First-trimester screening for preeclampsia: impact of maternal parity on modeling and screening effectiveness

Machine Learning Algorithms Versus Classical Regression Models in Pre-Eclampsia Prediction: A Systematic Review

PURPOSE OF REVIEW

Machine learning (ML) approaches are an emerging alternative for healthcare risk prediction. We aimed to synthesise the literature on ML and classical regression studies exploring potential prognostic factors and to compare prediction performance for pre-eclampsia.

RECENT FINDINGS

From 9382 studies retrieved, 82 were included. Sixty-six publications exclusively reported eighty-four classical regression models to predict variable timing of onset of pre-eclampsia. Another six publications reported purely ML algorithms, whilst another 10 publications reported ML algorithms and classical regression models in the same sample with 8 of 10 findings that ML algorithms outperformed classical regression models. The most frequent prognostic factors were age, pre-pregnancy body mass index, chronic medical conditions, parity, prior history of pre-eclampsia, mean arterial pressure, uterine artery pulsatility index, placental growth factor, and pregnancy-associated plasma protein A. Top performing ML algorithms were random forest (area under the curve (AUC) = 0.94, 95% confidence interval (CI) 0.91-0.96) and extreme gradient boosting (AUC = 0.92, 95% CI 0.90-0.94). The competing risk model had similar performance (AUC = 0.92, 95% CI 0.91-0.92) compared with a neural network. Calibration performance was not reported in the majority of publications. ML algorithms had better performance compared to classical regression models in pre-eclampsia prediction. Random forest and boosting-type algorithms had the best prediction performance. Further research should focus on comparing ML algorithms to classical regression models using the same samples and evaluation metrics to gain insight into their performance. External validation of ML algorithms is warranted to gain insights into their generalisability.

Maternal race and pre-eclampsia: Cohort study and systematic review with meta-analysis

OBJECTIVES

To examine the association between race and pre-eclampsia and gestational hypertension after adjustment for factors in maternal characteristics and medical history in a screening study from the Fetal Medicine Foundation (FMF) in England, and to perform a systematic review and meta-analysis of studies on pre-eclampsia.

DESIGN

Prospective observational study and systematic review with meta-analysis.

SETTING

Two UK maternity hospitals.

POPULATION

A total of 168 966 women with singleton pregnancies attending for routine ultrasound examination at 11-13 weeks of gestation without major abnormalities delivering at 24 weeks or more of gestation.

METHODS

Regression analysis examined the association between race and pre-eclampsia or gestational hypertension in the FMF data. Literature search to December 2021 was carried out to identify peer-reviewed publications on race and pre-eclampsia.

MAIN OUTCOME MEASURE

Relative risk of pre-eclampsia and gestational hypertension in women of black, South Asian and East Asian race by comparison to white women.

RESULTS

In black women, the respective risks of total-pre-eclampsia and preterm-pre-eclampsia were 2-fold and 2.5-fold higher, respectively, and risk of gestational hypertension was 25% higher; in South Asian women there was a 1.5-fold higher risk of preterm pre-eclampsia but not of total-pre-eclampsia and in East Asian women there was no statistically significant difference in risk of hypertensive disorders. The literature search identified 19 studies that provided data on several million pregnancies, but 17 were at moderate or high-risk of bias and only three provided risks adjusted for some maternal characteristics; consequently, these studies did not provide accurate contributions on different racial groups to the prediction of pre-eclampsia.

CONCLUSION

In women of black and South Asian origin the risk of pre-eclampsia, after adjustment for confounders, is higher than in white women.

