Next Article in Journal
Perinatal and Maternal Outcomes According to the Accurate Term Antepartum Ultrasound Estimation of Extreme Fetal Weights
Previous Article in Journal
Neurological Complications in Pregnancy and the Puerperium: Methodology for a Clinical Diagnosis
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
JCM
Volume 12
Issue 8
10.3390/jcm12082993
jcm-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Development and Validation of a Small for Gestational Age Screening Model at 21–24 Weeks Based on the Real-World Clinical Data

by
Jing Gao
1,2,3,†,
Zhongzhou Xiao
4,†,
Chao Chen
1,2,3,†,
Hu-Wei Shi
4,
Sen Yang
4,
Lei Chen
1,2,3,
Jie Xu
4,* and
Weiwei Cheng
1,2,3,*
1
International Peace Maternity and Child Health Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai 200030, China
2
Shanghai Key Laboratory of Embryo Original Disease, Shanghai 200040, China
3
Shanghai Municipal Key Clinical Specialty, Shanghai 200030, China
4
Shanghai Artificial Intelligence Laboratory, Shanghai 200030, China
*
Authors to whom correspondence should be addressed.
†
These authors contributed equally to this work.
J. Clin. Med. 2023, 12(8), 2993; https://doi.org/10.3390/jcm12082993
Submission received: 8 January 2023 / Revised: 26 March 2023 / Accepted: 19 April 2023 / Published: 20 April 2023
(This article belongs to the Section Obstetrics & Gynecology)
Browse Figures
Versions Notes

Abstract

:
Background: Small for gestational age (SGA) is a condition in which fetal birthweight is below the 10th percentile for the gestational age, which increases the risk of perinatal morbidity and mortality. Therefore, early screening for each pregnant woman is of great interest. We aimed to develop an accurate and widely applicable screening model for SGA at 21–24 gestational weeks of singleton pregnancies. Methods: This retrospective observational study included medical records of 23,783 pregnant women who gave birth to singleton infants at a tertiary hospital in Shanghai between 1 January 2018 and 31 December 2019. The obtained data were nonrandomly classified into training (1 January 2018 to 31 December 2018) and validation (1 January 2019 to 31 December 2019) datasets based on the year of data collection. The study variables, including maternal characteristics, laboratory test results, and sonographic parameters at 21–24 weeks of gestation were compared between the two groups. Further, univariate and multivariate logistic regression analyses were performed to identify independent risk factors for SGA. The reduced model was presented as a nomogram. The performance of the nomogram was assessed in terms of its discrimination, calibration, and clinical usefulness. Moreover, its performance was assessed in the preterm subgroup of SGA. Results: Overall, 11,746 and 12,037 cases were included in the training and validation datasets, respectively. The developed SGA nomogram, comprising 12 selected variables, including age, gravidity, parity, body mass index, gestational age, single umbilical artery, abdominal circumference, humerus length, abdominal anteroposterior trunk diameter, umbilical artery systolic/diastolic ratio, transverse trunk diameter, and fasting plasma glucose, was significantly associated with SGA. The area under the curve value of our SGA nomogram model was 0.7, indicating a good identification ability and favorable calibration. Regarding preterm SGA fetuses, the nomogram achieved a satisfactory performance, with an average prediction rate of 86.3%. Conclusions: Our model is a reliable screening tool for SGA at 21–24 gestational weeks, especially for high-risk preterm fetuses. We believe that it will help clinical healthcare staff to arrange more comprehensive prenatal care examinations and, consequently, provide a timely diagnosis, intervention, and delivery.

1. Introduction

The term small for gestational age (SGA) is exclusively used to describe newborns with birth weights below the 10th percentile. Its prevalence is 3–10% and 6.5% worldwide and in China, respectively [1]. Similarly, infants with fetal growth restriction (FGR) fail to achieve their full in utero growth potential and have a pathological condition whose causes can be broadly classified into maternal, fetal, and placental causes [2]. Notably, SGA or FGR can be mostly attributed to suboptimal uterine–placental perfusion and fetal nutrition, increasing the risks of perinatal morbidity and mortality and leading to adverse effects in later life (e.g., cognitive, behavioral, and socioemotional dysfunction in childhood, and metabolism and cardiovascular diseases in adulthood) [3]. Exact evaluation of FGR during antenatal care is difficult; hence, SGA is commonly used as a proxy for screening suspected FGR pregnancies. Indeed, the terms SGA and FGR are often used interchangeably [4]. The traditional method for diagnosing SGA neonates is maternal abdominal palpation; however, its performance is poor as its prediction accuracy is only 21% [5]. Recently, sonographic fetal biometry and Doppler velocimetry evaluation of the umbilical artery (UA) have been reported to improve the diagnostic accuracy of SGA. A Cochrane review published in 2017 reported that UA Doppler imaging in high-risk pregnancies could reduce the rate of perinatal child deaths by 29% [6]; however, this imaging is performed in the third trimester. Early detection will enable improved outcomes in these fetuses as it allows for the adjustment of follow-up and delivery times. Alexandros et al. developed a simple model to predict FGR and SGA by combining the predictors used in the first and second trimesters. In a previous study, the detection rate of 10% false positives was suboptimal for FGR (59.6%) and SGA (41.8%) [7]. Thus, some scientists are striving to achieve higher accuracy using novel predictors, such as serum placental growth factor, soluble fms-like tyrosine kinase-1 (sFlt-1), pregnancy-associated plasma protein-A (PAPP-A), and nuchal fold thickness, which are not routinely assessed in clinical practice [8].
This retrospective large cohort study aimed to establish an accurate widely applicable screening tool for SGA at 21–24 gestational weeks of singleton pregnancies. Accurately identifying fetuses at high risk for developing SGA can aid in their antenatal management and allow for timely interventions to reduce associated morbidity and mortality.

2. Materials and Methods

This study was conducted in accordance with the principles of the Declaration of Helsinki. The requirement for informed consent was waived by the Ethics Committee of the International Peace Maternity and Child Health Hospital owing to the retrospective nature of the study. Moreover, the study was reported in accordance with the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) statement [9]. The official TRIPOD checklist is presented in Supplemental Table S1.

