Dense phenotyping from electronic health records enables machine learning-based prediction of preterm birth

Computational Approaches for Predicting Preterm Birth and Newborn Outcomes

Preterm birth (PTB) and its associated morbidities are a leading cause of infant mortality and morbidity. Accurate predictive models and a better biological understanding of PTB-associated morbidities are critical in reducing their adverse effects. Increasing availability of multimodal high-dimensional data sets with concurrent advances in artificial intelligence (AI) have created a rich opportunity to gain novel insights into PTB, a clinically complex and multifactorial disease. Here, the authors review the use of AI to analyze 3 modes of data: electronic health records, biological omics, and social determinants of health metrics.

Data-driven insights can transform women's reproductive health

This perspective explores the transformative potential of data-driven insights to understand and address women's reproductive health conditions. Historically, clinical studies often excluded women, hindering comprehensive research into conditions such as adverse pregnancy outcomes and endometriosis. Recent advances in technology (e.g., next-generation sequencing techniques, electronic medical records (EMRs), computational power) provide unprecedented opportunities for research in women's reproductive health. Studies of molecular data, including large-scale meta-analyses, provide valuable insights into conditions like preterm birth and preeclampsia. Moreover, EMRs and other clinical data sources enable researchers to study populations of individuals, uncovering trends and associations in women's reproductive health conditions. Despite these advancements, challenges such as data completeness, accuracy, and representation persist. We emphasize the importance of holistic approaches, greater inclusion, and refining and expanding on how we leverage data and computational integrative approaches for discoveries so that we can benefit not only women's reproductive health but overall human health.

Development and validation of a risk prediction model for preterm birth in women with gestational diabetes mellitus

OBJECTIVES

This study aims to develop and validate a prediction model for preterm birth in women with gestational diabetes mellitus (GDM).

DESIGN

We conducted a retrospective study on women with GDM who gave birth at the Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, China, between November 2017 and July 2021. We divided 1879 patients into a development set (n = 1346) and a validation set (n = 533). The development set was used to construct the prediction model for preterm birth using the stepwise logistic regression model. A nomogram and a web calculator were established based on the model. Discrimination and calibration were assessed in both sets.

PATIENTS AND MEASUREMENTS

Patients were women with GDM. Data were collected from medical records. GDM was diagnosed with 75-g oral glucose tolerance test during 24-28 gestational weeks. Preterm birth was definied as gestational age at birth <37 weeks.

RESULTS

The incidence of preterm birth was 9.4%. The predictive model included age, assisted reproductive technology, hypertensive disorders of pregnancy, reproductive system inflammation, intrahepatic cholestasis of pregnancy, high-density lipoprotein, homocysteine, and fasting blood glucose of 75-g oral glucose tolerance test. The area under the receiver operating characteristic curve for the development and validation sets was 0.722 and 0.632, respectively. The model has been adequately calibrated using a calibration curve and the Hosmer-Lemeshow test, demonstrating a correlation between the predicted and observed risk.

CONCLUSION

This study presents a novel, validated risk model for preterm birth in pregnant women with GDM, providing an individualized risk estimation using clinical risk factors in the third trimester of pregnancy.

Predicting preterm birth using auto-ML frameworks: a large observational study using electronic inpatient discharge data

Background

To develop and compare different AutoML frameworks and machine learning models to predict premature birth.

Methods

The study used a large electronic medical record database to include 715,962 participants who had the principal diagnosis code of childbirth. Three Automatic Machine Learning (AutoML) were used to construct machine learning models including tree-based models, ensembled models, and deep neural networks on the training sample (N = 536,971). The area under the curve (AUC) and training times were used to assess the performance of the prediction models, and feature importance was computed via permutation-shuffling.

Results

The H2O AutoML framework had the highest median AUC of 0.846, followed by AutoGluon (median AUC: 0.840) and Auto-sklearn (median AUC: 0.820), and the median training time was the lowest for H2O AutoML (0.14 min), followed by AutoGluon (0.16 min) and Auto-sklearn (4.33 min). Among different types of machine learning models, the Gradient Boosting Machines (GBM) or Extreme Gradient Boosting (XGBoost), stacked ensemble, and random forrest models had better predictive performance, with median AUC scores being 0.846, 0.846, and 0.842, respectively. Important features related to preterm birth included premature rupture of membrane (PROM), incompetent cervix, occupation, and preeclampsia.

Conclusions

Our study highlights the potential of machine learning models in predicting the risk of preterm birth using readily available electronic medical record data, which have significant implications for improving prenatal care and outcomes.