Predictive Performance of Placental Protein 13 for Screening Preeclampsia in the First Trimester: A Systematic Review and Meta-Analysis

Preeclampsia is a pregnancy-specific syndrome that affects maternal and neonatal mortality. Several serum biomarkers can be used to predict preeclampsia. Among these proteins, placental protein 13 (PP13) has received progressively more interest in recent studies. The decrease in PP13 expression is one of the earliest signs for the development of preeclampsia and has shown its predictive performance for preeclampsia. In this meta-analysis, we collected 17 observational studies with 40,474 pregnant women. The overall sensitivity of PP13 to predict preeclampsia was 0.62 [95% confidence interval (CI) = 0.49–0.74], the specificity was 0.84 (95%CI = 0.81–0.86), and the diagnostic odds ratio was nine (95%CI = 5–15). The area under the curve for summary receiver operating characteristic was 0.84. We then chose the early-onset preeclampsia as a subgroup. The sensitivity of early-onset subgroup was 0.63 (95%CI = 0.58–0.76), the specificity was 0.85 (95%CI = 0.82–0.88), and the diagnostic odds ratio was 10 (95%CI = 6–18). The findings of our meta-analysis indicate that PP13 may be an effective serum biomarker for the predictive screening of preeclampsia. Nonetheless, large prospective cohort studies and randomized controlled trials are expected to uncover its application in clinical practice. The heterogeneity of the original trials may limit the clinical application of PP13.

Systematic Review Registration:https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=188948 The meta-analysis was registered in PROSPERO (CRD42020188948).

Preeclampsia Risk Prediction Model for Chinese Pregnant Patients With Systemic Lupus Erythematosus

OBJECTIVE

To screen for a high risk of preeclampsia in women with systemic lupus erythematosus (SLE).

METHODS

A total of 513 antenatal care records of pregnant patients with SLE were obtained, and the data were randomly assigned to either a development set (n = 342) or a validation set (n = 171). Preeclampsia predictors were identified with stepwise regression, and a coefficient B of each variable was used to establish a prediction model and risk scoring system. Goodness-of-fit was assessed by the Hosmer-Lemeshow and Omnibus tests, and the area under the receiver operating characteristic curve (area under the curve) was used to assess discrimination. Validation was performed using the validation set.

RESULTS

The preeclampsia incidence was 14.4% in the pregnant patients with SLE. A mean arterial pressure (MAP) ≥96.5 mm Hg (odds ratio [OR] 213.15 [95% confidence interval (95% CI) 24.39-999.99]), prepregnancy hypertension (OR 18.19 [95% CI 2.67-125.01]), a hematologic disorder (OR 4.13 [95% CI 1.03-16.67]), positive IgM anticardiolipin antibodies (aCLs) (OR 19.85 [95% CI 1.11-333.33]), serum albumin <31.5 grams/liter (OR 9.88 [95% CI 2.07-47.62]), serum uric acid ≥303 μmoles/liter (OR 5.58 [95% CI 1.40-22.22]), and 24-hour urinary protein ≥0.286 grams (OR 14.39 [95% CI 2.43-83.33]) were selected for the preeclampsia prediction model. The area under the curve was 0.975. Preeclampsia prediction model scores >4 indicated a high risk of preeclampsia. For the validation set, the preeclampsia prediction accuracy was 93.6% (sensitivity 88.5%, specificity 94.5%).

CONCLUSION

A model for predicting the risk of preeclampsia in pregnant patients with SLE was established on the basis of MAP, prepregnancy hypertension, hematologic disorders, IgM aCLs, albumin, uric acid, and 24-hour urinary protein. The model had good predictive efficiency and can help clinicians improve pregnancy outcomes in high-risk women with early interventions.

First-Trimester Preterm Preeclampsia Screening in Nulliparous Women: The Great Obstetrical Syndrome (GOS) Study

OBJECTIVES

To estimate the ability of a combination of first-trimester markers to predict preterm preeclampsia in nulliparous women.

