Multiple Indexes Estimating The Status Of Perinatal Fetal Outcome

BackgroundWith the improvement of the quality of perinatal care, antenatal fetal surveillance and neonatal outcome is getting more and more important in modern obstetrics. The goal of prenatal care is close surveillance of the pregnancy with the aim of avoiding intrauterine fetal death and other adverse prenatal and neonatal outcomes. A better understanding and using of fetal surveillance indexes has a great important clinical significance to achieve this goal. Antenatal fetal surveillance methods mainly include fetal movement account, non-stress test (NST), contraction stress test (CST), biophysical profile scores (BPS), Doppler ultrasound measurements of blood flow etc. NST, BPS and Doppler ultrasound measurements of umbilical artery (UA) blood flow are most commonly used because of their safety, non-invasiveness and repetition. However, any method has its limitations and can not get a full picture of the fetal status in the cavity of uterus.At present, foreign scholars have made broad researches on the fetal risk scoring system that incorporates several fetal surveillance indexes. Domestic researches on this field lag behind comparatively. The aim of this paper is to develop and analyze a new score that incorporates NST, an estimated fetal weight, BPS, amniotic fluid volume and fetal umbilical arterial blood flow to assess the risk for poor perinatal outcome.Objective1. To develop a perinatal fetal outcome assessment score (PFOAS) that incorporates NST, an estimation of fetal weight, BPS, amniotic fluid volume and fetal umbilical arterial blood flow.2. To analyze the value and limitation of NST for the prediction of poor perinatal outcome.3. To analyze the value and limitation of BPS for the prediction of poor perinatal outcome.4. To analyze the value and limitation of UA systolic/diastolic (S/D) ratio, UA pulsatility index (PI) and UA resistance index (RI) number for the prediction of poor perinatal outcome.5. To analyze the value and limitation of PFOAS for the prediction of poor perinatal outcome.Research data and methods1. Research dataData of748patients who were examined and delivered in The First Affiliated Hospital of Guangzhou University of Chinese Medicine from June2012to December2013were analysed. Gestation age were between34and42weeks (mean number was371/7±1.61weeks). Maternal age were between20and40years (mean number was28.2±3.4years). Inclusion criteria consisted of (1) The end of the menstrual was clear. The menstrual cycle was regular (28-32days) and the first trimester ultrasound examination determined gestational age, ultrasound testing indicators included the crown-rump length or biparietal diameter.(2)Singleton.(3) Pregnant women without abnormal history of pregnant and other pregnancy risk factors.(4)The fetuses had no structures or chromosomal abnormalities. 2. Research methodsNST:The non-stress test was performed using cardiotocography with the patient in the left lateral recumbent position. The recording lasted at least20minutes. If the fetal heart rate was "non-reactive" after20minutes of testing, the recording continued for another20minutes to account the average length of periods of non-rapid eye movement sleep when fetal movement and subsequently heart rate variability were reduced. NST was analyzed for the prediction of poor perinatal outcome using Chi-Square test.BPS:The biophysical profile was scored according to "Ten Minute" criteria using Ultrasound examination. BPS was analyzed for the prediction of poor perinatal outcome using receiver operating characteristic (ROC) curve.UA blood flow:UA waveforms were obtained from a free loop of the umbilical cord using color Doppler ultrasound sonography. UA PI number, UA RI number and S/D ration were scored. UA PI number and S/D ration were analyzed for the prediction of poor perinatal outcome using ROC curve.PFOAS:PFOAS incorporates NST, the estimation of fetal weight (EFW), amniotic fluid volume (AFV), fetal tone (FT) and UA blood flow indexes. Each parameter was arbitrarily assigned a numeric score. A nonreactive or suspicious NST, amniotic fluid volume between2and3centimeters, an estimated fetal weight less than the10th percentile for gestational age and an abnormal FT were each given a score of one. An estimated