Prediction of progression to severe disease in women with late preterm hypertensive disorders of pregnancy

Maternal birth weight as an indicator of early-onset and late-onset hypertensive disorders of pregnancy: The Japan Environment and Children's study

OBJECTIVES

This study aimed to investigate the association between maternal birth weight (MBW) and hypertensive disorders of pregnancy (HDP) according to the gestational age when HDP develops.

STUDY DESIGN

A total of 77,345 subjects were included in this prospective birth cohort study. The association between MBW and HDP was investigated by a multinomial logistic regression model.

MAIN OUTCOME MEASURES

Early-onset HDP (EO-HDP), preterm late-onset HDP (preterm LO-HDP), and term late-onset HDP (term LO-HDP).

RESULTS

Lower MBW was associated with higher odds of preterm and term LO-HDP (p-values for trend < 0.0001 and = 0.0005, respectively). A linear association between MBW and EO-HDP was observed (p-values for trend = 0.0496). The shape of the association between MBW and preterm LO-HDP was a combination of the associations between MBW with EO-HDP or LO-HDP. The effect size of the association between MBW < 2,500 g and EO-HDP was lower than that of MBW < 2,500 g with preterm or term LO-HDP. The adjusted odds ratios for EO-HDP, preterm LO-HDP, and term LO-HDP in subjects with MBW < 2,500 g were 1.052 (95 % confidence interval [CI]: 0.665-1.664), 1.745 (95 % CI: 1.220-2.496), and 1.496 (95 % CI: 1.154-1.939), respectively.

CONCLUSIONS

MBW was associated with HDP, regardless of gestational age when HDP developed. Furthermore, the association of MBW < 2,500 g with preterm or term LO-HDP was stronger than that with EO-HDP.

Factors Associated with Progression to Preeclampsia with Severe Features in Pregnancies Complicated by Mild Hypertensive Disorders

In this retrospective cohort study, we aimed to investigate the variables associated with progression to preeclampsia with severe features in parturients already diagnosed with mild hypertensive disorders of pregnancy. The study was conducted in a single university-affiliated medical center between 2018 and 2020. All women admitted due to hypertensive disorders were included. Data collected was compared between parturients who progressed and did not progress to preeclampsia with severe features. Among 359 women presenting without severe features, 18 (5%) developed severe features, delivered smaller babies at lower gestational age, and with higher rates of cesarean delivery (p < 0.001 for all). Chronic hypertension, maternal diabetes, any previous gestational hypertensive disorder, gestational diabetes, number of hospitalizations, earlier gestational age at initial presentation, and superimposed preeclampsia as the preliminary diagnosis were all associated with preeclampsia progression to severe features. Previous delivery within 2-5 years was a protective variable from preeclampsia progression. Following regression analysis and adjustment to confounders, only gestational age at initial presentation and superimposed preeclampsia remained significant variables associated with progression to severe features (aOR 0.74 (0.55-0.96) and 34.44 (1.07-1111.85), aOR (95% CI), respectively, p < 0.05 for both) with combined ROC-AUC prediction performance of 0.89, 95% CI 0.83-0.95, p < 0.001. In conclusion, according to our study results, early gestational age at presentation and superimposed preeclampsia as the preliminary diagnosis are the only independent factors that are associated with progression to severe features in women already diagnosed with mild hypertensive disorders during pregnancy.

Pre-eclampsia

Pre-eclampsia is a life-threatening disease of pregnancy unique to humans and a leading cause of maternal and neonatal morbidity and mortality. Women who survive pre-eclampsia have reduced life expectancy, with increased risks of stroke, cardiovascular disease and diabetes, while babies from a pre-eclamptic pregnancy have increased risks of preterm birth, perinatal death and neurodevelopmental disability and cardiovascular and metabolic disease later in life. Pre-eclampsia is a complex multisystem disease, diagnosed by sudden-onset hypertension (>20 weeks of gestation) and at least one other associated complication, including proteinuria, maternal organ dysfunction or uteroplacental dysfunction. Pre-eclampsia is found only when a placenta is or was recently present and is classified as preterm (delivery <37 weeks of gestation), term (delivery ≥37 weeks of gestation) and postpartum pre-eclampsia. The maternal syndrome of pre-eclampsia is driven by a dysfunctional placenta, which releases factors into maternal blood causing systemic inflammation and widespread maternal endothelial dysfunction. Available treatments target maternal hypertension and seizures, but the only 'cure' for pre-eclampsia is delivery of the dysfunctional placenta and baby, often prematurely. Despite decades of research, the aetiology of pre-eclampsia, particularly of term and postpartum pre-eclampsia, remains poorly defined. Significant advances have been made in the prediction and prevention of preterm pre-eclampsia, which is predicted in early pregnancy through combined screening and is prevented with daily low-dose aspirin, starting before 16 weeks of gestation. By contrast, the prediction of term and postpartum pre-eclampsia is limited and there are no preventive treatments. Future research must investigate the pathogenesis of pre-eclampsia, in particular of term and postpartum pre-eclampsia, and evaluate new prognostic tests and treatments in adequately powered clinical trials.