Computational Phenotyping of OMOP CDM Normalized EHR for Prenatal and Postpartum Episodes: An Informatics Framework and Clinical Implementation on All of Us

The use of Electronic Health Records (EHR) in pregnancy care and obstetrics-gynecology (OB/GYN) research has increased in recent years. In pregnancy, timing is important because clinical characteristics, risks, and patient management are different in each stage of pregnancy. However, the difficulty of accurately differentiating pregnancy episodes and temporal information of clinical events presents unique challenges for EHR phenotyping. In this work, we introduced the concept of time relativity and proposed a comprehensive framework of computational phenotyping for prenatal and postpartum episodes based on the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). We implemented it on the All of Us national EHR database and identified 6,280 pregnancies with accurate start and end dates among 5,399 female patients. With the ability to identify different episodes in pregnancy care, this framework provides new opportunities for phenotyping complex clinical events and gestational morbidities for pregnant women, thus improving maternal and infant health.

Transforming neonatal care with artificial intelligence: challenges, ethical consideration, and opportunities

Artificial intelligence (AI) offers tremendous potential to transform neonatology through improved diagnostics, personalized treatments, and earlier prevention of complications. However, there are many challenges to address before AI is ready for clinical practice. This review defines key AI concepts and discusses ethical considerations and implicit biases associated with AI. Next we will review literature examples of AI already being explored in neonatology research and we will suggest future potentials for AI work. Examples discussed in this article include predicting outcomes such as sepsis, optimizing oxygen therapy, and image analysis to detect brain injury and retinopathy of prematurity. Realizing AI's potential necessitates collaboration between diverse stakeholders across the entire process of incorporating AI tools in the NICU to address testability, usability, bias, and transparency. With multi-center and multi-disciplinary collaboration, AI holds tremendous potential to transform the future of neonatology.

Sex-stratified phenotyping of comorbidities associated with an inpatient delirium diagnosis using real world data

Delirium is a heterogeneous and detrimental mental condition often seen in older, hospitalized patients and is currently hard to predict. In this study, we leverage large-scale, real- world data using the electronic health records (EHR) to identify two cohorts comprised of 7,492 UCSF patients and 19,417 UC health system patients (excluding UCSF patients) with an inpatient delirium diagnosis and the same number of propensity score-matched control patients without delirium. We found significant associations between comorbidities or laboratory test values and an inpatient delirium diagnosis which were validated independently. Most of these associations were those previously-identified as risk factors for delirium, including metabolic abnormalities, mental health diagnoses, and infections. Some of the associations were sex- specific, including those related to dementia subtypes and infections. We further explored the diagnostic associations with anemia and bipolar disorder by conducting longitudinal analyses from the time of first diagnosis of the risk factor to development of delirium demonstrating a significant relationship across time. Finally, we show that an inpatient delirium diagnosis leads to dramatic increases in mortality outcome across both cohorts. These results demonstrate the powerful application of leveraging EHR data to shed insights into prior diagnoses and laboratory test values that could help predict development of inpatient delirium and emphasize the importance of considering patient demographic characteristics including documented sex when making these assessments.

One Sentence Summary

Longitudinal analysis of electronic health record data reveals associations between inpatient delirium, comorbidities, and mortality.

Multiomic signals associated with maternal epidemiological factors contributing to preterm birth in low- and middle-income countries

Preterm birth (PTB) is the leading cause of death in children under five, yet comprehensive studies are hindered by its multiple complex etiologies. Epidemiological associations between PTB and maternal characteristics have been previously described. This work used multiomic profiling and multivariate modeling to investigate the biological signatures of these characteristics. Maternal covariates were collected during pregnancy from 13,841 pregnant women across five sites. Plasma samples from 231 participants were analyzed to generate proteomic, metabolomic, and lipidomic datasets. Machine learning models showed robust performance for the prediction of PTB (AUROC = 0.70), time-to-delivery (r = 0.65), maternal age (r = 0.59), gravidity (r = 0.56), and BMI (r = 0.81). Time-to-delivery biological correlates included fetal-associated proteins (e.g., ALPP, AFP, and PGF) and immune proteins (e.g., PD-L1, CCL28, and LIFR). Maternal age negatively correlated with collagen COL9A1, gravidity with endothelial NOS and inflammatory chemokine CXCL13, and BMI with leptin and structural protein FABP4. These results provide an integrated view of epidemiological factors associated with PTB and identify biological signatures of clinical covariates affecting this disease.