| Chapter One:Analysis of the related factors of pregnancy in IVF-ET and build the forecast modelObjectiveTo investigate the relationship of pregnancy outcome with clinical characteristics, this investigative purpose is to try to look for which factors will affect or reflect to the occurrence of pregnancy in vitro fertilization and embryo transfer(IVF-ET). analyze the diagnosis of node of these factors, and establish the forecast model of pregnancy.explore the reasons for the occurrence of pregnancy and provide a basis to improve the pregnancy outcome.Materials and methodsA retrospective study of1441embryo-transfer cycles about ivf and intracytoplasmic sperm injection (icsi) was performed at Reproductive Medicine Center of Nanfang Hospital between January2008and December2010, The study group included780pregnancy cycles, while the control group consisted of661cycles in patients who is failure of ET.To investigate the relationship of pregnancy outcome with clinical characteristics, basal state, ovulation program, serum hormone levels on the day of administration of hCG, oocyte and embryo quality of pregnancy. This investigative purpose is to analyze the diagnosis of node of these factors, and build the forecast model, and then verificate it’s effectiveness.SPSS13.0statistical package was used for data analysis. Enumeration data were analyzed by non-parametric rank sum test. Comparison between groups was done by t test and that between rates of samples was done by χ2test. Unconditional logistic regression model was used to analyzed the significant risk factors. Predictive critical value was determined by the ROC curve and Youden index or based on accepted numerical obtained from reference. All P-values were two-sided, and p<0.05was considered statistically significant.80%of data was selected to analyzing with non-conditional logistic regression. The factors that was proved affect to the occurrence of pregnancy, and establish the forecast model of pregnancy. Other20%of data were ineluded according to the same criteria with the model group, the validation group. The validation group was analyzed retrospectively. Using the probability theory to draw the receiver operating characteristic curve (roc), In order to validate the equation of test efficiency.Results1.Single factor analysisContrasting to control group, pregnancy group had lower ages, but had more basic ovarian follicles. Lower total dosage of Gn, but had more times of Gn. pregnancy group also had lower serum LH and P on the day of hCG injection, lower ratio of P/E2. There were aslo significant diversities in oocyte retrieval rate, M Ⅱ oocyte percentage, fertilization rate, normal fertilization rate, the number of high-grade embryo between the two groups (p<0.05).2. Multivariate analysis and buliding the forecast model.Multivariate logistic regression was employed based on the univariate analysis. Datas was divided into four parts with clinical characteristics, basal state, ovulation program, serum hormone levels on the day of administration of hCG, oocyte and embryo quality of pregnancy. Pregnancy was positive outcomes. Eitght affecting factors was summarized with statistical significance in IVF. Ages (OR=0.974, p=0.048, with the95%confidence interval0.948-1.000). Number of cycles (OR=0.794, p=0.014, with the95%confidence interval0.661-0.955).Basic ovarian follicles (OR=1.039, p=0.000, with the95%confidence interval1.023-1.056); total dosage of Gn (OR=0.904, p=0.000, with the95%confidence interval0.786-0.869); times of Gn (OR=0.630, p=0.001, with the95%confidence interval1.070-1.112); The number of follicles which is bigger than14mm (OR=1.031, p=0.028, with the95%confidence interval1.003-1.059); P on the day of hCG injection (OR=0.630, p=0.000, with the95%confidence interval0.488-0.813); Number of high-grade embryo (OR=1.149, p=0.000, with the95%confidence interval1.105-1.194). The data above were significant differences statistically (p<0.05).3. Analysis the diagnosis of node of these factors, using ROC curves or layered logistic regression analysis. The area under the curve for ages to pregnancy was0.544, with the95%confidence interval0.514-0.573,(p=0.004).The maximum youden’s index corresponded with35.5years. The pregnancy probability in cycles less than3group, is2.137times to cycles more than2group, with the95%confidence interval1.280-3.569,(p=0.004). The area under the curve for basic ovarian follicles to pregnancy was0.586, with the95%confidence interval0.557-0.616,(p=0.000).The maximum youden’s index corresponded with number12.5. The area under the curve for total dosage of Gn to pregnancy was0.561, with the95%confidence interval0.531-0.590,(p=0.000).The maximum youden’s index corresponded with3237.5IU/L. The area under the curve for times of Gn to pregnancy was0.538, with the95%confidence interval0.508-0.568,(p=0.013). The maximum youden’s index corresponded with10.5days. The pregnancy probability in group two (The number of follicles which is bigger than14mm is more than5), is0.731times to group one (The number of follicles which is bigger than14mm is less than6), with the95%confidence interval0.549-0.973. The pregnancy probability of the first group (P<0.6ng/ml) about P on the day of hCG injection was1.757times more than the forth (P≥1.2ng/ml). The second group (0.6ng/ml<P<0.9ng/ml) was1.384times more than the forth group,(p<0.05).The third groups (0.9ng/ml≤P<1.2ng/ml) was1.361times more than the forth group,(p=0.075). The first group (P<0.6ng/ml) lead to the optimum pregnancy probability. The area under the curve for number of high-grade embryo to pregnancy was0.625, with the95%confidence interval0.596-0.654,(p=0.000). The maximum youden’s index corresponded with number2.5.4. Random80%of data were selected form all the1441cycle to