METHODS

We conducted a prospective cohort study of nulliparous women with singleton gestations, recruited between 11⁰ and 13⁶ weeks gestation. Data on the following were collected: maternal age; ethnicity; chronic diseases; use of fertility treatment; body mass index; mean arterial blood pressure (MAP); serum levels of pregnancy-associated plasma protein A (PAPP-A), placental growth factor (PlGF), soluble fms-like tyrosine kinase-1 (sFlt-1), alpha fetoprotein (AFP), free beta human chorionic gonadotropin (ß-hCG); and mean uterine artery pulsatility index (UtA-PI). We constructed a proportional hazard model for the prediction of preterm preeclampsia selected based on the Akaike information criterion. A receiver operating characteristic curve was created with the predicted risk from the final model. Our primary outcome was preterm preeclampsia and our secondary outcome was a composite of preeclampsia, small for gestational age, intrauterine death, and preterm birth.

RESULTS

Among 4659 nulliparous women with singleton gestations, our final model included 4 variables: MAP MoM, log₁₀PlGF MoM, log₁₀AFP MoM and log₁₀UtA-PI MoM. We obtained an area under the curve of 0.84 (95% CI 0.75-0.93) with a detection rate of preterm preeclampsia of 55% (95% CI 37%-73%) and a false-positive rate of 10%. Using a risk cut-off with a false-positive rate of 10%, the positive predictive value for our composite outcome was 33% (95% CI 29%-37%).

CONCLUSIONS

The combination of MAP, maternal serum PlGF and AFP, and UtA-PI are useful to identify nulliparous women at high risk of preterm preeclampsia but also at high risk of other great obstetrical syndromes.

Systematic review of prediction models for gestational hypertension and preeclampsia

INTRODUCTION

Prediction models for gestational hypertension and preeclampsia have been developed with data and assumptions from developed countries. Their suitability and application for low resource settings have not been tested. This review aimed to identify and assess the methodological quality of prediction models for gestational hypertension and pre-eclampsia with reference to their application in low resource settings.

METHODS

Using combinations of keywords for gestational hypertension, preeclampsia and prediction models seven databases were searched to identify prediction models developed with maternal data obtained before 20 weeks of pregnancy and including at least three predictors (Prospero registration CRD 42017078786). Prediction model characteristics and performance measures were extracted using the CHARMS, STROBE and TRIPOD checklists. The National Institute of Health quality assessment tools for observational cohort and cross-sectional studies were used for study quality appraisal.

RESULTS

We retrieved 8,309 articles out of which 40 articles were eligible for review. Seventy-seven percent of all the prediction models combined biomarkers with maternal clinical characteristics. Biomarkers used as predictors in most models were pregnancy associated plasma protein-A (PAPP-A) and placental growth factor (PlGF). Only five studies were conducted in a low-and middle income country.

CONCLUSIONS

Most of the studies evaluated did not completely follow the CHARMS, TRIPOD and STROBE guidelines in prediction model development and reporting. Adherence to these guidelines will improve prediction modelling studies and subsequent application of prediction models in clinical practice. Prediction models using maternal characteristics, with good discrimination and calibration, should be externally validated for use in low and middle income countries where biomarker assays are not routinely available.

Effect of Parity on Pregnancy-Associated Hypertension Among Asian American Women in the United States

OBJECTIVES

Pregnancy-associated hypertension (PAH) includes gestational hypertension, preeclampsia and eclampsia. Although a protective effect of multi-parity on PAH has been reported in previous studies, the association has not been examined among Asian American women in the U.S.

METHODS

Using data from 2014 U.S. National Vital Statistics System, we examined the prevalence of PAH among Asian American women who had singleton live births (N = 235,303), and its association with parity (number of previous pregnancies including live births and fetal deaths) controlling for potential confounders. We estimated adjusted odds ratios (aORs) and 95% confidence intervals (CI) using multivariable logistic regression analysis.

RESULTS

Overall, 2.72% (95% CI 2.66%, 2.79%) of Asian American women were recorded to have PAH during pregnancy. Parity was inversely associated with PAH in our study, where Asian American women who had 1-2 and 3 or more previous pregnancies had significantly lower odds of PAH (aOR 0.61, 95% CI 0.58, 0.65; and aOR 0.62, 95% CI 0.57, 0.68, respectively) compared to nulliparous women, after controlling for potential confounders.