fetal weight less than the5th percentile for gestational age and amniotic fluid volume of less than2centimeters were each given two scores of one. Abnormal fetal UA blood flow was given a score of between one and four for increasing severity of the blood flow patterns. The scoring system was then applied to each of the patient’s test results. PFOAS was analyzed for the prediction of poor perinatal outcome using ROC curve. PFOAS was compared with the NST, BPS, UA S/D, UA RI and U A PI for their ability to predict poor perinatal outcome.Poor perinatal outcome:Poor perinatal outcome is a composite set of variables, including the following:amniotic fluid contamination, neonatal weight was less than the10th percentile of the fetal weight for gestational age, a1-minute Apgar score≤7, admission to neonatal intensive care unit (NICU) and intrauterine fetal death.Statistical Analysis:Data were expressed as the geometric means±tandard error of mean of experiments. The prediction value of NST, BPS, UA S/D, UA PI, UA RI and PFOAS were done using Chi-Square test and ROC, using statistical software packages of SPSS19.0.Results1. The incidence of poor perinatal outcome in suspicious or non-reactive NST group was significantly higher than that in the reactive NST one. To predicate poor perinatal outcome, NST had a sensitivity of75.22%, a specificity of86.14%, a positive predictive value of95.13%, a negative predictive value of49.13%, a positive likelihood ration of0.2877and a negative likelihood ration of5.4279.2. The area under the ROC of BPS was0.825. The incidence of poor perinatal outcome in the BPS<5group was significantly higher than that in the BPS≥5one. To predicate poor perinatal outcome, BPS had a sensitivity of61.06%, a specificity of94.02%, a positive predictive value of93.14%, a negative predictive value of64.49%, a positive likelihood ration of0.4142and a negative likelihood ration of10.2038.3. The area under the ROC of UA PI was0.759. The incidence of poor perinatal outcome in the PI≥0.95group was significantly higher than that in the PI<0.95one. To predicate poor perinatal outcome, UA PI had a sensitivity of56.64%, a specificity of93.39%, a positive predictive value of92.37%, a negative predictive value of60.38%, a positive likelihood ration of0.4643and a negative likelihood ration of8.5630. 4. The area under the ROC of UA S/D was0.653. The incidence of poor perinatal outcome in the S/D≥2.87group was significantly higher than that in the S/D<2.87one. To predicate poor perinatal outcome, UA S/D had a sensitivity of40.71%, a specificity of90.24%, a positive predictive value of89.53%, a negative predictive value of42.59%, a positive likelihood ration of0.6571and a negative likelihood ration of4.1693.5. The area under the ROC of UA RI was0.693. The incidence of poor perinatal outcome in the RI≥0.65group was significantly higher than that in the R1<0.65one. To predicate poor perinatal outcome, UA RI had a sensitivity of27.43%, a specificity of94.80%, a positive predictive value of88.01%, a negative predictive value of48.44%, a positive likelihood ration of0.7654and a negative likelihood ration of5.2789.6. The area under the ROC of PFOAS was0.966. The incidence of poor perinatal outcome in the PFOAS≥4group was significantly higher than that in the PFOAS<4one. To predicate poor perinatal outcome, UA S/D had a sensitivity of85.84%, a specificity of96.22%, a positive predictive value of97.45%, a negative predictive value of80.17%, a positive likelihood ration of0.1472and a negative likelihood ration of22.7120.Conclusions1. NST had a limited value for the prediction of poor perinatal outcome. It can not play a unique role because of its high false-positive rate.2. UA PI, UA RI and S/D had some value for the prediction of poor perinatal outcome. Their diagnostic accuracies were general. The value of PI was higher than S/D and RI.3. BPS had a high value for the prediction of poor perinatal outcome. Its diagnostic accuracy was medium. It can play an important role in the prediction of poor perinatal outcome.4. The false-positive rate was less than10%when BPS<3, and false-negative rate was less than10%when BPS>5. This can provided reference for clinical diagnosis.5. PFOAS had a highest value for the prediction of poor perinatal outcome with a high clinical value.