Materials and Methods
Study design and participants
This study retrospectively collected the clinical data of pregnant woman who received routine pregnancy care and gave birth at the Department of Obstetrics and Gynecology of Peking University First Hospital from January 1, 2021, to June 30, 2021. We included all singleton pregnancies during the period. The exclusion criteria were as follows: fetal major structural abnormalities, chromosomal abnormalities and missing data (Figure 1). This study was approved by the Ethics Committee of Peking University First Hospital (2013-572). Informed consent was obtained from all participants involved in the study.
Protocols and definitions
In this study, demographic data and clinical characteristics of all enrolled subjects were collected, including maternal age, height, weight, body mass index (BMI), gravida, parity, pregnancy method (spontaneous / in vitro fertilization and embryo transfer (IVF-ET)), maternal complications during pregnancy (pregestational diabetes mellitus, gestational diabetes mellitus, gestational hypertension, preeclampsia, chronic hypertension, chronic hypertension with preeclampsia, nephropathy, autoimmune diseases such as lupus erythematosus, rheumatoid arthritis, antiphospholipid syndrome, and Sjogren’s syndrome, thyroid diseases, and placental umbilical cord abnormalities such as single umbilical artery and velamentous placenta), gestational age at delivery, delivery method, and results of ultrasound examinations during pregnancy. Neonatal outcome included neonatal 1-minute and 5-minute Apgar scores, umbilical arterial blood pH value, admission (or not) to the neonatal intensive care unit (NICU) and hospital stays, admission (or not) to the pediatric ward and hospital stays, need for mechanical ventilation and phototherapy, neonatal hypoglycemia, acute respiratory distress syndrome (ARDS), intraventricular hemorrhage (IVH), anemia, necrotizing enterocolitis, retinopathy, sepsis, convulsions, and pneumonia.
The gestational age was dated by the last menstrual period (LMP) or assisted reproduction technology dating. If LMP was unknown or the period was irregular, the first trimester ultrasound would be used. The Hadlock formula III was used to calculate the EFW(15). Three guidelines were used to define FGR: (1) the ACOG criteria for FGR: EFW less than the 10th percentile; (2) The SMFM criteria for FGR: EFW or AC less than the 10th percentile; (3) The ISUOG criteria for FGR: EFW or AC < 3rdpercentile, or EFW or AC < 10th percentile with abnormal blood flow. Doppler blood flow abnormalities in the ISUOG guideline included abnormal uterine artery pulsation index (PI) > 95th percentile, abnormal umbilical artery PI > 95th percentile, and abnormal brain-placental ratio < 5th percentile. The definition of SGA newborn was birth weight lower than the 10th percentile of neonatal growth standards.
A composite ANO included one of these adverse outcomes: neonatal umbilical arterial blood pH<7.1, 5-minute Apgar score<7, acute respiratory distress syndrome (ARDS), intraventricular hemorrhage, neonatal seizures, and admission to the neonatal intensive care unit (NICU).
Statistical analysis
We used SAS 9.4 for the statistical analysis. Kolmogorov-Smimov normality test checks the continuous data for normal distribution. If the data complied with the normal distribution, measurement data are expressed as the mean (standard deviation). Comparisons between groups were performed using the variance analysis. If the data was not consistent with the normal distribution, measurement data were expressed as the median (quartile). Comparisons between groups were performed using the Kruskal-Wallis H test. Categorical data are expressed as the number (percentage). The chi-square test (or Fisher’s exact test) was used for comparisons between groups. P < 0.05 was considered statistically significant.
The sensitivity, specificity, positive predictive value and negative predictive value of ACOG, SMFM and ISUOG for neonatal SGA and ANO were calculated using the fourfold (2X2) contingency table. The discriminatory capacities of three FGR definitions to predict outcome SGA and ANO was compared using the area under receiver-operating-characteristics curves (AUC). Risk ratio (RR) and 95% confidence interval (CI) to predict SGA and ANO were also compared.