2.1. Study Population

For this retrospective study, we collected data from the digital medical records system of the International Peace Maternity and Child Health Hospital—a tertiary-care hospital in Shanghai—between 1 January 2018 and 31 December 2019. The inclusion criteria for the present study were as follows: (1) a singleton pregnancy; (2) gestation ≥ 28 weeks; and (3) no severe fetal chromosomal or structural abnormalities. After data screening, 30,757 individuals were included in the study (Figure 1).

2.2. Variables Measurements

Using studies and reviews in the relevant literature, we searched for SGA variables that can be easily evaluated in different settings by examiners with different levels of clinical experience and as a part of routine examinations during pregnancy. In China, ultrasound screening of fetal malformations is currently performed at 21–24 gestational weeks. Therefore, we used maternal data (including demographic characteristics, blood glucose levels, and blood lipid levels) and results of sonographic examinations (first ultrasound screening of fetal malformations at 21–24 weeks of gestation) as variables in our model. Further, we collected data regarding the mother’s and father’s demographic characteristics, medical history, and reproductive history at the first antenatal visit during 9–13 weeks of gestation. Moreover, data regarding maternal height, weight, gravidity, parity, and educational level were obtained via face-to-face interviews. The pre-pregnancy body mass index (pre-pregnancy BMI) was calculated by dividing the pre-pregnancy weight (kg) by the square of the pre-pregnancy height (m2). The obtained BMI values were classified into four levels using the cutoff for Asian adults, as proposed by the World Health Organization [10]: <18.5 kg/m2, underweight; 18.5–24.9 kg/m2, normal weight; 25.0–29.9 kg/m2, overweight; and 30 kg/m2, obesity. Gestational age was derived from sonographic measurement of the fetal crown–rump length or biparietal diameter. Maternal fasting lipid serum samples were obtained in the first trimester (9–14 weeks), collected in 10-mL vacutainer tubes, and centrifuged. The following laboratory indices were calculated: triglycerides (TG), total cholesterol, high-density lipoprotein, and low-density lipoprotein. Furthermore, values of blood pressure (systolic blood pressure and diastolic blood pressure) and glucose index (fasting plasma glucose [FPG], one-hour glucose [GLU-1H], two-hour glucose [GLU-2H], and glycosylated hemoglobin [HbA1c]) were recorded on the day of the 75 g oral glucose tolerance test between 24 and 28 weeks of gestation. Moreover, we used the Phillips IU22 ultrasound with a probe frequency of 3.5 MHz. Pregnancy was determined after the woman was placed in a prone position, slightly filling her bladder and controlling her respiratory intensity. The following parameters were measured in accordance with standardized recommendations [11]: biparietal diameter (BPD), head circumference (HC), abdominal circumference (AC), humerus length (HL), transverse trunk diameter (TTD), anteroposterior trunk diameter (APTD), and max amniotic fluid volume (AFV). UA Doppler flow indices included the following: systolic/diastolic ratio (S/D), pulsatility index (PI), and resistance index (RI). Moreover, we evaluated the placental thickness, placental location (normal, placenta previa, or low-lying), placental sinusoids, single umbilical artery (SUA), and velamentous cord insertion placenta. All examiners were senior doctors with standard training and >5 years of experience in obstetric ultrasonography.

2.3. Outcomes

Each neonate’s birthweight (in grams) was routinely measured by registered midwives using an electronic weighing scale within 30 min of delivery. SGA was defined as a birth weight below the 10th percentile as per the Chinese neonatal birth weight curve for different gestational ages [12].

2.4. Data Processing

The data were collected and stored in Microsoft Office Excel® 2019 (Microsoft Corporation, Santa Rosa, CA, USA); subsequently, they were exported to R software version 4.2.1 (R Core Team, Vienna, Austria) for data preprocessing. Regarding obstetric and ultrasonographic characteristics, the proportions of missing data varied and are summarized in Supplemental Table S4. Based on our clinical experience, we assumed that the data were missing at random (MAR). Thus, multivariate imputation by chained equations was performed for these missing values [13]. Subsequently, a tiny proportion (<1%) of the missing values remained. We excluded these values to ensure data integrity. Further, sensitivity analyses were performed to determine whether the imputation values were robust and whether the assumption of MAR was valid. The results provided approximate measurements, which were comparable (Supplemental Materials).

2.5. Statistical Analysis

First, preliminary statistical analyses were performed, including the normality test and the correlation analysis of covariates. In the normality test, we assessed whether the data followed a normal distribution using a combination of the Kolmogorov–Smirnov test and QQ plots. Medians and interquartile ranges (IQRs) were used to present continuous variables, whereas counts and percentages were used to present categorical variables. The Wilcoxon rank sum test was used for comparisons of continuous variables between the groups. The chi-squared or Fisher’s exact test was used for categorical variables, as appropriate. Two-sided p-values of <0.05 were considered significant. All statistical analyses were performed using R version 4.2.1 (23 June 2022).
According to the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) statement, data were nonrandomly divided into a training set (1 January 2018 to 31 December 2018) and a validation set (1 January 2019 to 31 December 2019) in terms of the year of data collection. Notably, this is a better design for evaluating model performance than random division because it allows for evaluating nonrandom variation between the two data sets. First, we conducted a univariate logistic regression analysis to acquire information about the relationship between each potential determinant and SGA. All statistically significant covariates were selected for a subsequent multivariate logistic regression analysis. The multivariate logistic regression analysis with a backward stepwise selection (using Akaike’s information criterion [AIC]) was then used for the training set. Subsequently, multicollinearity was evaluated by assessing the value of the variance inflation factor (VIF). Thereafter, a nomogram was created using the results of the multivariate logistic regression analysis.
The calibration curve was used to assess the calibration ability of the nomogram in addition to the Hosmer–Lemeshow test. Further, we measured the area under the receiver operating characteristic curve (AUC) to quantify the discrimination performance of the nomogram. Bootstrapping validation with 1000 chosen resamples was used to validate AUC and its 95% confidence interval, both internally (training set) and externally (validation set), as well as the calibration curve. The model’s predictive performance after bootstrapping could be applied to other similar populations to some extent. Regarding its usefulness in clinical practice, a decision curve analysis (DCA) was used to investigate the overall benefits in patients with pre-specified threshold probabilities in the validation dataset [14]. Further, we compared the prediction model with two reasonable clinical strategies: intervention for all and intervention for none.
Focusing on the cases of preterm births (gestational age < 37 weeks), a post hoc analysis was performed to compare the differences between cases that were successfully predicted (true positives) and cases that were not predicted (false negatives) using the Wilcoxon test.

3. Result

3.1. Baseline Demographic, Blood Testing, and Ultrasonographic Characteristics of Pregnant Women

This study enrolled 23,783 pregnant women, including 11,746 who gave birth in 2018 (training dataset) and 12,037 who gave birth in 2019 (validation dataset). Figure 1 shows the flow chart of the inclusion of the study population. We compared the differences between the SGA and non-SGA groups within the two datasets and found no statistically significant differences in SGA prevalence between the datasets (4.07% vs. 4.32%; p = 0.486).
The baseline demographic characteristics of pregnant women are presented in Table 1. With a similar distribution, we found an inconsistency in the significant differences in some characteristics, including age, father’s education level, and family history of diabetes or hypertension, and between the training and validation datasets; however, other characteristics did not differ significantly. The median ages of mothers and fathers were similar (31 [IQR: 28–34] and 32 [IQR: 29–35], respectively). Moreover, in both datasets, the education levels of mothers and fathers were comparable, and most of them had a bachelor’s degree. However, the two groups (SGA or non-SGA) significantly differed in terms of BMI, GA, father’s age, gravidity, and parity.
Table 2 shows the medical characteristics of pregnant women, including obstetric and ultrasonographic results. Although we found some inconsistencies, there were no significant differences in clinical features between the two datasets, regardless of the group (SGA or non-SGA), further proving the validity of the split datasets. In contrast, there were significant differences between the two groups in terms of BPD, AC, HC, FL, HL, TTD, APTD, AFV, placental thickness, umbilical artery systolic/diastolic ratio (UA-S/D), UA (PI), UA (RI), FPG, and GLU-2H (p < 0.05).

3.2. Model Development: Univariate and Multivariate Analysis

Among the covariates included, 24 potential risk factors of 38 variables were excluded by univariate analysis of the training dataset based on a p-value of <0.05. Furthermore, the 24 variables were used in the backward multivariate logistic regression analysis, resulting in 12 remaining variables: age, gravidity, parity, BMI, gestational age, SUA, AC, HL, abdominal APTD, UA-S/D, TTD, and FPG. Owing to the strong correlation between TTD and APTD and a strong VIF value, the variable TTD was excluded based on the Delphi method, which involves the consultation of expert-based opinions to make informed decisions [15]. Furthermore, we developed a logistic regression incorporating the above factors. The odds ratio (OR) and its 95% confidence interval for each factor are shown in Table 3. These data indicate that the independent predictors for SGA were advanced maternal age women (OR: 1.23), SUA (OR: 3.25), and UA-S/D (OR: 1.4). The results of the univariate analysis are provided in the Supplemental Materials.

3.3. Model Performance and Validation

A nomogram was created to quantify and illustrate the prediction model based on the above-mentioned 12 predictors to predict the risk of SGA (Figure 2). Each predictor was assigned a score based on the characteristics of each pregnant woman. The total score was then calculated to obtain the risk probability [16]. The prediction nomogram achieved acceptable performance, with an AUC of 0.70 (95% CI: 0.67–0.72) based on the 2018 data extracted from the internal validation dataset. This finding was confirmed to be reliable using a bootstrapping method with binormal smoothing. A perfect concordance was found between observation and prediction (Figure 3). Moreover, the Hosmer–Lemeshow test yielded a nonsignificant result (p = 0.84), suggesting no evidence of a poor fit.
The discriminative capability of the model slightly improved in the external validation dataset (Figure 4), with an AUC of 0.71 (95% CI: 0.68–0.73). Moreover, the calibration curve of the nomogram indicated good agreement between the observed and predicted probability in the external validation dataset, except for a slight departure on the upper tail of the curve (Figure 4). Furthermore, the Hosmer–Lemeshow test provided a nonsignificant result (p = 0.39).

3.4. Model Cost-Benefit Analysis (DCA)

The DCA for the SGA prediction model was performed (Figure 5). The net benefit was calculated for all threshold probabilities, ranging from 0 to 1. The decision curve of our model indicated that the use of the nomogram would be more beneficial than the use of intervention for all and intervention for none scheme to predict SGA if the chosen threshold probability is between 4% and 15%.

3.5. The Performance of the Preterm Subgroup

Among the true SGA fetuses, 33 and 40 in the training and validation datasets were born preterm, respectively. Of them, 27 (81.8%) and 36 (90.0%) were successfully predicted using the model (average prediction rate: 86.3%). Regarding the comparison between the SGA fetuses truly predicted by our model and those not predicted by our model, the mean gestational ages of truly predicted infants were significantly lower in both training (38.73 vs. 39.08, p < 0.05) and validation (38.78 vs. 39.20, p < 0.05) datasets. This indicates that the model has the advantage of accurately predicting SGA at a relatively small gestational age. To better illustrate these differences, box-and-whisker plots are shown in Figure 6.

4. Discussion

SGA is a complex and multifactorial disorder that affects fetal development and often results in stillbirth or other perinatal complications. Early screening to predict the likelihood of FGR in fetuses is expected to help with timely diagnosis through intensive follow-up or detailed examination [17]. Based on the clinical data from a large cohort of pregnant women, our study retrospectively identified the risk factors for SGA and developed a good predictive model advanced to 21–24 weeks of gestation.
Considering the heterogeneous causes of SGA, we used both maternal information and the results of the sonographic examination in utero at 22–25 weeks of gestation as variables in our model. Furthermore, using logistic regression analyses, we identified 12 risk factors: age, gravidity, parity, BMI, gestational age, SUA, AC, HL, abdominal APTD, UA-S/D, TTD, and FPG. The values of these indicators are available in routine medical records, indicating that our model is widely applicable. Various studies have reported predictive models with novel biochemical markers, such as PAPP-A, alfa-fetoprotein, AFP, and human chorionic gonadotropin, which are known to be partly associated with placental function. For example, Sotiriadis et al. [18] combined first- and second-trimester markers, including PAPP-A, for establishing their model. The detection rate for 10% false positives was good for late FGR (78.6%). However, the causes and mechanisms of these novel markers remain unconfirmed [19]. Unless the pregnancies are considered high-risk, the use of these novel markers for all pregnant women may be considered excessive medical care.
The results of the present study indicated that gravida and parity are negatively associated with SGA; this finding is similar to that of previous studies. Li Lin et al. [20] conducted a retrospective study of Chinese individuals and found that multiparity was associated with a reduced risk of LBW (aRR = 0.74, 95% CI: 0.72–0.77) and SGA (aRR = 0.67, 95% CI: 0.66–0.69) compared with nulliparity. Furthermore, a meta-analysis of 41 studies suggested that nulliparous mothers have an 89% increased risk of SGA [21]. This can be attributed to less uteroplacental blood flow and smaller uterine cavities in women who have never given birth or conceived before [22]. Results of a previous study that used uterine artery Doppler velocimetry revealed that the prevalence of uterine artery notches is significantly higher in nulliparous women, suggesting a higher uteroplacental blood impedance to flow [23]. Furthermore, pregnancy at an advanced age (≥35 years) was identified as an independent risk factor for SGA. Another study of 137,791 women reported that the risk of SGA increased with maternal age and that the risk increased earlier in nulliparous women [24]. Interestingly, in the subgroup analysis by Palatnik et al. [25], nulliparous women aged ≥30 years (but not multiparous women) and all women aged ≥40 years had a high risk of developing SGA. Delayed childbearing has been a growing trend in many countries over the last few decades [26], increasing the number of nulliparous women at an advanced age. Therefore, monitoring and treating primiparous women aged >35 years is a challenge for all involved in the fight to minimize the consequences of SGA fetuses.
To the best of our knowledge, there is no effective treatment available to reverse the course of FGR, except for delivery [27]. According to the 2020 Society for Maternal-Fetal Medicine Consult Series, delivery of an FGR fetus is recommended after 37 weeks of gestation based on the percentile of the estimated fetal weight [28]. Moreover, earlier delivery is indicated in cases of absent or reverse UA flow owing to the high impedance uteroplacental perfusion [29]. Furthermore, poor maternal conditions, such as hypertensive disease of pregnancy, cardiac disease, and immune system disease, could result in preterm birth, indicating that the earlier the delivery, the worse the intrauterine environment and/or maternal condition. Our model achieved a relatively acceptable accuracy of SGA prediction and demonstrated good performance in identifying SGA pregnancy with a high risk of preterm birth.
Our nomogram model can serve as a practical tool for the clinical screening of SGA. Once it indicates the risk of SGA in pregnant women, it suggests that the pregnant woman needs more attention. Notably, it has been recommended to shorten the visit intervals. In addition, finding possible etiologies, such as maternal malnutrition, poor weight gain, smoking, congenital fetal infections, and fetal genetic or structural disorders, should be the first goal. However, in most cases, doctors often fail to identify the etiology. Doppler sonography plays a significant role in assessing fetal developmental conditions and trends [30]. Currently, the middle cerebral artery and ductus venosus Doppler, except for UA Doppler, represent chronic fetal hypoxia, thereby helping to determine the timing of delivery. Cardiac and aortic isthmus Doppler allows the evaluation of heart functionality with the deterioration of FGR [31]. MacDonald et al. [32] reported a novel predictor, the ratio of the middle cerebral artery to the UA pulsatility indices (CPUR), which indicated the strongest association with indicators of placental insufficiency. However, Rial Crestelo et al. [33] indicated that the added value of CPUR at 33 weeks of gestation for detecting defective growth is poor in nonselected pregnancies.
Our study has some limitations. Although the overall sample size was considerable, all participants were recruited from the same obstetric hospital, principally covering low to moderate risk pregnant women. This cohort may not completely represent obstetric practice in the community. Second, the accuracy of our model was not optimal. One of the reasons for this could be the limited number of variables. The policy at our institution is not to reveal the fetal sex before delivery; hence, we did not select fetal sex as one of the variables. Third, in our retrospective study, we could not collect relevant information regarding gestational hypertension in the mid-trimester. Some studies reported that 30–40% of cases of FGR are complicated by placental function impairment due to pre-eclampsia, chronic hypertension, and gestational hypertension [34]. The predictive power of our model can be improved if we collect more data regarding hypertension changes, drug treatment, complications, and other variables. Further studies are warranted to include more appropriate influencing factors and use prospective design to optimize our model.
Our SGA model focused on the issue within the medical framework of Prediction, Prevention, Personalization, and Precision, and it highlighted the significance of timely detection of high-risk pregnancies with high accuracy. In clinical practice, once the model indicates the possibility of SGA, obstetricians are advised to search for possible etiologies, shorten visit intervals, and seek detailed investigations (for example, blood pressure, urine protein, and ultrasound Doppler examination). This will guide clinical healthcare workers to arrange for more prenatal care examinations and, thus, enable timely diagnosis, intervention, and delivery.

5. Conclusions

Our model represents a reliable screening tool for SGA at 21–24 weeks gestation, particularly for fetuses at high risk of preterm delivery. In fact, our model has a wide application for improving pregnancy outcomes and expectations for newborns and optimizing medical cost-efficiency.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/jcm12082993/s1.

Author Contributions

Conceptualization, J.X. and W.C.; data curation, S.Y. and L.C.; formal analysis, Z.X.; funding acquisition, W.C.; investigation, W.C., J.X., J.G., Z.X. and S.Y.; methodology, C.C.; project administration, H.-W.S.; statistical analysis, H.-W.S.; resources, L.C.; supervision, J.X.; validation, C.C.; writing—original draft, J.G. and Z.X.; writing—review and editing, J.X. and W.C. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the scientific research project of Shanghai Municipal Health Commission (Project NO. 202140091) and Shanghai Commission of Science and Technology (Project NO. 20DZ230500). The funding was not involved in the study design, data collection, analysis, interpretation of data, writing of the report, or the decision to submit the article for publication.

Institutional Review Board Statement

Our research strictly adheres to the principles of the Declaration of Helsinki. Ethical approval for the use of patient information was obtained from the Research Ethics Committee of the International Peace Maternity and Child Health Hospital (GKLW2021–20).

Informed Consent Statement

The informed consent was waived by the Ethics Committee of the International Peace Maternity and Child Health Hospital due to the retrospective nature of the study.

Data Availability Statement

Further data inquiries can be directed to the corresponding author.

Acknowledgments

We are grateful for the materials provided by the pregnant women and we thank all the clinical staff who participated in the study.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Francis, A.; Hugh, O.; Gardosi, J. Customized vs INTERGROWTH-21st standards for the assessment of birthweight and stillbirth risk at term. Am. J. Obstet. Gynecol. 2018, 218, S692–S699. [Google Scholar] [CrossRef] [PubMed]
  2. American College of Obstetricians and Gynecologists’ Committee on Practice Bulletins. Obstetrics and the Society for Maternal-Fetal Medicin ACOG Practice Bulletin No. 204, ACOG Practice Bulletin No. 204: Fetal growth restriction. Obstet. Gynecol. 2019, 133, e97–e109. [Google Scholar] [CrossRef] [PubMed]
  3. Melamed, N.; Baschat, A.; Yinon, Y.; Athanasiadis, A.; Mecacci, F.; Figueras, F.; Berghella, V.; Nazareth, A.; Tahlak, M.; McIntyre, H.D.; et al. FIGO (International Federation of Gynecology and Obstetrics) initiative on fetal growth: Best practice advice for screening, diagnosis, and management of fetal growth restriction. Int. J. Gynecol. Obstet. 2021, 152 (Suppl. 1), 3–57. [Google Scholar] [CrossRef] [PubMed]
  4. Nawathe, A.; Lees, C. Early onset fetal growth restriction. Best Pr. Res. Clin. Obstet. Gynaecol. 2016, 38, 24–37. [Google Scholar] [CrossRef]
  5. Pay, A.; Frøen, J.F.; Staff, A.C.; Jacobson, B.; Gjessing, H.K. Prediction of small-for-gestational-age status by symphysis-fundus height: A registry-based population cohort study. BJOG Int. J. Obstet. Gynaecol. 2016, 123, 1167–1173. [Google Scholar] [CrossRef]
  6. Bais, J.M.; Eskes, M.; Pel, M.; Bonsel, G.J.; Bleker, O.P. Effectiveness of detection of intrauterine growth retardation by abdominal palpation as screening test in a low risk population: An observational study. Eur. J. Obstet. Gynecol. Reprod. Biol. 2004, 116, 164–169. [Google Scholar] [CrossRef]
  7. Alfirevic, Z.; Stampalija, T.; Dowswell, T. Fetal and umbilical Doppler ultrasound in high-risk pregnancies. Cochrane Database Syst. Rev. 2017, 6, CD007529. [Google Scholar] [CrossRef]
  8. Sotiriadis, A.; Figueras, F.; Eleftheriades, M.; Papaioannou, G.K.; Chorozoglou, G.; Dinas, K.; Papantoniou, N. First-trimester and combined first- and second-trimester prediction of small-for-gestational age and late fetal growth restriction. Ultrasound Obstet. Gynecol. 2019, 53, 55–61. [Google Scholar] [CrossRef]
  9. Poon, L.C.; Lesmes, C.; Gallo, D.M.; Akolekar, R.; Nicolaides, K.H. Prediction of small-for-gestational-age neonates: Screening by biophysical and biochemical markers at 19–24 weeks. Ultrasound Obstet. Gynecol. 2015, 46, 437–445. [Google Scholar] [CrossRef]
  10. Ciobanu, A.; Rouvali, A.; Syngelaki, A.; Akolekar, R.; Nicolaides, K.H. Prediction of small for gestational age neonates: Screening by maternal factors, fetal biometry, and biomarkers at 35–37 weeks’ gestation. Am. J. Obstet. Gynecol. 2019, 220, 486.e1–486.e11. [Google Scholar] [CrossRef]
  11. Collins, G.S.; Reitsma, J.B.; Altman, D.G.; Moons, K.G.M. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): The TRIPOD statement. BMJ 2014, 350, g7594. [Google Scholar] [CrossRef]
  12. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004, 363, 157–163. [Google Scholar] [CrossRef]
  13. Salomon, L.J.; Alfirevic, Z.; Berghella, V.; Bilardo, C.; Hernandez-Andrade, E.; Johnsen, S.L.; Kalache, K.; Leung, K.; Malinger, G.; Munoz, H.; et al. Practice guidelines for performance of the routine mid-trimester fetal ultrasound scan. Ultrasound Obstet. Gynecol. 2010, 37, 116–126. [Google Scholar] [CrossRef]
  14. Zhu, L.; Zhang, R.; Zhang, S.; Shi, W.; Yan, W.; Wang, X.; Lyu, Q.; Liu, L.; Zhou, Q.; Qiu, Q.; et al. Chinese neonatal birth weight curve for different gestational age. Chin. J. Pediatr. 2015, 53, 97–103. [Google Scholar]
  15. Sterne, J.A.C.; White, I.R.; Carlin, J.B.; Spratt, M.; Royston, P.; Kenward, M.G.; Wood, A.M.; Carpenter, J.R. Multiple imputation for missing data in epidemiological and clinical research: Potential and pitfalls. BMJ 2009, 338, b2393. [Google Scholar] [CrossRef]
  16. Vickers, A.J.; van Calster, B.; Steyerberg, E.W. A simple, step-by-step guide to interpreting decision curve analysis. Diagn. Progn. Res. 2019, 3, 18. [Google Scholar] [CrossRef]
  17. Niederberger, M.; Spranger, J. Delphi Technique in Health Sciences: A Map. Front. Public Health 2020, 8, 457. [Google Scholar] [CrossRef]
  18. Park, S.Y. Nomogram: An analogue tool to deliver digital knowledge. J. Thorac. Cardiovasc. Surg. 2018, 155, 1793. [Google Scholar] [CrossRef]
  19. Schlembach, D. Fetal growth restriction—Diagnostic work-up, management and delivery. Geburtshilfe Frauenheilkd 2020, 80, 1016–1025. [Google Scholar] [CrossRef]
  20. Christians, J.K.; Lennie, K.I.; Munoz, M.F.H.; Binning, N. PAPP-A2 deficiency does not exacerbate the phenotype of a mouse model of intrauterine growth restriction. Reprod. Biol. Endocrinol. 2018, 16, 58. [Google Scholar] [CrossRef]
  21. Lin, L.; Lu, C.; Chen, W.; Li, C.; Guo, V.Y. Parity and the risks of adverse birth outcomes: A retrospective study among Chinese. BMC Pregnancy Childbirth 2021, 21, 257. [Google Scholar] [CrossRef] [PubMed]
  22. Shah, P.S.; Knowledge Synthesis Group on Determinants of LBW/PT Births. Parity and low birth weight and preterm birth: A systematic review and meta-analyses. Acta Obstet. Gynecol. Scand. 2010, 89, 862–875. [Google Scholar] [CrossRef] [PubMed]
  23. Canteiro, R.; Bahamondes, M.V.; Fernandes, A.D.S.; Espejo-Arce, X.; Marchi, N.M.; Bahamondes, L. Length of the endometrial cavity as measured by uterine sounding and ultrasonography in women of different parities. Contraception 2010, 81, 515–519. [Google Scholar] [CrossRef] [PubMed]
  24. Prefumo, F.; Bhide, A.; Sairam, S.; Penna, L.; Hollis, B.; Thilaganathan, B. Effect of parity on second-trimester uterine artery Doppler flow velocity and waveforms. Ultrasound Obstet. Gynecol. 2003, 23, 46–49. [Google Scholar] [CrossRef] [PubMed]
  25. Desplanches, T.; Bouit, C.; Cottenet, J.; Szczepanski, E.; Quantin, C.; Fauque, P.; Sagot, P. Combined effects of increasing maternal age and nulliparity on hypertensive disorders of pregnancy and small for gestational age. Pregnancy Hypertens. 2019, 18, 112–116. [Google Scholar] [CrossRef]
  26. Palatnik, A.; De Cicco, S.; Zhang, L.; Simpson, P.; Hibbard, J.; Egede, L.E. The Association between Advanced Maternal Age and Diagnosis of Small for Gestational Age. Am. J. Perinatol. 2019, 37, 37–43. [Google Scholar] [CrossRef]
  27. Li, H.; Fan, C.; Mubarik, S.; Nabi, G.; Ping, Y.X. The trend in delayed childbearing and its potential consequences on pregnancy outcomes: A single center 9-years retrospective cohort study in Hubei, China. BMC Pregnancy Childbirth 2022, 22, 514. [Google Scholar] [CrossRef]
  28. Behboudi-Gandevani, S.; Ziaei, S.; Farahani, F.K.; Jasper, M. The perspectives of Iranian women on delayed childbearing: A qualitative study. J. Nurs. Res. 2015, 23, 313–321. [Google Scholar] [CrossRef]
  29. Martins, J.G.; Biggio, J.R.; Abuhamad, A. Diagnosis and management of fetal growth restriction: (Replaces Clinical Guideline Number 3, April 2012), Society For Maternal Fetal Medicine Consult Series. No. 52. Am. J. Obstet. Gynecol. 2020, 223, B2–B17. [Google Scholar] [CrossRef]
  30. He, J.; Chen, Z.; Wen, T.; Xu, L.; Chen, C.; Liu, P. Utility of placental diffusion-weighted magnetic resonance imaging in prenatal diagnosis of small for gestational age infants and pregnancy outcome prediction. Placenta 2022, 121, 91–98. [Google Scholar] [CrossRef]
  31. Baschat, A.A.; Weiner, C.P. Umbilical artery Doppler screening for detection of the small fetus in need of antepartum surveillance. Am. J. Obstet. Gynecol. 2000, 182 Pt. 1, 154–158. [Google Scholar] [CrossRef]
  32. Macdonald, T.M.; Hui, L.; Robinson, A.J.; Dane, K.M.; Middleton, A.L.; Tong, S.; Walker, S.P. Cerebral–placental–uterine ratio as novel predictor of late fetal growth restriction: Prospective cohort study. Ultrasound Obstet. Gynecol. 2019, 54, 367–375. [Google Scholar] [CrossRef]
  33. Rial-Crestelo, M.; Portilla, R.J.M.; Cancemi, A.; Caradeux, J.; Fernandez, L.; Peguero, A.; Gratacos, E.; Figueras, F. Added value of cerebro-placental ratio and uterine artery Doppler at routine third trimester screening as a predictor of SGA and FGR in non-selected pregnancies. J. Matern. Neonatal Med. 2018, 32, 2554–2560. [Google Scholar] [CrossRef]
  34. Villar, J.; Altman, D.G.; Purwar, M.; Noble, J.A.; Knight, H.E.; Ruyan, P.; Ismail, L.C.; Barros, F.C.; Lambert, A.; Papageorghiou, A.T.; et al. The objectives, design and implementation of the INTERGROWTH-21st Project. BJOG Int. J. Obstet. Gynaecol. 2013, 120 (Suppl. 2), 9–26. [Google Scholar] [CrossRef]
Jcm 12 02993 g001 550
Figure 1. Flowchart illustrating pregnant women included in the study.
Figure 1. Flowchart illustrating pregnant women included in the study.
Jcm 12 02993 g001
Jcm 12 02993 g002 550
Figure 2. Nomogram model for predicting the risk of small for gestational age in the mid-trimester pregnancy. BMI, body mass index; GA, gestational age; SUA, single umbilical artery; AC, abdominal circumference; humeral length APTD, anteroposterior trunk diameter; UA-SD, Umbilical artery systolic/diastolic ratio; FBG, fasting blood glucose; SGA, small for gestational age.
Figure 2. Nomogram model for predicting the risk of small for gestational age in the mid-trimester pregnancy. BMI, body mass index; GA, gestational age; SUA, single umbilical artery; AC, abdominal circumference; humeral length APTD, anteroposterior trunk diameter; UA-SD, Umbilical artery systolic/diastolic ratio; FBG, fasting blood glucose; SGA, small for gestational age.
Jcm 12 02993 g002
Jcm 12 02993 g003 550
Figure 3. Receiver operating characteristic and calibration curves in the training dataset. ROC, receiver operating characteristic; SGA, small for gestational age.
Figure 3. Receiver operating characteristic and calibration curves in the training dataset. ROC, receiver operating characteristic; SGA, small for gestational age.
Jcm 12 02993 g003
Jcm 12 02993 g004 550
Figure 4. Receiver operating characteristic and calibration curves in the validation dataset. ROC, receiver operating characteristic; SGA, small for gestational age.
Figure 4. Receiver operating characteristic and calibration curves in the validation dataset. ROC, receiver operating characteristic; SGA, small for gestational age.
Jcm 12 02993 g004
Jcm 12 02993 g005 550
Figure 5. Decision curve analysis for the small for gestational age (SGA) prediction model.
Figure 5. Decision curve analysis for the small for gestational age (SGA) prediction model.
Jcm 12 02993 g005
Jcm 12 02993 g006 550
Figure 6. Box-and-whisker plots of gestational age between small for gestational age fetuses truly predicted by our model and not predicted by our model.
Figure 6. Box-and-whisker plots of gestational age between small for gestational age fetuses truly predicted by our model and not predicted by our model.
Jcm 12 02993 g006
Table
Table 1. Baseline characteristics.
Table 1. Baseline characteristics.
Training SetValidation Set
CharacteristicNon-SGA 1
(n = 11,259)		SGA 1
(n = 487)		p-Value 2Non-SGA 1
(n = 11,516)		SGA 1
(n = 521)		p-Value 2
BMI 20.9 (19.5, 22.9) 20.0 (18.7, 21.6) <0.001 20.9 (19.5, 22.6) 20.2 (18.7, 21.5) <0.001
GA 39.1 (38.4, 40.0) 39.1 (38.1, 39.9) 0.002 39.1 (38.6, 40.0) 39.0 (38.3, 39.9) 0.010
SBP 110 (103, 118) 110 (102, 119) 0.8 110 (103, 118) 109 (103, 116) 0.14
DBP 68 (63, 74) 69 (63, 75) 0.5 69 (63, 75) 68 (63, 74) 0.4
Age 0.046 0.3
<35 8820 (78%) 400 (82%) 9113 (79%) 422 (81%)
≥35 2439 (22%) 87 (18%) 2403 (21%) 99 (19%)
Hus age 0.009 0.018
<35 7788 (69%) 364 (75%) 8386 (73%) 404 (78%)
≥35 3471 (31%) 123 (25%) 3130 (27%) 117 (22%)
Edu >0.9 0.6
Bachelor 5865 (52%) 256 (53%) 5708 (50%) 252 (48%)
Above 2264 (20%) 101 (21%) 2206 (19%) 99 (19%)
Below 2778 (25%) 114 (23%) 2546 (22%) 113 (22%)
Unknown 352 (3.1%) 16 (3.3%) 1056 (9.2%) 57 (11%)
Hus edu 0.6 0.030
Bachelor 5648 (50%) 255 (52%) 5427 (47%) 248 (48%)
Above 2519 (22%) 111 (23%) 2433 (21%) 98 (19%)
Below 2405 (21%) 91 (19%) 2280 (20%) 92 (18%)
Unknown 687 (6.1%) 30 (6.2%) 1376 (12%) 83 (16%)
Gravidity <0.001 <0.001
1 5202 (46%) 296 (61%) 5655 (49%) 318 (61%)
2 3384 (30%) 120 (25%) 3392 (29%) 133 (26%)
≥3 2673 (24%) 71 (15%) 2469 (21%) 70 (13%)
Parity <0.001 <0.001
1 7584 (67%) 390 (80%) 8095 (70%) 432 (83%)
≥2 3675 (33%) 97 (20%) 3421 (30%) 89 (17%)
Smoking 0.6 0.4
No 10,831 (96%) 471 (97%) 10,371 (90%) 460 (88%)
Yes 54 (0.5%) 3 (0.6%) 81 (0.7%) 4 (0.8%)
Unknown 374 (3.3%) 13 (2.7%) 1064 (9.2%) 57 (11%)
Alcohol 0.4 0.4
No 10,700 (95%) 469 (96%) 9765 (85%) 432 (83%)
Yes 185 (1.6%) 5 (1.0%) 687 (6.0%) 32 (6.1%)
Unknown 374 (3.3%) 13 (2.7%) 1064 (9.2%) 57 (11%)
Family history of diabetes or hypertension 0.8 0.018
No 8583 (76%) 377 (77%) 7603 (66%) 313 (60%)
Yes 2294 (20%) 94 (19%) 2317 (20%) 120 (23%)
Unknown 382 (3.4%) 16 (3.3%) 1596 (14%) 88 (17%)
Conception >0.9 0.2
Natural 10,175 (90%) 441 (91%) 9705 (84%) 434 (83%)
ART 789 (7.0%) 34 (7.0%) 841 (7.3%) 32 (6.1%)
Unknown 295 (2.6%) 12 (2.5%) 970 (8.4%) 55 (11%)
BMI, body mass index; GA, gestational age; SBP, systolic blood pressure; DBP, diastolic blood pressure; Hus age, husband’s age; Edu, education; Hus edu, husband’s education; SGA, small for gestational age. 1 Median (IQR); n (%). 2 Wilcoxon rank sum test; Pearson’s chi-squared test; Fisher’s exact probability test.
Table
Table 2. Medical characteristics.
Table 2. Medical characteristics.
Training SetValidation Set
CharacteristicNon-SGA 1
(n = 11,259)		SGA 1
(n = 487)		p-Value 2Non-SGA 1
(n = 11,516)		SGA 1
(n = 521)		p-Value 2
Placental location 0.3 0.7
Normal10,831 (96%)475 (98%) 11,035 (96%)503 (97%)
Low-lying192 (1.7%)5 (1.0%) 197 (1.7%)8 (1.5%)
Previa236 (2.1%)7 (1.4%) 284 (2.5%)10 (1.9%)