CONCLUSIONS

Recent U.S. vital statistics data revealed that nulliparity is significantly associated with PAH among Asian American women. Future studies should identify specific factors that are associated with PAH and factors contributing to disparities in PAH risk among Asian American women.

Prediction models for preeclampsia: A systematic review

BACKGROUND

Preeclampsia is a disease specific to pregnancy that can cause severe maternal and foetal morbidity and mortality. Early identification of women at higher risk for preeclampsia could potentially aid early prevention and treatment. Although a plethora of preeclampsia prediction models have been developed in recent years, individualised prediction of preeclampsia is rarely used in clinical practice.

OBJECTIVES

The objective of this systematic review was to provide an overview of studies on preeclampsia prediction.

STUDY DESIGN

Relevant research papers were identified through a MEDLINE search up to 1 January 2017. Prognostic studies on the prediction of preeclampsia or preeclampsia-related disorders were included. Quality screening was performed with the Quality in Prognostic Studies (QUIPS) tool.

RESULTS

Sixty-eight prediction models from 70 studies with 425,125 participants were selected for further review. The number of participants varied and the gestational age at prediction varied widely across studies. The most frequently used predictors were medical history, body mass index, blood pressure, parity, uterine artery pulsatility index, and maternal age. The type of predictor (maternal characteristics, ultrasound markers and/or biomarkers) was not clearly associated with model discrimination. Few prediction studies were internally (4%) or externally (6%) validated.

CONCLUSIONS

To date, multiple and widely varying models for preeclampsia prediction have been developed, some yielding promising results. The high degree of between-study heterogeneity impedes selection of the best model, or an aggregated analysis of prognostic models. Before multivariable preeclampsia prediction can be clinically implemented universally, further validation and calibration of well-performing prediction models is needed.

Prediction of first-trimester preeclampsia: Relevance of the oxidative stress marker MDA in a combination model with PP-13, PAPP-A and beta-HCG

OBJECTIVE

Early diagnosis of preeclampsia (PE) is very important and various parameters, individually or in combined models, are reported useful for prediction of PE. The objective of this study is to investigate the predictive value of pregnancy-associated plasma protein-A (PAPP-A), placental protein-13 (PP-13), human Chorionic Gonadotropin (B-HCG), and oxidative stress marker malondialdehyde (MDA), individually and in combination.

MATERIALS AND METHODS

Maternal sera of 38 cases with PE and 122 controls were collected for first trimester screening and tested for PAPP-A and B-HCG by chemiluminescence, for PP-13 by using ELISA, and for MDA by high-performance liquid chromatography. Combined models of parameters were constituted as "MDA + PP-13", "PP-13 + PAPP-A + B-HCG" and "MDA + PP-13 + PAPP-A + B-HCG". The diagnostic performances of serum markers of preeclampsia were examined by nonparametric receiver-operator characteristics (ROC) analysis.

RESULTS

PP-13 levels were significantly lower (p < 0.001) and MDA levels were significantly higher (p < 0.001) in PE. The area under the ROC curve (AUC) for MDA and PP-13 were greater than those for PAPP-A and B-HCG (p < 0.001). The AUCs of the combined models were significantly larger than those of individual parameters. The combined model "MDA + PP-13 + PAPP-A + B-HCG" exhibited the best predictive outcome with an AUC of 0.91 [95% CI 0.86-0.95], 97% [95% CI 86.2-99.9] sensitivity and 75% [95% CI 66.5-82.6] specificity, and was significantly different from that of "PAPP-A + PP-13 + B-HCG" model, but similar to that of "MDA + PP-13" model.

CONCLUSION

Combined models consisting of various parameters of different origin, may provide better predictive outcomes, and oxidative markers should be considered in combination with other placental biomarkers in prediction of PE.