Comparison of Multivariable Logistic Regression and Other Machine Learning Algorithms for Prognostic Prediction Studies in Pregnancy Care: Systematic Review and Meta-Analysis

BACKGROUND

Predictions in pregnancy care are complex because of interactions among multiple factors. Hence, pregnancy outcomes are not easily predicted by a single predictor using only one algorithm or modeling method.

OBJECTIVE

This study aims to review and compare the predictive performances between logistic regression (LR) and other machine learning algorithms for developing or validating a multivariable prognostic prediction model for pregnancy care to inform clinicians' decision making.

METHODS

Research articles from MEDLINE, Scopus, Web of Science, and Google Scholar were reviewed following several guidelines for a prognostic prediction study, including a risk of bias (ROB) assessment. We report the results based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Studies were primarily framed as PICOTS (population, index, comparator, outcomes, timing, and setting): Population: men or women in procreative management, pregnant women, and fetuses or newborns; Index: multivariable prognostic prediction models using non-LR algorithms for risk classification to inform clinicians' decision making; Comparator: the models applying an LR; Outcomes: pregnancy-related outcomes of procreation or pregnancy outcomes for pregnant women and fetuses or newborns; Timing: pre-, inter-, and peripregnancy periods (predictors), at the pregnancy, delivery, and either puerperal or neonatal period (outcome), and either short- or long-term prognoses (time interval); and Setting: primary care or hospital. The results were synthesized by reporting study characteristics and ROBs and by random effects modeling of the difference of the logit area under the receiver operating characteristic curve of each non-LR model compared with the LR model for the same pregnancy outcomes. We also reported between-study heterogeneity by using τ2 and I2.

RESULTS

Of the 2093 records, we included 142 studies for the systematic review and 62 studies for a meta-analysis. Most prediction models used LR (92/142, 64.8%) and artificial neural networks (20/142, 14.1%) among non-LR algorithms. Only 16.9% (24/142) of studies had a low ROB. A total of 2 non-LR algorithms from low ROB studies significantly outperformed LR. The first algorithm was a random forest for preterm delivery (logit AUROC 2.51, 95% CI 1.49-3.53; I2=86%; τ2=0.77) and pre-eclampsia (logit AUROC 1.2, 95% CI 0.72-1.67; I2=75%; τ2=0.09). The second algorithm was gradient boosting for cesarean section (logit AUROC 2.26, 95% CI 1.39-3.13; I2=75%; τ2=0.43) and gestational diabetes (logit AUROC 1.03, 95% CI 0.69-1.37; I2=83%; τ2=0.07).

CONCLUSIONS

Prediction models with the best performances across studies were not necessarily those that used LR but also used random forest and gradient boosting that also performed well. We recommend a reanalysis of existing LR models for several pregnancy outcomes by comparing them with those algorithms that apply standard guidelines.

TRIAL REGISTRATION

PROSPERO (International Prospective Register of Systematic Reviews) CRD42019136106; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=136106.

Development and Validation of a Nomogram for Predicting the Risk of Pregnancy-Induced Hypertension: A Retrospective Cohort Study

Objective: To develop and validate a prediction model for identifying pregnant women at risk of developing pregnancy-induced hypertension (PIH) to guide treatment decision and classification of management. Methods: This study retrospectively enrolled 907 consecutive pregnant women with de novo hypertension from the Antenatal Care Center of Henan Provincial People's Hospital between June 1, 2018 and May 31, 2019. The cohort was randomly divided into two subgroups: the development cohort (n = 635) and validation cohort (n = 272). Univariate analysis and backward elimination of multivariate logistic regression analyses were utilized to identify predictive factors, and a nomogram was established. The performance was assessed using the area under the curve (AUC), the mean AUC of k-fold cross-validation, and calibration plots. Based on the classification and regression tree model, risk classification was performed. Results: The score included five commonly available predictors: body mass index, proteinuria, age, uric acid, and mean arterial pressure (BPAUM score). When applied to internal validation, the score revealed good discrimination with stratified fivefold cross-validation in the development cohort (AUC = 0.91) and validation cohort (AUC: 0.89) at fixed 10% false-positive rates, and the calibration plots showed good calibration. The total score point was divided into three risk classifications: low risk (0 - 179 points), medium risk (179 - 204 points), and high risk (>204 points). Conclusions: This study established a prediction model for predicting PIH, which could be used in clinical decision-making to improve maternal health and birth outcomes.