analyzing with non-conditional logistic regression. The eight factors that are significant in single factor analysis were bring into logistic regression, using Neter method, to establish a model for predicting the pregnancy probability in patients with IVF-ET. The formula for calculation is as follows:P=1/(1+y)(P means the probability to pregnancy); y=exp{—(0.588-0.014×Ages-0.236×Number of cycles+0.021xBasic ovarian follicles-0.076xTotal dosage of Gn/1000+0.021×Times of Gn-0.021×The number of follicles which is bigger than14mm-0.405×P on the day of hCG injection+0.145×Number of high-grade embryo)}Other20%of data were ineluded according to the same criteria with the model group, the validation group. The validation group was analyzed retrospectively. Using the probability theory to draw the receiver operating characteristic curve (roc), In order to validate the equation of test efficiency. The area under the curve for pregnancy probability was0.640, with the95%confidence interval0.574-0.707,(p=0.000).Conclusion1. The factors related to pregnancy in IVF-ET are very complicated, such as ages, number of cycles, basic number of Ovarian follicles, total dosage of Gn, the day Gn used, number of follicles which is bigger than14mm, P on the day of hCG injection, high-grade embryo.2. The patients over35.5years old were harder to pregnant. The patients below3times of number of cycles, pregnancy is higher than the patients equal to or older than3times. The patients whose Basic sinus follicular number greater than12.5were more likely to pregnancy. The patients whose total dosage of Gn less than3237.5IU/L were more likely to pregnancy. The patients whose times of Gn more than10.5days were more likely to pregnancy. The patients whose number of follicles which is bigger than14mm more than6were more likely to pregnancy. The patients whose progesterone on the day of hCG injection below0.6ng/ml, lead to the optimum pregnancy probability. While the patients whose P on the day of hCG injection upon 1.2ng/ml lesd to lowest pregnancy rate. The patients whose number of high-grade embryo more than2.5were more likely to pregnancy.3. HCG injection decision can reference the progesterone and the number of follicles which is bigger than14mm. The patients whose progesterone on the day of hCG injection below0.6ng/ml, lead to the optimum pregnancy probability. While the patients whose progesterone on the day of hCG injection upon1.2ng/ml lesd to lowest pregnancy rate. So Frozen embryo transfer should be used when progesterone on the day of hCG injection was more than1.2ng/ml. When the number of follicles was bigger than14mm, should Decided to hCG injection.Chapter Two:Analysis of the related factors of the pregnancy of frozen embryo transfers after fresh embryo transfer failureObjectiveTo investigate the relationship of pregnancy outcome with clinical characteristics, this investigative purpose is to try to look for which factors will affect or reflect to the pregnancy probability of frozen embryo transfers after fresh embryo transfer failure. Explore the reasons for the occurrence of pregnancy and provide a basis to improve the pregnancy outcome.Materials and methodsThe study group included780pregnancy cycles. while the control group consisted of170cycles, who obtained pregnancy of frozen embryo transfers after fresh embryo transfer failure.To investigate the relationship of pregnancy outcome with clinical characteristics, basal state, ovulation program, serum hormone levels on the day of administration of hCG, oocyte and embryo quality of pregnancy. Statistical analysis of univariate and multiple logistic regression were carried out for analysis. This investigative purpose is to analyze the Influence factors to pregnancy outcome in FET.SPSS13.0statistical package was used for data analysis. Enumeration data were analyzed by non-parametric rank sum test. Comparison between groups was done by T test and that between rates of samples was done by x2test. Unconditional logistic regression model was used to analyzed the significant risk factors. Predictive critical value was determined by the ROC curve and Youden index or based on accepted numerical obtained from reference. All P-values were two-sided, and p<0.05was considered statistically significant.Results1.Single factor analysisContrasting to control group, pregnancy group had lower serum E2and P on the day of hCG injection, the number of follicles which is bigger than10mm and14mm, oocyte retrieval rate, M Ⅱ oocyte percentage, number of high-grade embryo, and E2concentration in every ovarian follicles which is bigger than14mm, The data above were significant differences statistically (p<0.05).2. Multivariate analysisSerum E2/1000and P on the day of hCG injection are the independent influential factors in FET pregnancy cycles in fresh cycles. Serum progesterone on the day of hCG injection (OR=1.653, p=0.015, with the95%confidence interval1.101-2.482). Serum estrogen/1000on the day of hCG injection (OR=1.219, p=0.001, with the95%confidence interval1.085-1.369)3. The area under the curve for Serum progesterone on the day of hCG injection to pregnancy was0.566, with the95%confidence interval0.518-0.613,(P=0.007).The maximum youden’s index corresponded with0.6ng/ml. The area under the curve for Serum estrogen/1000on the day of hCG injection to pregnancy was0.600, with the95%confidence interval0.555-0.645,(p=0.000).The maximum youden’s index corresponded with2330pg/ml.CONCLUSIONSelevated progesterone and estradiol on the day of hCG injection was detrimental to pregnancy rates. In combination with the elevated progesterone, high estradiol concentrations had a potential negative effect. For these patients, FET should be suggested to improve the pregnancy outcomes. |
PDF Full Text Request