The maternal plasma proteome changes as a function of gestational age in normal pregnancy: a longitudinal study

OBJECTIVE

Pregnancy is accompanied by dramatic physiological changes in maternal plasma proteins. Characterization of the maternal plasma proteome in normal pregnancy is an essential step for understanding changes to predict pregnancy outcome. The objective of this study was to describe maternal plasma proteins that change in abundance with advancing gestational age and determine biological processes that are perturbed in normal pregnancy.

STUDY DESIGN

A longitudinal study included 43 normal pregnancies that had a term delivery of an infant who was appropriate for gestational age without maternal or neonatal complications. For each pregnancy, 3 to 6 maternal plasma samples (median, 5) were profiled to measure the abundance of 1125 proteins using multiplex assays. Linear mixed-effects models with polynomial splines were used to model protein abundance as a function of gestational age, and the significance of the association was inferred via likelihood ratio tests. Proteins considered to be significantly changed were defined as having the following: (1) >1.5-fold change between 8 and 40 weeks of gestation; and (2) a false discovery rate-adjusted value of P < .1. Gene ontology enrichment analysis was used to identify biological processes overrepresented among the proteins that changed with advancing gestation.

RESULTS

The following results were found: (1) Ten percent (112 of 1125) of the profiled proteins changed in abundance as a function of gestational age; (2) of the 1125 proteins analyzed, glypican-3, sialic acid-binding immunoglobulin-type lectin-6, placental growth factor, C-C motif-28, carbonic anhydrase 6, prolactin, interleukin-1 receptor 4, dual-specificity mitogen-activated protein kinase 4, and pregnancy-associated plasma protein-A had more than a 5-fold change in abundance across gestation (these 9 proteins are known to be involved in a wide range of both physiological and pathological processes, such as growth regulation, embryogenesis, angiogenesis immunoregulation, inflammation etc); and (3) biological processes associated with protein changes in normal pregnancy included defense response, defense response to bacteria, proteolysis, and leukocyte migration (false discovery rate, 10%).

CONCLUSION

The plasma proteome of normal pregnancy demonstrates dramatic changes in both the magnitude of changes and the fraction of the proteins involved. Such information is important to understand the physiology of pregnancy and the development of biomarkers to differentiate normal vs abnormal pregnancy and determine the response to interventions.

Revisiting the discriminatory accuracy of traditional risk factors in preeclampsia screening

BACKGROUND

Preeclampsia (PE) is associated with a high risk of perinatal morbidity and mortality. However, there is no consensus in the definition of high-risk women.

AIM

To question current definition of high PE risk and propose a definition that considers individual heterogeneity to improves risk classification.

METHODS

A stratified analysis by parity was conducted using the Swedish Birth Register between 2002-2010 including 626.600 pregnancies. The discriminatory accuracy (DA) of traditional definitions of high-risk women was compared with a new definition based on 1) specific combinations of individual variables and 2) a centile cut-off of the probability of PE predicted by a multiple logistic regression model.

RESULTS

None of the classical risk-factors alone reached an acceptable DA. In multiparous, any combination of a risk-factor with previous PE or HBP reached a +LR>10. The combination of obesity and multiple pregnancy reached a good DA particularly in the presence of previous preeclampsia (positive likelihood ratio (LR+) = 26.5 or chronic hypertension (HBP) LR+ = 40.5. In primiparous, a LR+>15 was observed in multiple pregnancies with the simultaneous presence of obesity and diabetes mellitus or with HBP. Predicted probabilities above 97 centile in multiparous and 99 centile in primiparous provided high (LR+ = 12.5), and moderate (LR+ = 5.85), respectively. No one risk factor alone or in combination provided a LR- sufficiently low to rule-out the disease.

CONCLUSIONS

In preeclampsia prediction the combination of specific risk factors provided a better discriminatory accuracy than traditional single risk approach. Our results contribute to a more personalized risk estimation of preeclampsia.