Maternal cardiac risk tools show limitations in predicting neonatal outcomes in CHD pregnancies

Pregnant patients with congenital heart disease (CHD) remain at elevated risk for adverse maternal and neonatal outcomes, yet current cardiac risk stratification tools may not adequately identify neonatal risk, according to findings presented at the Pediatric Academic Societies Meeting 2026.^1,2

Investigators from Albert Einstein College of Medicine and Montefiore Medical Center evaluated whether commonly used maternal cardiac risk models could predict neonatal complications among pregnant adults with CHD. The study found that although higher maternal risk scores generally correlated with increased neonatal complications, the predictive performance of the existing tools remained limited.

The poster, titled “Risk of Adverse Neonatal Outcomes in Pregnant Patients with Congenital Heart Disease (CHD),” included a retrospective review of 198 pregnancies managed through the Maternal-Fetal Medicine–Cardiology Joint Program at Montefiore Medical Center/Albert Einstein College of Medicine between January 2015 and December 2024.

Why current maternal cardiac risk models may not predict neonatal outcomes

During a recent interview, study investigator Manoj K. Gupta, MD, pediatric cardiologist at the Children’s Hospital at Montefiore Einstein, explained that current cardiac risk models were originally designed to predict maternal cardiovascular complications rather than neonatal events.

“The most commonly used maternal risk models were built to predict maternal cardiovascular events, so their core variables are heavily weighted towards the lesions that they have, their prior cardiac history, their maternal functional status,” Gupta said.

According to Gupta, the physiologic factors driving maternal cardiac risk do not fully overlap with those associated with fetal and neonatal outcomes.

“These factors do not influence the pregnancy overall, so they only partially overlap with the mechanism that drive the neonatal outcomes,” he said.

The investigators evaluated the predictive performance of 4 established maternal cardiac risk models: modified World Health Organization (mWHO), ZAHARA, CARPREG I, and CARPREG II. Receiver operating characteristic curve analysis demonstrated poor discrimination across all models, with area under the curve values approximating 0.5. The study documented neonatal complications in 65 of 198 pregnancies (32.8%).

Gupta noted that placental function and physiologic changes during pregnancy may contribute substantially to neonatal outcomes but are not adequately reflected in current maternal risk models.

“So, as a result, several gaps emerge to predict neonatal outcomes, which may be attributed to the mismatch in the maternal versus neonatal physiology, the placental function during pregnancy, and these models may not reflect the dynamic physiological changes of pregnancy which are critical in determining the fetal well being,” Gupta said.

Clinical interpretation of higher maternal risk scores

Although statistical discrimination was limited, the investigators observed a general trend linking higher maternal risk scores with increased neonatal complication rates.

Gupta said clinicians should still view elevated maternal risk scores as clinically meaningful indicators requiring closer monitoring and multidisciplinary planning.

“In practice, this means that the maternal risk scores, which should be interpreted as a risk gradients, rather than binary predictors,” Gupta said.

He added that higher-risk patients may warrant enhanced surveillance even if the tools are imperfect.

“So, high risk course is cleaning clinically meaningful, particularly in settings where decisions are made with uncertainty,” Gupta said. “So the practical takeaways use higher maternal risk scores as a prompt for increased vigilance and planning, while recognizing that they are incomplete proxies for neuronal risks and should be complemented by broader clinical judgment.”

Medication use, anticoagulation, and placental health

Secondary analyses identified several significant univariate predictors associated with adverse neonatal outcomes, including pulmonary hypertension or cyanosis, New York Heart Association class greater than II, cardiac medication use during pregnancy, anticoagulation during pregnancy, high-risk valve disease or left heart obstruction, and smoking during pregnancy.

During the interview, Diana S. Wolfe, MD, MPH, FACOG, FACC, obstetric director of Maternal Fetal Medicine Cardiology Joint Program at Montefiore Einstein, emphasized the importance of balancing maternal cardiac management with fetal well-being.

“The majority of medications are safe in pregnancy, despite what most physicians believe, and there are very few medications that we absolutely cannot give during pregnancy,” Wolfe said.

However, Wolfe noted that maternal cardiac disease and medication exposure together can influence placental function and fetal growth.

“The maternal condition, along with medications, can influence that interface and result in what we call adverse pregnancy outcomes,” Wolfe said.

She added that fetal status should remain a central component of clinical assessment throughout pregnancy.

“So while modifying medications and administering medications to optimize maternal conditions, we have to remember that the fetus is another vital sign of what's going on,” Wolfe said.

Future neonatal risk models may incorporate broader predictors

The investigators concluded that future neonatal risk prediction tools should extend beyond traditional maternal cardiac variables.

Wolfe said future models may need to incorporate social determinants of health, maternal comorbidities, and fetal findings.

“Although it's more challenging to integrate this kind of data, you know, we're learning more and more that social determinants of health and maternal mental health are important predictors for outcomes of both the mom the maternal health as well as neonatal health,” Wolfe said.

She also highlighted maternal obesity, gestational diabetes, hypertensive disorders of pregnancy, and fetal growth restriction as potentially important contributors to neonatal risk.

According to Gupta, multidisciplinary cardio-obstetric care models may help bridge the gap between maternal and neonatal risk assessment while supporting individualized delivery planning and neonatal management.

Disclosure

The Gupta and Wolfe report no relevant disclosures.

References

1. Dasika P, Ramineni S, Wolfe DS, et al. Risk of adverse neonatal outcomes in pregnant patients with congenital heart disease (CHD). Poster presented at: Pediatric Academic Societies Meeting 2026; April 24-28, 2026; Boston, MA.

2. Gupta M, Wolfe D. Q&A interview on neonatal outcomes in pregnant patients with congenital heart disease. Interview by Contemporary Pediatrics. Conducted May 6, 2026. Unpublished transcript.