Key result
AI models may improve maternal cardiovascular risk prediction over CARPREG, ZAHARA, and mWHO scores.
Why the study?
Established pregnancy risk prediction models for women with CHD have important limitations—including variable performance across lesions, reliance on static clinical variables, and limited personalization—while AI applications specifically focused on pregnant women with CHD remain limited.
Does artificial intelligence and machine learning improve pregnancy risk stratification for maternal cardiovascular complications in women with congenital heart disease compared to established risk scores?
Does artificial intelligence and machine learning improve pregnancy risk stratification for maternal cardiovascular complications in women with congenital heart disease compared to established risk scores?
This narrative review highlights the potential of artificial intelligence and machine learning to overcome limitations of traditional risk scores (such as CARPREG, ZAHARA, and mWHO) for personalizing pregnancy risk stratification in women with congenital heart disease.
Answers are grounded in this paper and cite their sources.
Risk stratification guides counseling and surveillance in CHD pregnancies; leaves open validation of contemporary models in prospective cohorts.
Congenital heart disease (CHD) is the most common congenital anomaly worldwide, and advances in pediatric cardiology and cardiac surgery have enabled a growing number of women with CHD to reach reproductive age and pursue pregnancy. While most women with CHD can successfully complete pregnancy, maternal cardiovascular complications and adverse fetal outcomes remain significant concerns. Accurate risk stratification is therefore fundamental to preconception counseling, antenatal surveillance, and delivery planning. Over the past two decades, several validated risk prediction models, including the Cardiac Disease in Pregnancy (CARPREG) score, CARPREG II, the ZAHARA score, and the modified World Health Organization (mWHO) classification, have transformed pregnancy risk assessment in women with heart disease. Nevertheless, these tools have important limitations, including variable performance across different congenital lesions, dependence on static clinical variables, and limited personalization of risk estimates. Recent advances in artificial intelligence (AI) and machine learning (ML) offer opportunities to enhance traditional risk prediction approaches through analysis of complex clinical datasets and identification of nonlinear relationships among risk factors. Emerging evidence from cardio-obstetric research suggests that AI-driven models may improve prediction of maternal cardiovascular complications and facilitate individualized patient management. However, AI applications specifically focused on pregnant women with congenital heart disease remain limited. This narrative review examines current approaches to pregnancy risk stratification in women with congenital heart disease, evaluates the strengths and limitations of established risk prediction models, reviews emerging evidence regarding AI-assisted cardiovascular risk prediction during pregnancy, and discusses future opportunities for integrating artificial intelligence into congenital heart disease pregnancy care.Mn
No takes yet. Share an insight, caveat, or question.
Yash Goyal (2026) conducted a review in Congenital heart disease in pregnancy. Artificial intelligence and machine learning risk prediction models vs. Traditional risk prediction models (CARPREG, CARPREG II, ZAHARA, modified WHO) was evaluated on Maternal cardiovascular complications and pregnancy risk stratification. AI-driven models may improve maternal cardiovascular complication prediction over CARPREG, ZAHARA, and mWHO scores, though CHD-specific pregnancy evidence remains limited.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: