Logo Medical Science Monitor

Call: +1.631.470.9640
Mon - Fri 10:00 am - 02:00 pm EST

Contact Us

Logo Medical Science Monitor Logo Medical Science Monitor Logo Medical Science Monitor

21 December 2024: Clinical Research  

Evaluation of Pregnancy Risks in Women with Subchorionic Hematoma Using Machine Learning Models

Lan Wang ORCID logo123BCDEF, Aiping Qin ORCID logo2A*, Yihua Yang2C, Yufu Jin2B, Qiuyan Huang2B, Xinyue Huang2B, Yu Feng2B, Ting Liang4B

DOI: 10.12659/MSM.945472

Med Sci Monit 2024; 30:e945472

0 Comments

Abstract

0:00

BACKGROUND: Subchorionic hematoma (SCH) can lead to blood accumulation and potentially affect pregnancy outcomes. Despite being a relatively common finding in early pregnancy, the effects of SCH on pregnancy outcomes such as miscarriage, stillbirth, and preterm birth remain debated. This study aims to address these gaps by systematically evaluating the influence of SCH-related clinical factors on pregnancy outcomes using robust analytical techniques.

MATERIAL AND METHODS: Data from SCH and non-SCH pregnant women were collected and split into training and test datasets. Machine learning classifiers and regression models were used to assess the impact of clinical indices on outcomes such as delivery type, NICU transfer, gestational age, and birth weight. Results were evaluated using ROC and calibration plots.

RESULTS: (1) SCH women had a significantly higher risk of stillbirth or miscarriage than non-SCH women (P<0.001). Logistic regression and XGB models showed AUCs of 0.858 and 0.916, respectively. Key factors affecting delivery outcomes included the first positive HCG level, hematoma duration, CA125 level, gestational sac diameter, fibrinogen level, and spouse age. (2) 12.7% of successfully delivered SCH newborns required NICU transfer, but clinical indices did not predict NICU need (AUC 0.589 and 0.629). (3) Successfully delivered SCH women had longer gestational ages than those with miscarriage/stillbirth (38.8 vs 10.1 weeks), but indices did not predict preterm/full-term birth (AUCs 0.449 and 0.503). (4) Birth weight was significantly affected by live birth times and gestational age (P<0.05), though the adjusted R-square was 0.226.

CONCLUSIONS: (1) SCH increases miscarriage or stillbirth risk. (2) the first positive HCG level, the hematoma duration, serum CA125 level, the gestational sac maximum diameter, fibrinogen, and the spouse age highly impacted the delivery outcome. (3) SCH indices do not affect NICU transfer or birth weight. (4) Miscarriage/stillbirth mainly occurs in the first trimester; passing this stage often leads to successful delivery. (5) The birth weight of full-term newborns is significantly higher than that of preterm infants. The clinical indices of SCH pregnant women have no impact on the birth weight of the newborn.

Keywords: Hematoma, Pregnancy, Chorionic Gonadotropin, beta Subunit, Human

Introduction

Subchorionic hematoma (SCH) refers to the bleeding between the chorionic plate and the bottom decidua, causing blood to accumulate between the chorion and the bottom decidua. Under the membrane, it is mostly crescent-shaped, and its lower edge is mostly connected with the internal cervix [1,2].

The most critical issue after SCH is whether it influence pregnancy outcome. This is also the focus of the current study. Some scholars got some results. yet, whether SCH affects pregnancy is controversial or contradictory. Based on the controversy, we believe that either some studies lack sufficient samples, or some studies ignore the effects of confounding factors on pregnancy outcomes with SCH. Also, most studies do not use advanced statistical methods, such as logistic regression, and classification methods (Gradient boost) that are often used in machine learning.

Machine learning (ML) is used in medical research to analyze complex datasets, predict disease outcomes, and personalize medicine. ML models improve prediction and diagnosis accuracy by processing clinical, genomic, and imaging data to assess risks for heart disease, diabetes, and cancer [3]. In personalized medicine, ML helps tailor treatments, especially in cancer therapy, by analyzing genetic and clinical data [4]. ML also supports real-time clinical decisions, aiding doctors, enhancing efficiency, and reducing errors [5].

The current study collected clinical data (indices) of SCH pregnant women and uses advanced statistical methods and classification methods in ML to analyze whether clinical indices affect the pregnancy outcomes of SCH pregnant women and which of them play the import roles. ML is applied in this study due to its ability to manage large datasets and reveal patterns, enhancing research accuracy and reliability.

Material and Methods

STUDY DESIGN AND APPROVAL:

This article is a retrospective study that was conducted after obtaining approval from the FIRST AFFILIATED HOSPITAL of GUANGXI MEDICAL UNIVERSITY ETHICAL REVIEW COMMITTEE (Approval Number: 2024-E331-01).

TIME FRAME:

Clinical data were collected from December 1, 2016, to February 28, 2019.

SUBJECTS:

The study involved pregnant women diagnosed with subchorionic hematoma (SCH) who visited the outpatient and emergency departments of the First Affiliated Hospital of Guangxi Medical University, located in the Guangxi Zhuang Autonomous Region, China. A total of 302 pregnant women with SCH were initially considered. Due to incomplete data or information, 11 cases were not followed up, and 291 cases were finally included. Women who were diagnosed with single-pregnancy in our hospital and took routine birth tests were selected as the control group.

INCLUSION CRITERIA:

(1) B-ultrasound examination shows intrauterine gestational sac and primitive fetal heartbeat and SCH is found (diagnostic criteria: B-ultrasound during pregnancy shows that there is an echoless zone between the uterine wall and the gestational sac (or fetal membrane), the shape is usually a typical crescent, triangle, polygon, and irregular shape, the most common in the placenta from the lower edge of the cervix to the inside of the chorion, mostly crescent-shaped pregnant women); (2) without severe cardiovascular, liver, kidney and infectious diseases; (3) without conscious disorder; (4) without mental illness.

EXCLUSION CRITERIA:

pregnant women with (1) severe medical diseases and endocrine dysfunction; (2) twins and multiple births; (3) vaginal bleeding or threatened abortion during early pregnancy; (4) abortion of chromosomal abnormalities or termination of pregnancy for personal reasons, uterus Congenital malformations, endometrial polyps, submucosal uterine fibroids, cervical insufficiency [6].

CLINICAL INDICES COLLECTING AND RELEVANCE EXPLANATION TO SCH AND PREGNANCY OUTCOMES:

Collecting clinical indices of SCH pregnant women at SCH diagnosed, including: (1) age (Advanced maternal age may increase the risk of miscarriage and pregnancy complications); (2) spouse age (Spouse’s age may indirectly impact pregnancy outcomes due to genetic or socioeconomic factors); (3) previous pregnancy times (Multiple pregnancies suggest fertility but may raise risks of uterine issues); (4) live birth times (The number of past live births may impact the current pregnancy’s prognosis);(5) endometrial injury history (endocervical curettage, uterine curettage, intrauterine device, hysteroscope, cesarean section, myomectomy); (6) number of spontaneous abortions (A history of miscarriages may indicate uterine issues, increasing the risk of SCH and pregnancy loss);(7) body mass index (BMI, kg/m2, high or low BMI can increase pregnancy risks, such as hypertension and diabetes);(8) Systolic blood pressure (SBP, mmHg); (9) Diastolic blood pressure (DBP, mmHg); (10) prothrombin time (PT, s); (11) activated partial thromboplastin time (APTT, s); (12) thrombin time (TT, s); (13) serum fibrinogen (Fib, g/l) level (fibrinogen is key to coagulation, and abnormal levels can affect bleeding risk and placental function); (14) serum Carbohydrate antigen-125 (CA125) level (U/mL) after SCH diagnosed; (15) SCH diagnosed time (d); (16) gestational sac maximum diameter detected by B-ultrasound test (mm); (17) hematoma maximum diameter B-ultrasound test (mm); (18) gestational sac maximum diameter to hematoma maximum diameter ratio (gsmd_hmd_ratio); (19) hematoma duration (prolonged hematoma may raise the risk of miscarriage or preterm birth); (20) the time serum human chorionic gonadotropin (HCG) level were measured >3I U/L for the first time at 27–36 days after menopause (the first positive HCG time) and (21) the level serum HCG level were measured >3 IU/L for the first time at 27–36 days after menopause (the first positive HCG level, IU/L, HCG is vital in early pregnancy, indicating placental function and health).

Pregnancy outcomes were collected by follow-up, including whether fetal loss (including abortion and stillbirth, abortion is defined as fetal loss less than 28 weeks of gestation, stillbirth is defined as pregnancy over 28 to fetal death); whether the delivery was successful (including premature and term births); whether the neonates are transferred (the transfer standard is 1 minute, or 5 minutes, or 10 minutes with an Apgar score <7, or birth weight <2500 g, or pregnancy week <37 weeks) to the Neonatal Intensive Care Unit (NICU); neonatal weight and gestational age.

FOLLOW-UP:

A total of 291 cases were followed up to February 2020 with outpatient service, telephone, WeChat, SMS, and case information inquiry. The total number of follow-ups was 42 days after delivery (4 patients were lost, and the follow-up rate was 96.36%).

TREATMENT METHODS:

All cases were pregnant women with a singleton pregnancy (HCG >3 IU/l) after natural pregnancy. fundamental treatment: Transvaginal ultrasound was dynamically monitored every 7 days after SCH diagnosed until SCH disappeared, and then routinely checked by obstetrics and obstetrics. If abdominal pain or extensive vaginal bleeding during the ultrasound interval, an emergency ultrasound examination is performed. Serum HCG level is reviewed every 2–7 days, progesterone and estradiol are stopped until 10–11 weeks of pregnancy. After SCH diagnosed, the subjects were instructed to reduce daily activity, progesterone capsules 40mg, orally BID or TID and 5000 IU low molecular weight heparin subcutaneous injection; dynamically monitor patients once a day or twice a day with the poor doubling of HCG level; if fetal heartbeat disappeared by ultrasound examination, or SCH women have a large amount of vaginal bleeding, blood volume greater than usual menstrual flow, or severe abdominal pain, or inevitable miscarriage were treated with Mifepristone and Misoprostol tablets orally or urgent uterine termination of pregnancy [7].

DATA DIVISION:

The dataset of SCH pregnant women was divided into a training set and a test set at a 6: 4 ratio. This division is essential for developing and validating predictive models.

SOFTWARE AND LIBRARIES: Python (3.7.0) and Jupyter Notebook (6.0.1) were used for data processing, modeling, and visualization. Key libraries included: NumPy (1.18.4) and Pandas (0.25.1) for data manipulation and analysis. Matplotlib (1.18.4) and Seaborn (1.18.4) for data visualization. Scikit-learn (1.18.4) for machine learning, including modules such as model_selection, linear_model, and metrics. XGBoost (1.18.4) for gradient boosting models. SciPy (1.18.4) for scientific computing. We used Tableone (1.18.4) for generating summary statistics tables, providing a clear overview of the dataset characteristics [8,9].

STATISTICAL TESTS:

t test: Applied to compare the means of continuous variables between 2 groups, assuming normally distributed data. Wilcoxon Rank Test: A non-parametric alternative to the t test, used when data do not follow a normal distribution. Chi-square Test: Used to assess the association between categorical variables. Fisher’s Exact Test: Applied for categorical data, especially when sample sizes are small, to determine if there are nonrandom associations between 2 categorical variables.

INDEPENDENT AND DEPENDENT VARIABLES:

Independent Variables (Features): The clinical indices collected from the training dataset. Dependent Variables (Targets): Various pregnancy outcomes, such as delivery outcome (live birth, stillbirth, or miscarriage), whether neonates entered the NICU (yes/no), success or failure of delivery (abortion/successful delivery), gestational age (term delivery/premature birth), and neonatal weight.

MODELING AND VALIDATION:

Machine Learning Classifiers: Used to model the relationship between clinical indices and pregnancy outcomes. Algorithms such as logistic regression and gradient boosting were employed to handle classification tasks.

Multivariate Linear Regression: Utilized to assess the impact of multiple independent variables on continuous dependent variables like neonatal weight, providing insights into how various factors contribute to outcomes.

MODEL EVALUATION:

Receiver Operating Characteristic (ROC) Curve: Plotted to evaluate the performance of classification models, illustrating the trade-off between sensitivity and specificity. Area Under Curve (AUC): Calculated to quantify the overall ability of the model to discriminate between different outcomes, with higher values indicating better performance. Calibration Plot: Used to assess the accuracy of predicted probabilities, ensuring that predicted risks correspond well to actual outcomes.

Residual Analysis: Conducted to evaluate the goodness-of-fit of regression models, identifying any patterns that suggest model improvements.

These statistical methods provide a comprehensive approach to analyzing the data, developing predictive models, and validating their effectiveness in predicting pregnancy outcomes associated with SCH.

Results

CLINICAL INDICES COMPARISON OF SCH WOMEN WITH THE DIFFERENT DELIVERY OUTCOME:

We found that 72.85% (n=212) of women with SCH delivered successfully and 27.15% (n=79) had unsuccessful delivery (stillbirth or miscarriage). In the same period, 4475 non-SCH women underwent routine prenatal examinations, 88.9% (n=3978) of which successful delivery, and 11.1% (n=497) unsuccessful delivery. The probability of failure delivery in women with SCH was significantly higher than that of non-SCH women (chi-square value=64.675, P<0.001) (Table 1).

The age, spouse age, APTT, fib, CA125, gestational sac maximum diameter, gsmd_hmd_ratio, and the first positive HCG level were significantly different (P<0.05, Table 1). The data of women with SCH were divided into the training dataset and test dataset (Table 2).

ML classifiers’ fitting was verified on the validation dataset, and the ROC is drawn, the AUC is 0.858 and 0.916, respectively. The recognition efficacy of XGB is better than logistic regression (Figure 1). The weight of the influence of each variable on the delivery outcome of SCH pregnant women is found from the XGB method, the first HCG positive level, hematoma duration, CA125 level, gestational sac maximum diameter, fibrinogen level and spouse age have a high impact on the outcome of SCH pregnant women’s delivery outcome (Figure 2).

IMPACT OF CLINICAL INDICES OF SCH WOMEN ON NEWBORNS ENTERING THE NICU:

Of the 212 women with SCH successfully delivered, 27 (12.7%) newborns were need to transfer into the NICU and 15 (87.3%) newborns were not; the age (t=6.080, P=0.000) and newborn birthweight (t=3.982, P=0.000) were significantly different, yet other clinical indices were not (P>0.05, Table 3). The data of the successful delivered were divided into the training dataset and test dataset (Table 4). The ML classifiers failed to find on the validation dataset that the clinical indices affecting whether whose newborns needed to transferred into the NICU (AUCs were 0.589 and 0.629, respectively, Figure 3)

IMPACT OF CLINICAL INDICES OF SCH WOMEN ON GESTATIONAL AGE:

The gestational age of the successfully delivered were statistical significantly longer than that of the unsuccessful delivered (38.8 (1.8) weeks vs 10.1 (4.1) weeks, t=−82.674, P=0.000). The gestational age of preterm infants is 34.8 (2.2) weeks, and which of full-term infants is 39.2 (1.0) weeks. There is no statistical difference in clinical indices of women with SCH in preterm and full-term (P>0.05, Table 5). The clinical indices of successfully delivered women with SCH were randomly divided into the training dataset and test dataset (Table 6). The ML classifiers failed to find on the validation set that the SCH pregnant women clinical indices affected whether whose newborns were born preterm or full-term (AUCs were 0.449 and 0.503, respectively) (Figure 4).

IMPACT OF CLINICAL INDICES OF SCH WOMEN ON NEONATAL WEIGHT:

The birth weight of full-term newborns was 3169.6 (454.00) g, and the birth weight of preterm newborns were 2538.4 (746.30) g, the difference was statistically significant (t=−4.208, P=0.001, Table 7).

Univariate and multivariate linear regression analysis was performed with neonatal birth weight as the response variable and clinical indices as independent variables. The result showed that the live birth times, PT, and gestational age were predictors to birth weight (P<0.05, Table 8); the result of multivariate linear regression analysis showed that the live birth times and gestational age were independent predictors to birth weights of the newborn (P<0.05). However, R-square and adjusted R-square were 0.233 and 0.226, respectively, which indicated the live birth times and gestational age can only explain about 23% of the newborn’s birth weight (Table 9).

Discussion

IMPACT OF CLINICAL INDICES OF PREGNANT WOMEN WITH SCH ON DELIVERY OUTCOME:

The proportion of pregnant women with SCH who failed to deliver was significantly higher than that of non-SCH pregnant women (27.15% vs 11.11%, P=0.000), which may indicate that the occurrence of SCH affected the delivery outcome, and the probability of abortion or stillbirth of SCH pregnant women is significantly increased. This aligns with Tiantian Xu’s study, showing higher miscarriage rates in women with SCH compared to those without [10].

Second, we explored the impact of SCH pregnant women’s clinical indices on the delivery outcome. By comparison, the age, spouse age, APTT, fib, CA125, gestational sac maximum diameter, gsmd_hmd_ratio, and the first positive HCG level were significantly different. And multivariate logistic regression analysis found that CA125 and hematoma duration were independent predictors to poor delivery outcome in SCH pregnant women (P<0.05). Finally, the XGBoost modeling result indicated that the first HCG positive level, hematoma duration, CA125, gestational sac maximum diameter, fib, and spouse age highly impacted the success of SCH pregnant women delivery outcome, especially the first HCG positive level and hematoma duration have the highest weight, and the AUC on the test dataset exceeds 0.9, which is higher than the AUC in the multivariate logistic regression analysis. This indicated that the XGOOST method is more accurate than logistic regression in predicting whether SCH pregnant women deliver successfully.

Our results are similarly and differently with those of other researchers. similarly, SCH affects the delivery outcome of pregnant women, which is consistent with the results of scholar mentioned above [11]; some scholars have reported that the size of the hematoma affects the delivery outcome, the larger the hematoma, the higher the miscarriage rate [12,13], due to the irregular shape of the hematoma, currently there is no standard for measuring the size of the hematoma. We use the hematoma maximum diameter to measure the size of the hematoma, and it was not found to have an impact on the delivery outcome. Differently, studies have reported that the older the SCH pregnant woman, the higher the miscarriage rate [14], but there was no significant difference in the age between them, and age has no impact on delivery outcome. Some studies found that the earlier the hematoma occurs, the higher the miscarriage rate [14], but there was no significant difference in the time when the hematoma was diagnosed.

The result showed that the first positive HCG level had the greatest influence on the outcome of SCH pregnant women, and no related studies have been found to specifically discuss this. Serum HCG levels significantly impact the delivery outcome in non-SCH pregnant women. Some scholars found that the low level of HCG in the first trimester is a risk factor for spontaneous abortion and intrauterine death [15]. Given the strong association between low HCG levels and adverse pregnancy outcomes, it is hypothesized that in SCH patients experiencing stillbirth or miscarriage, the expansion of the chorion towards the decidua may trigger the release of proteolytic enzymes from the outer syncytiotrophoblast cells due to various factors, resulting in decidual lesions [16], and eventually leads to more severe reduction of trophoblast cells and syncytia cells, which leads to lower HCG levels. This remains to be confirmed by further research.

We found that the hematoma duration is the second most important clinical feature that affects the delivery outcome of SCH pregnant women. Few related studies have specifically discussed it. Some scholars believe that the longer the hematoma duration, the worse the delivery outcome. Especially in pregnant women with persistent SCH, delivery is significantly earlier [17]. Another study found that longer hematoma duration is linked to higher risks of term premature rupture of membranes, gestational diabetes, and fetal growth restriction, but not other pregnancy outcomes [18]. Whether hematoma duration affects live birth has not been reported. Studies found that SCH will gradually shrink and even disappear with pregnancy, mostly disappearing around 1 to 3 months, and a few will run through the entire pregnancy [19,20]. This is similar to our result, but, strangely, the duration of hematoma in SCH pregnant women who successfully delivered was significantly longer than that of those who unsuccessfully delivered – 63.3 (16.8) days vs 45.7 (11.0) days (P<0.001). The reason may be that some pregnant women with SCH have had an abortion, so the duration of SCH is shorter. However, it was also suggested that pregnant women with SCH who can continue pregnancy beyond the first trimester are likely to eventually succeed in delivery.

The third most important clinical feature that impacted the delivery outcome of SCH pregnant women is serum CA125 level. Few studies have specifically explored this indicator for SCH pregnant women’s delivery outcomes. CA125 level increased significantly at 1–3 weeks after menopause. The increase of CA125 level in maternal serum may be due to the interaction between the placental trophoblast and the maternal endometrium in the first trimester; when normally fertilized eggs implanted, trophoblasts penetrate and penetrate the endometrium, and the endometrium and muscle tissue are destroyed. During blastocyst implantation, the blastocyst becomes fully embedded within and covered by the endometrium, initiating the release of CA125. Post-implantation, the endometrial tissue experiences damage and undergoes remodeling, resulting in a marked elevation in CA125 levels. This elevation is attributed to the interactions between endometrial cells and trophoblast cells. As gestation advances, the blastocyst continues to develop, and the endometrium gradually completes its repair process. The subsequent healing and stabilization of the endometrium lead to reduced tissue damage, thereby gradually decreasing the release of CA125. Ultimately, serum CA125 levels return to normal, indicating the stabilization of the endometrial environment and the healthy progression of the pregnancy. The normalization of CA125 levels is indicative of successful endometrial repair following implantation, facilitating the smooth progression of the pregnancy [21]. Because CA125 is present in the chorionic-decidual plate after pregnancy, maternal serum CA125 level also increases after pregnancy. SCH pregnant women have bleeding under the chorion and can release CA125 into the blood. A study found that the more severe the trophoblast damage, the higher the CA125 level [22]. Researchers found that CA125 levels in women with hematomas were significantly higher than in those without hematomas, which indicated serum CA125 level was an effective predictor of delivery outcome in early pregnancy. Although the study did not focus on SCH pregnant women [23], its results were similar to our findings.

IMPACT OF SCH PREGNANT WOMEN’S CLINICAL INDICES ON WHETHER NEWBORNS NEED TO BE TRANSFERRED INTO THE NICU:

Except for gestational age and neonatal birth weight (P<0.05), we found no significant association between clinical indices of pregnant women with SCH whose newborns needed to be transferred into the NICU and those who did not go to the NICU. Logistic regression analysis and XGBOOST modeling showed that the clinical indices of pregnant women with SCH did not have an impact because lower gestational age and lower birth weight are criteria for transfer to the NICU.

There are few reports on the status of newborns of pregnant women with SCH after delivery, especially whether they need to be transferred to the NICU. A study reported that pregnant women with hematomas have a significantly higher probability of having newborns transferred into the NICU after delivery than pregnant women with non-hematomas (P=0.015), which is inconsistent with our result. However, that study included SCH and post-placental hematoma, and there did not specifically assess pregnant women with SCH [6].

SCH PREGNANT WOMEN’S CLINICAL INDICES AND THE GESTATIONAL AGE OF NEWBORNS:

The mean gestational age of the successful delivered newborns was 38.8 (1.8) weeks, and the gestational age of aborted or stillborn fetuses was only 10.1 (4.1) weeks. The difference was statistically significant. It is suggested that the miscarriage or stillbirth of SCH pregnant women occurred in the first trimester. If they can get through the first trimester, they are likely to achieve a successful delivery.

Besides, the gestational age of premature infants of SCH pregnant women was 34.8 (2.2) weeks, and the gestational age of full-term infants was 39.2 (1.0) weeks; there was no statistical difference in clinical indices (P>0.5). Logistic regression analysis and XGBOOST modeling results also did not find that the clinical indices of SCH pregnant women affected whether the newborns were born preterm or full-term. This suggested that the clinical indices of pregnant women with SCH have no relation with gestational age. This is similar to reports that SCH does not affect whether a birth is preterm or not [10,24]; however, a meta-analysis found that SCH increases the probability of preterm birth [1].

ASSOCIATION OF CLINICAL INDICES OF PREGNANT WOMEN WITH SCH WITH NEWBORN BIRTH WEIGHT:

The birth weight of full-term infants was 3169.6 (454.00) g, and that of preterm infants was 2538.4 (746.30) g. The difference was statistically significant (P=0.001). This result is consistent with the fact that the birth weight of preterm infants is commonly lower than that of full-term infants [25]. Multivariate linear regression analysis showed that the live birth times and gestational age were independent factors associated with newborn birth weight (P<0.05). However, R-square and adjusted R-square were only 0.233 and 0.226, respectively, which means the live birth times and gestational age can only explain about 23% of the newborn’s birth weight, which suggested that if pregnant women with SCH can deliver successfully, their clinical indices have no impact on the birth weight of the newborn.

CLINICAL SIGNIFICANCE AND MANAGEMENT IMPLICATIONS OF SCH IN PREGNANT WOMEN:

This study’s findings significantly impact managing pregnant women with SCH. Key measures included the following: monitoring the first positive HCG level as a predictor of delivery outcomes; recognizing that persistent hematomas beyond the first trimester may not predict poor outcomes, which can guide management and reassure patients; Incorporating CA125 levels into assessments to identify those needing closer monitoring; focusing on gestational age and birth weight for NICU admission, considering SCH as a factor in preterm birth risk; and using advanced modeling like XGBOOST for more accurate delivery outcome predictions, allowing personalized care plans through machine learning.

SAMPLE SIZE AND SELECTION BIAS:

This was a retrospective study collecting clinical data of pregnant women with and without SCH and with the prenatal examination. We enrolled nearly 300 pregnant women and it was a single-center study, so there may be selection bias; however, the sample size of pregnant women with SCH is one of the largest in recent years.

INCOMPLETE COVERAGE OF CLINICAL INDICES:

As the pregnant women with SCH and non-SCH ones had no systematic and comparable clinical indices during prenatal examinations or other examinations, which means important clinical parameters that could provide insights into the health status and risk factors of these women are not consistently recorded or analyzed. Consequently, the ability to fully identify and control for confounding factors that might influence the pregnancy outcomes for women with SCH is significantly compromised.

LACK OF STANDARDS FOR EVALUATING SCH SIZE AND SEVERITY: The parameters of SCH can only be calculated by the B-ultrasound test to evaluate the size and the severity of the hematoma. However, there is no consensus on which assessment methods are best. Some researchers propose using the chorionic area to gestational sac area ratio to evaluate the severity of SCH [26], while others use hematoma volume calculation method and hematoma circumference and depth to comprehensively evaluate the size and severity [27]. We assessed the size of the hematoma with the maximum hematoma diameter and evaluated the severity with the maximum gestational sac diameter to the maximum hematoma diameter ratio, but further research is needed to determine the best assessment methods.

Conclusions

The current study collected clinical indices of pregnant women with SCH and non-SCH pregnant women. Using statistical analysis methods and classification methods in ML, we conclude the following: (1) SCH affected the delivery outcome, and the probability of miscarriage or stillbirth in SCH pregnant women is significantly higher than that of non-SCH counterparts. (2) The level of the first positive HCG, the hematoma duration, serum CA125 level, the gestational sac maximum diameter, fibrinogen, and the age of the spouse highly impacted the delivery outcome of pregnant women with SCH. (3) Clinical indices of pregnant women with have no impact on whether the newborns were transferred to the NICU. (4) Abortion or stillbirth of pregnant women with SCH commonly occurred in the first trimester, and if they can get through this period, they are likely to have a successful delivery. SCH was not associated with gestational age of pregnant women with SCH. (5) The birth weight of full-term newborns delivered by pregnant women with SCH is significantly higher than that of preterm infants. The clinical indices of pregnant women with SCH were not associated with newborn birth weight.

References

1. Tuuli MG, Norman SM, Odibo AO, Perinatal outcomes in women with subchorionic hematoma: A systematic review and meta-analysis: Obstet Gynecol, 2011; 117(5); 1205-12

2. Ogawa M, Nagao D, Mori K, Elastography for differentiation of subchorionic hematoma and placenta previa: Ultrasound Obstet Gynecol, 2012; 39(1); 112-14

3. Alowais SA, Alghamdi SS, Alsuhebany N, Revolutionizing healthcare: The role of artificial intelligence in clinical practice: BMC Med Educ, 2023; 23(1); 689

4. Javaid M, Haleem A, Pratap Singh R, Significance of machine learning in healthcare: Features, pillars and applications: Int J Intell Netw, 2022; 3; 58-73

5. Zawacki-Richter O, Marín VI, Bond M, Gouverneur F, Systematic review of research on artificial intelligence applications in higher education – where are the educators?: Int J Educ Technol High Educ, 2019; 16(1); 39

6. Hashem A, Sarsam SD, The impact of incidental ultrasound finding of subchorionic and retroplacental hematoma in early pregnancy: J Obstet Gynaecol India, 2019; 69(1); 43-49

7. Lemmers M, Verschoor MA, Kim BV, Medical treatment for early fetal death (less than 24 weeks): Cochrane Database Syst Rev, 2019; 6; CD002253

8. Ogunleye AA, Qing-Guo W, XGBoost model for chronic kidney disease diagnosis: IEEE/ACM Trans Comput Biol Bioinform, 2020; 17(6); 2131-40

9. Li S, Zhang X, Research on orthopedic auxiliary classification and prediction model based on XGBoost algorithm: Neural Comput Appl, 2020; 32(59) s00521-019-04378-4

10. Xu T, Lun W, Wang P, He Y, Analysis of risk factors and pregnancy outcomes in pregnant women with subchorionic hematoma: Medicine (Baltimore), 2023; 102(47); e35874

11. Li Q, Zhu J, Hua KEffects of subchorionic hematoma on pregnancy outcome: A meta analysis: Zhonghua Yi Xue Za Zhi, 2016; 96(17); 1383-85

12. Janowiczgrelewska A, Sieroszewski P, Prognostic significance of subchorionic hematoma for the course of pregnancy: Ginekol Pol, 2013; 84(11); 944-49

13. Dongol A, Mool S, Tiwari P, Outcome of pregnancy complicated by threatened abortion: Kathmandu Univ Med J, 2011; 9(1); 41-44

14. Bennett GL, Bromley B, Lieberman E, Benacerraf BR, Subchorionic hemorrhage in first-trimester pregnancies: Prediction of pregnancy outcome with sonography: Radiology, 1996; 200(3); 803-6

15. Zhao T, Zhang H, Zhao X, Cheng W, Tao J, Association between adverse obstetric outcomes and abnormal maternal serum markers in the second trimester screening: J Shanghai Jiaotong Univ (Med Sci), 2017; 37(8); 1127-31

16. Pillai RN, Konje JC, Tincello DG, Potdar NJ, Role of serum biomarkers in the prediction of outcome in women with threatened miscarriage: A systematic review and diagnostic accuracy meta-analysis: Hum Reprod Update, 2015; 22(2); 228-39

17. Xiang L, Wei Z, Cao Y, Symptoms of an intrauterine hematoma associated with pregnancy complications: a systematic review: PLoS One, 2014; 9(11); e111676

18. Pan S, Lan Y, Zhou Y, Associations between the size and duration of asymptomatic subchorionic hematoma and pregnancy outcomes in women with singleton pregnancies: BMC Pregnancy Childbirth, 2023; 23(1); 555

19. Asada K, Furukawa R, Saito Y, Massive subchorionic thrombohematoma: Termination should not be discouraged by antenatal diagnosis: Acta Obstet Gynecol Scand, 2011; 90(8); 927-28

20. Windrim C, Athaide G, Gerster T, Kingdom JCP, Sonographic findings and clinical outcomes in women with massive subchorionic hematoma detected in the second trimester: J Obstet Gynaecol Can, 2011; 33(5); 475-79

21. Shehata NAA, Ali HAA, Hassan A, Doppler and biochemical assessment for the prediction of early pregnancy outcome in patients experiencing threatened spontaneous abortion: Int J Gynaecol Obstet, 2018; 143(2); 150-55

22. Madendag Y, Col-Madendag I, Kanat-Pektas M, Danisman N, Predictive power of serum CA-125 and LDH in the outcome of first trimester pregnancies with human chorionic gonadotropin levels below discriminatory zone: Arch Gynecol Obstet, 2009; 279(5); 661-66

23. Mohamady MA, Fattah GA, Elkattan E, Correlation of serum CA-125 and progesterone levels with ultrasound markers in the prediction of pregnancy outcome in threatened miscarriage: Int J Fertil Steril, 2016; 9(4); 506-11

24. Pedersen JF, Mantoni M, Prevalence and significance of subchorionic hemorrhage in threatened abortion: A sonographic study: Am J Roentgenol, 1990; 154(3); 535-37

25. Del Rio R, Thio M, Bosio MPrediction of mortality in premature neonates. An updated systematic review: An Pediatr (Barc), 2020; 93(1); 24-33 [in Spanish]

26. Aoki S, Inagaki M, Kurasawa K, Retrospective study of pregnant women placed under expectant management for persistent hemorrhage: Arch Gynecol Obstet, 2014; 289(2); 307-11

27. Podrasky AE, Javitt MC, Glanc P, ACR appropriateness Criteria(R) second and third trimester bleeding: Ultrasound Q, 2013; 29(4); 293-301

Tables

Table 1. Clinical indices comparison of SCH women with the different delivery outcome.Table 2. Clinical indices of SCH women with different delivery outcome divided into training and test dataset.Table 3. Comparison of clinical indices of SCH women whose newborns whether entering the NICU (mean, std).Table 4. The training and test datasets of women with SCH whose newborns entering the NICU (mean, std).Table 5. Comparison of clinical indices of premature and term women with SCH (mean, std).Table 6. Successful delivery women with SCH divided into training dataset and test dataset (mean, std).Table 7. Birth weight comparison between preterm and full-term newborns of women with SCH [mean (std)].Table 8. Univariate linear regression analysis of newborn birth weight.Table 9. Result of multivariate linear regression analysis to newborns birth weight.Table 1. Clinical indices comparison of SCH women with the different delivery outcome.Table 2. Clinical indices of SCH women with different delivery outcome divided into training and test dataset.Table 3. Comparison of clinical indices of SCH women whose newborns whether entering the NICU (mean, std).Table 4. The training and test datasets of women with SCH whose newborns entering the NICU (mean, std).Table 5. Comparison of clinical indices of premature and term women with SCH (mean, std).Table 6. Successful delivery women with SCH divided into training dataset and test dataset (mean, std).Table 7. Birth weight comparison between preterm and full-term newborns of women with SCH [mean (std)].Table 8. Univariate linear regression analysis of newborn birth weight.Table 9. Result of multivariate linear regression analysis to newborns birth weight.

In Press

Clinical Research  

Evaluation of Knowledge, Attitudes, and Practices Toward Thyroid Nodules in 456 Patients with Thyroid Nodules

Med Sci Monit In Press; DOI: 10.12659/MSM.945732  

Review article  

Comparative Efficacy of 3D-Printed Insoles in Managing Common Foot Conditions: A Review

Med Sci Monit In Press; DOI: 10.12659/MSM.947252  

Clinical Research  

Correlation Descriptive Study on Nurses' Professional Values and Job Satisfaction and the Ethical Climate o...

Med Sci Monit In Press; DOI: 10.12659/MSM.945639  

Clinical Research  

Pre- and Post-Surgical MRI Analysis of Levator Ani in Pelvic Organ Prolapse Patients: A Single-Center Study

Med Sci Monit In Press; DOI: 10.12659/MSM.945993  

Most Viewed Current Articles

17 Jan 2024 : Review article   6,967,723

Vaccination Guidelines for Pregnant Women: Addressing COVID-19 and the Omicron Variant

DOI :10.12659/MSM.942799

Med Sci Monit 2024; 30:e942799

0:00

16 May 2023 : Clinical Research   701,731

Electrophysiological Testing for an Auditory Processing Disorder and Reading Performance in 54 School Stude...

DOI :10.12659/MSM.940387

Med Sci Monit 2023; 29:e940387

0:00

01 Mar 2024 : Editorial   25,246

Editorial: First Regulatory Approvals for CRISPR-Cas9 Therapeutic Gene Editing for Sickle Cell Disease and ...

DOI :10.12659/MSM.944204

Med Sci Monit 2024; 30:e944204

0:00

28 Jan 2024 : Review article   19,861

A Review of IgA Vasculitis (Henoch-Schönlein Purpura) Past, Present, and Future

DOI :10.12659/MSM.943912

Med Sci Monit 2024; 30:e943912

0:00

Your Privacy

We use cookies to ensure the functionality of our website, to personalize content and advertising, to provide social media features, and to analyze our traffic. If you allow us to do so, we also inform our social media, advertising and analysis partners about your use of our website, You can decise for yourself which categories you you want to deny or allow. Please note that based on your settings not all functionalities of the site are available. View our privacy policy.

Medical Science Monitor eISSN: 1643-3750
Medical Science Monitor eISSN: 1643-3750