Velamentous Placenta >0.9 0.5
No11,226 (99%)486 (99%) 11,472 (100%)518 (99%)
Yes33 (0.3%)1 (0.2%) 44 (0.4%)3 (0.6%)
Placental sinusoids 0.7 0.4
No11,205 (99%)486 (99%) 11,443 (99%)516 (99%)
Yes54 (0.5%)1 (0.2%) 73 (0.6%)5 (1.0%)
SUA 0.046 0.2
No11,230 (99%)483 (99%) 11,479 (100%)518 (99%)
Yes29 (0.3%)4 (0.8%) 37 (0.3%)3 (0.6%)
BPD56.0 (54.0, 58.0)55.0 (53.0, 58.0)<0.00156.0 (54.0, 58.0)55.0 (53.0, 57.0)<0.001
AC175 (169, 183)171 (164, 178)<0.001174 (167, 180)169 (163, 175)<0.001
HC199 (193, 206)195 (189, 203)<0.001198 (193, 204)195 (189, 201)<0.001
FL38.0 (37.0, 40.0)38.0 (36.0, 39.0)<0.00138.0 (37.0, 40.0)38.0 (36.0, 39.0)<0.001
HL36.0 (35.0, 38.0)36.0 (34.0, 37.0)<0.00136.0 (35.0, 38.0)35.0 (34.0, 37.0)<0.001
TTD55.0 (53.0, 58.0)54.0 (51.0, 56.0)<0.00155.0 (52.0, 57.0)53.0 (51.0, 56.0)<0.001
APTD 57.0(54.0, 60.0)55.0 (53.0, 58.0)<0.00156.0 (54.0, 59.0)55.0 (52.0, 57.0)<0.001
AFV 49 (43.0, 55.0)48 (42.0, 53.0)0.00250 (45.0, 55.0)48 (43.0, 55.0)0.010
Pl thickness 26.0 (23.0, 28.0)25.0 (23.0, 28.0)0.01226.0 (23.0, 28.0)25.0 (23.0, 28.0)0.002
UA-S/D 3.20 (2.85, 3.60)3.30 (2.91, 3.80)<0.0013.19 (2.81, 3.60)3.29 (2.91, 3.70)<0.001
UA-PI 1.13 (1.02, 1.24)1.17 (1.05, 1.28)<0.0011.12 (1.01, 1.24)1.16 (1.04, 1.28)<0.001
UA-RI 0.69 (0.65, 0.72)0.70 (0.66, 0.74)<0.0010.68 (0.65, 0.72)0.69 (0.66, 0.73)<0.001
HBA1C 4.90 (4.80, 5.10)4.90 (4.80, 5.10)0.155.00 (4.80, 5.20)4.90 (4.80, 5.10)<0.001
FPG 4.20 (3.95, 4.47)4.12 (3.87, 4.37)<0.0014.21 (3.97, 4.47)4.11 (3.88, 4.34)<0.001
GLU-1H7.66 (6.77, 8.80)7.50 (6.65, 8.54)0.0087.68 (6.62, 8.81)7.62 (6.58, 8.77)0.3
GLU-2H6.46 (5.64, 7.41)6.27 (5.49, 7.28)0.0146.47 (5.67, 7.43)6.32 (5.44, 7.55)0.043
TC4.47 (4.01, 4.96)4.39 (3.95, 4.88)0.0544.43 (4.01, 4.91)4.46 (4.06, 4.95)0.3
TG1.27 (1.01, 1.63)1.19 (0.96, 1.52)<0.0011.30 (1.03, 1.66)1.27 (1.01, 1.59)0.053
HDL1.91 (1.65, 2.17)1.93 (1.67, 2.17)0.41.96 (1.70, 2.25)2.02 (1.73, 2.31)0.014
LDL2.59 (2.16, 3.05)2.53 (2.12, 2.98)0.0702.45 (2.08, 2.86)2.44 (2.10, 2.90)0.4
SUA, single umbilical artery; BPD, biparietal diameter; AC, abdominal circumference; HC, head circumference; FL, fetal length; HL, humeral length; TTD, transverse trunk diameter; APTD, anteroposterior trunk diameter; AFV; Pl thickness, placenta thickness; UA-S/D, umbilical artery systolic/diastolic ratio; UA-PI, umbilical artery pulsatility index; UA-RI, umbilical artery resistance index; HBA1C, glycosylated hemoglobin; FBG, fasting blood glucose; GLU-1H, one-hour glucose; GLU-2H, 2 h glucose; TC, total cholesterol; TG, triglycerides; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SGA, small for gestational age. 1 Median (IQR); n (%). 2 Wilcoxon rank sum test; Pearson’s chi-squared test; Fisher’s exact probability test.
Table
Table 3. Factors associated with SGA in women at the International Peace Maternity and Child Health Hospital (n = 11,746).
Table 3. Factors associated with SGA in women at the International Peace Maternity and Child Health Hospital (n = 11,746).
CharacteristicOR 195% CI 2p-Value
Age
<35 ——
≥351.23(0.94, 1.59)0.12
Gravidity
1——
20.730.56, 0.930.014
≥30.620.43, 0.860.006
Parity
1——
≥20.690.51, 0.950.020
BMI0.890.85, 0.92<0.001
GA0.860.82, 0.91<0.001
SUA
No——
Yes3.250.94, 8.610.033
AC0.980.96, 0.990.012
HL0.920.87, 0.980.004
APTD0.980.93, 1.030.4
UA-S/D1.401.20, 1.64<0.001
FPG0.700.55, 0.880.003
BMI, body mass index; GA, gestational age; SUA, single umbilical artery; AC, abdominal circumference; HL, humeral length; APTD, anteroposterior trunk diameter; UA-S/D, umbilical artery systolic/diastolic ratio; FBG, fasting blood glucose. 1 OR Odds ratio. 2 CI Confidence interval.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Gao, J.; Xiao, Z.; Chen, C.; Shi, H.-W.; Yang, S.; Chen, L.; Xu, J.; Cheng, W. Development and Validation of a Small for Gestational Age Screening Model at 21–24 Weeks Based on the Real-World Clinical Data. J. Clin. Med. 2023, 12, 2993. https://doi.org/10.3390/jcm12082993

AMA Style

Gao J, Xiao Z, Chen C, Shi H-W, Yang S, Chen L, Xu J, Cheng W. Development and Validation of a Small for Gestational Age Screening Model at 21–24 Weeks Based on the Real-World Clinical Data. Journal of Clinical Medicine. 2023; 12(8):2993. https://doi.org/10.3390/jcm12082993

Chicago/Turabian Style

Gao, Jing, Zhongzhou Xiao, Chao Chen, Hu-Wei Shi, Sen Yang, Lei Chen, Jie Xu, and Weiwei Cheng. 2023. "Development and Validation of a Small for Gestational Age Screening Model at 21–24 Weeks Based on the Real-World Clinical Data" Journal of Clinical Medicine 12, no. 8: 2993. https://doi.org/10.3390/jcm12082993

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop