzhou Municipal Hospital, Wuxi Hospital of Maternal and Child Health Carefrom February1,2002to November30,2012. Among1439cases,255are scar uterus ofpregnant women who delivered in those hospitals, including Trial of labor (TOLAC) groupof cases, and rest1184cases are elective repeat cesarean delivery (ERCD) group. Inclusioncriteria are the last cesarean operation performed more than2years, single live births, nosurgical complications and health complications of pregnancy in pregnant women.2. In lights of Flamm, Grobman, Smith, Gonen’s prediction models to measure therisk of the TOLAC group, the theoretical and practical values were compared andapplicability for formulas were verified.3. By using the multivariate Logistic regression analysis and combining with theformula and clinical practice, we randomly selected60%TOLAC patients (153). Weperformed multiple factor Logistic regression analysis on ages, occupations, history ofvaginal delivery, cervical dilatation, amniotic fluid contamination, perinatal construction,height, weight, prenatal full-term newborns. We screened out the factors which fit thesituation of China and have significance for the prediction of VBAC and ERCD risk, andfurther explored the prediction formulas. Statistical processing:Numerical variables are shown as (X±S). SAS9.2statistical software was used. Themeasurement data were compared with Z test, t test. Calculation data were compared byusing χ2test, Fisher’s exact test, Kappa test. The data were also analyzed by operatingcharacteristic (receiver operating characteristic, ROC) logistic curves and multivariateregression analysis and nonparametric data were compared with the Wilcoxon rank sumtest.Result:1.255cases were trial of labor after cesarean(TOLAC), including233cases ofsuccessful trial (vaginal birth after cesarean, VBAC),22cases of failure of emergencycesarean section.1184cases were elective repeat cesarean delivery(ERCD),131caseswere parturient.The subjective intention of pregnant women on the choice of deliverymode played a dominant role.2. Relative to ERCD group, TOLAC group tend to be more no industry, no regularproduction inspection, emergency admission, light maternal weight; the fetus is smallerand gestational age is smaller; Dominating were women with histories of vaginal deliveryof pregnant, whose differences were statistically significant (P<0.05).3. In the maternal complications of TOLAC group and ERCD group, exceptincomplete uterine rupture of [0%(0/255)1.01%(12/1184)](χ2=4.53, P=0.03), otherdifferences were not statistically significant. Nevertheless, the vaginal trial group hospitalstay, hospital costs, the use of antibiotics, blood loss in the TOLAC group weresignificantly less than the ERCD group; postoperative hemoglobin loss was more than theRECD group. These differences were statistically significant (P<0.05).4. Neonatal complications of TOLAC group and ERCD group: the Apgar score≤7(1minute after birth)and neonatal transfer rate showed no significant difference (P>0.05)both in full-term infants group and premature infants group.5. In the Flamm formula, we used ROC analysis and the area under the curve is0.52,P=0.71. Then we used t test. t=-0.72, P=0.48, suggests that the formula of diagnostic efficiency is low. Similar to the Grobman formula, the area under the curve is0.67, P <0.05. t test shows t=2.58, P <0.05, suggesting that the formula had a certain diagnosticvalue. Then we used the same method into Gonen formula and the area under the curve is0.51, P=0.90, t test shows t=-0.19, P=0.85, suggesting that the formula of diagnosticefficiency is low. In Smith formula, the area of the curve (uncorrection of fetal sex) is0.65,P＜0.05，the critical value of42.16%. The AUC(correction of fetal sex) is0.66, P <0.05,the critical value of39.65%.t test shows t=-2.32, P <0.05(uncorrection of fetal sex) andt=-2.26, P <0.05(correction of fetal sex), which is with40%as the boundary. The Kappatest was used on the prediction results and the actual Kappa value=0.20, P<0.05(uncorrection of fetal sex). Kappa value is0.24, P <0.05(correction of fetal sex),suggesting that our experimental results are consistent with Smith formula. Smith formulahas a certain value in the diagnosis and the40%is defined as high-risk groups, which isapplicable in China.6. Among multivariate logistic regression analysis, amniotic fluid contamination,full-term or not and neonatal weight are meaningful factors: P <0.05. A new formula wasbuilt according to the results, then analyzed by ROC. The area under the curve is0.89, P <0.05. The AUC of the new prediction formula, the Grobman formula and the Smithformula(correction of fetal sex) were compared. We found that the AUC of the newformula was bigger than others with significant difference, P <0.05, suggesting that thenew prediction formula has better diagnostic performance than the Grobman and the Smithformula.Conclusion:1. VBAC is safe when the hospital has the ability of emergency caesarean operation,the obstetrician and anesthesia doctors, and controlled strictly. What is more, VBAC haslow cost, shorten the days in hospital patients, reduce the rate of cesarean section and has anumber of benefits.2. The subjective intention of maternal on the delivery mode selection plays adominant role, which suggests the subjective intention of the patients is one of the reasons causing the increase of cesarean section rate. Guiding pregnant women with the history ofcesarean section to choose the correct delivery mode should be one of the importantcontents of prenatal care.3. In the Flamm, Grobman, Gonen, Smith prediction formula, the Smith formula ismore applicable to the population of our country and the results of40%defined ashigh-risk groups is applicable in China. Our experimental results suggest that the newprediction formula diagnostic value is better than the Smith formula and Grobman formula,but the new formula still needs further improvement and verification.