Background

Coarctation of the aorta (CoA) is a rare yet clinically significant congenital heart disease (CHD) that can cause significant morbidity and mortality during the newborn period if not promptly diagnosed and treated. Accurate prenatal detection enables safe perinatal planning and reduces neonatal morbidity and death due to critical CoA. Advances in obstetrical ultrasound screening and fetal echocardiographic techniques have greatly improved detection rates for most CHD. However, prenatal diagnosis and prognostication for critical CoA are challenged by the natural patency of the ductus arteriosus in utero, the sometimes progressive nature of CoA in utero and other changes in aortic and pulmonary blood flow that occur only after birth. Several prior studies examining a number of fetal echocardiographic measurements revealed limited sensitivity and specificity for individual variables [1,2,3,4,5]. Indeed, two recent systematic reviews and metanalyses concluded that the diagnostic accuracy of fetal echocardiographic variables in predicting CoA is only moderate [6, 7]. Multiparameter scoring systems and even artificial-intelligence-assisted automated analysis have been developed to address this gap but they have not yet been validated [8,9,10,11,12]. Consequently, population studies have demonstrated that only 3–37% of CoA is diagnosed prenatally, with 30–54% of newborns with CoA being discharged undiagnosed [13,14,15]. Furthermore, false positive rates of prenatally suspected CoA by fetal echocardiography are estimated at over 50%, with one study reporting a false positive rate of 94% [2, 16].

As an adjunct to fetal echocardiography, the implementation of critical CHD (CCHD) screening for the newborn prior to hospital discharge has helped uncover CHD lesions that were not diagnosed prenatally. Although CCHD screening has effectively improved postnatal diagnosis of many types of CHDs, CoA continues to have a high false negative rate as the ductus arteriosus may not have fully evolved by the time of the CCHD screening. Previous reports confirm that only 20% of significant CoA lesions were diagnosed through this CCHD screening algorithm [17].

The impact of these limitations in the diagnosis and risk-stratification for CoA are clearly seen in the systems for postnatal management. On one hand, the high false positive rates of prenatal coarctation prediction can lead to over-medicalization of otherwise healthy newborns, including parent-child separation, delay of enteral feedings, and unnecessary medical tests and procedures [16, 18]. On the other hand, under-recognition and false negative diagnoses of CoA can result in significant lifelong morbidity and mortality. To address these challenges, our center has recently developed a CoA personalized longitudinal algorithm for newborns (Coarc-PLAN). This pathway is stratified according to the predicted postnatal risk of CoA to minimize unnecessary medicalization of neonates while simultaneously ensuring safety through timely cardiac evaluation before discharge. This study evaluates the ability of this tiered and risk-stratified postnatal management pathway to safely provide the appropriate degree of monitoring and postnatal level of care for fetuses with suspected CoA through interventions and monitoring based on the prenatal risk category.

Methods

Study design

This is a single-center, retrospective cohort study.

Patient selection

All fetuses with concerns for postnatal CoA on their fetal echocardiogram performed at our tertiary care cardiac center between 2016 and 2022 were candidates for our study. Fetuses with isolated coarctation or coarctation with minor intracardiac lesions that may not require surgical intervention, such as small VSD, left-sided superior vena cava, and other minor intra-cardiac findings, were included in the study if they were delivered at our institution and had postnatal information available for review. Those with termination of pregnancy, in-utero fetal demise, and delivery at outside institutions were excluded due to a lack of postnatal data. Additionally, those with hemodynamically significant CHDs were excluded as their postnatal management is driven by more significant intracardiac CHD rather than the management of CoA by itself. The chart of each pregnant patient with fetal concerns for CoA was electronically flagged in the delivery hospital electronic medical record to send an alert upon the patient’s admission to Labor and Delivery. This automated alert was sent in real time to the fetal team and the neonatal ICU team at the delivery hospital. The institutional review board approved this study with a waiver of informed consent.

Concern for developing postnatal coarctation of the aorta was categorized as mild, moderate, or high based on the fetal cardiologist’s interpretation of the fetal echocardiogram according to Table 1. Ultimately, risk categorization was determined at the fetal cardiologist’s discretion. One author (SP) reviewed the fetal echocardiograms and assigned prenatal CoA risk category to reduce interobserver variability. This prenatal risk assessment stratification of mild, moderate, and high concern for critical coarctation then determined postnatal recommendations. The risk categories were reassessed and adjusted based on follow-up fetal echocardiogram findings throughout gestation, with the most recent risk assessment at the time of delivery determining the postnatal recommendations.

Table 1 Fetal echocardiogram classification characteristics for postnatal coarctation concern.
Full size table

The prenatal CoA risk assessment determined the initial postnatal management recommendations as described in Table 2. If any additional or unexpected concerns were present in the delivery room, neonates were managed per standard of care. We defined enhanced observation for the mild concern cohort as (1) daily right upper extremity and right lower extremity blood pressure assessments and (2) brachial and femoral pulse checks twice daily (day shift check performed by rounding pediatrician, evening shift check performed by a NICU provider). The Newborn Nursery unit nursing team had additional training to provide these enhanced observations with the NICU nursing team available for assistance as needed. At our institution, the delivery hospital is adjacent to the children’s hospital and adjoined via a bridge. Thus, for the high concern patients transferred to the CCU, parents are still in close proximity to their child while the mother receives postpartum care.

Table 2 Resource allocation, initial disposition, monitoring, and management protocol for newborns with mild, moderate, or high concerns for postnatal coarctation of aorta based on fetal echocardiography.
Full size table

Timing of the postnatal echocardiogram was determined by risk category as per Table 2, or was obtained sooner for clinical concerns. For newborns with mild or moderate prenatal concern for CoA who stayed at the delivery hospital as per our protocol, their initial postnatal transthoracic echocardiogram was performed by the sonographers on staff at the delivery hospital, who are primarily trained in adult echocardiography but received additional focused training in congenital heart defect echocardiography. This additional training is provided by the children’s hospital echo lab educator via didactic talks focused on congenital heart defects as well as dedicated study review with each sonographer to provide individual feedback. The NICU team was available in-house 24/7 to assess neonates, regardless of unit location, and escalate care as needed. All mild-to-moderate concern patients were evaluated by pediatric cardiology consultation after the initial postnatal echocardiogram to determine recommendations for repeat inpatient postnatal echocardiograms, escalation of care, and timing of discharge. All patients without concerns for hemodynamically significant coarctation needing early surgical repair were discharged to outpatient follow-up with pediatric cardiology with/without repeat outpatient postnatal echocardiogram if there was any aortic arch hypoplasia, patent ductus arteriosus, or other cardiac concern. The newborn delivery, admission location, and the hospital course was monitored by the delivery hospital NICU team until discharge to home or transfer to another hospital as instructed in Table 2. In addition, each newborn clinical course was reviewed by the research team for possible adverse safety events, defined as hemodynamic instability requiring resuscitation or initiation of cardiac medications (except PGE), emergent CICU transfer, emergent readmission, or death due to delayed recognition of clinically significant CoA in the Newborn Nursery unit, NICU, or after discharge home.

Results

There were 139 fetuses with concern for coarctation on their initial fetal echocardiogram during the study period. Of those, 87/139 patients met the inclusion criteria for the final study analysis (Fig. 1). The remaining patients were excluded for the following reasons: pregnancy termination (8), fetal demise (4), other major cardiac anomaly or congenital diaphragmatic hernia (18), delivery at another institution (22).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Study patient outcomes.

For patients in the mild, moderate, and high concern categories, 5%, 62%, and 82% required surgical coarctation repair during the first year of life. Among these, 3%, 38%, and 82% underwent neonatal repair prior to initial hospital discharge from the mild, moderate, and high concern categories. In the mild category, one patient born prematurely had surgery before discharge on the 41st day of life at corrected gestational age 38w5d. For patients in the moderate and high concern groups who had surgery before discharge, the average age of surgery was 5.6 days (n = 5) and 5.9 days (n = 28), respectively. No infants required urgent readmission for management of coarctation of the aorta or surgical repair.

In the mild concern cohort of 40 neonates, 32 (80%) were initially able to be admitted to the Newborn Nursery unit for routine newborn care, with the majority (n = 22/32) remaining in the parent’s room during the entire hospitalization. Ten out of 32 of these patients were later transferred to the NICU for various indications, including (a) more significant concern for coarctation on postnatal imaging or exam (five patients), (b) other cardiac concern (two patients), (c) maternal discharge prior to ductal closure (two patients), and (d) non-cardiac hypoglycemia (one patient). All ten of these patients were discharged from the NICU without CoA repair. One of these patients was readmitted for elective CoA repair at 9 days, prompted by findings on a scheduled follow-up outpatient postnatal echocardiogram.

There were eight mild-concern patients admitted directly to the NICU from the delivery room due to non-coarctation concerns [prematurity (three patients), atrial tachycardia (one patient), respiratory distress (one patient), and other anomalies (three patients: trisomy 18, abdominal wall defect, dural sinus malformation)].

Of the 13 newborns with moderate concern for CoA initially observed in the NICU, six (46%) had postnatal imaging consistent with significant coarctation of aorta, triggering escalation of care to the high concern pathway with umbilical line placement, initiation of PGE infusion, and transfer to the CICU. Five of these six patients underwent repair before discharge. The remaining eight newborns with moderate concern CoA that were discharged without surgery from initial newborn hospitalization were followed by pediatric cardiology as outpatients, and three had elective repair of CoA at a later age, with the average age of surgery of 50 days. Five moderate concern patients (38%) did not require CoA repair.

All 34 high-concern patients were transferred to the children’s hospital after initiation of PGE infusion and umbilical line placement. One high-concern patient died of non-cardiac causes prior to surgery. Five patients (15%) were discharged without surgery and did not require coarctation repair. Only patients transferred to the CICU had enteral feedings delayed and umbilical catheters placed unless another indication was present. There were no adverse safety events noted in entire cohort.

Discussion

Our study cohort demonstrated that the Coarc-PLAN can be safely implemented and reduce unnecessary ICU admission or admission to the inappropriate ICU while potentially reducing overall costs and enhancing critical parental-newborn bonding. Although we did not directly perform a cost analysis, given that the cost of a newborn nursery day is less than the cost of an ICU day, it could be presumed that overall healthcare costs would be lower in those patients that avoided an ICU admission. The safety of this approach for the mild-concern patients was demonstrated by the lack of emergent readmissions or emergent CICU transfers in this group while maintaining a protocol for “coarctation watch” during otherwise routine newborn care in the parent’s room, thus decreasing parental separation, feeding interruption, and medical interventions. As a result, 80% of the mild-concern patients (32/40) remained with their parents after the initial assessment checkpoint and 69% (n = 22/32) remained until discharge. All mild-concern patients needing NICU admission at the initial clinical assessment (8/40) were admitted for non-CoA reasons.

The patients risk stratified to the high concern group were all admitted to the ICU with PGE, umbilical lines, intravenous fluids, and delayed enteral feedings. However, since 15% of high concern patients were discharged without CoA surgery and the 46% of the moderate concern cohort that had escalation to the high concern pathway for clinical concern had no adverse safety events associated with the delay, one could argue that deferring umbilical catheter placement and initiation of PGE until after postnatal admission echocardiogram could be considered as an area of future study. This approach of delaying PGE initiation until after first postnatal echocardiogram is already practiced at other institutions and could likely be safely implemented when the delivery unit is within the children’s hospital or adjoining buildings, as our delivery hospital is, but may be less optimal when the delivery hospital is farther away. (https://www.hopkinsmedicine.org/-/media/files/allchildrens/clinical-pathways/heart---arch-watch-clinical-pathway-final-2025-04-30.pdf).

For moderate concern babies, 38% required CoA repair prior to discharge and 62% had CoA repair in the first year of life. Despite this, there were no adverse safety events or emergencies in this population and escalation of care happened appropriately within the Coarc-PLAN algorithm, although conclusions drawn from this cohort may be limited by small sample size. Prior to universal standardization, there were 2 moderate risk babies who remained with mother with enhanced observation for the first 6–12 h with successful escalation of care as needed. There may be an opportunity to extend duration that patients in the moderate concern cohort remain in the parents’ room while still following the algorithm but replacing telemetry with vital sign checks every 4 h (Table 2). Successful implementation of enhanced observation in this cohort would require sufficient in-hospital support for the recommended monitoring while in the newborn nursery and escalation of care process for clinical or imaging concern that allows for rapid NICU assessment as needed.

This tiered coarctation risk approach minimizes medicalization of low-to-moderate concern newborns until further clinical indications arise in comparison to the high concern group, that were all transferred to the cardiac ICU with umbilical line placement and initiation of PGE. These newborns started enteral feedings without automatic parenteral fluids unless other indication was present. Evidence in preterm babies and emerging literature in term and CHD babies supports the benefits of early enteral feedings and potentially minimizing early fluid initiation in some populations [19,20,21,22]. The low-moderate concern group also had umbilical line placement deferred until postnatal echocardiogram confirmed CoA, thus avoiding the risk of umbilical catheter-associated complications like infection, thrombosis, and cardiac tamponade [23,24,25,26]. As previously discussed, delaying umbilical catheter placement until confirmation of diagnosis could be considered even in the high-concern population.

In addition to decreased medicalization of this at-risk population, there are also broader family and neonatal benefits to a tiered management strategy of infants suspected to have CoA. Early skin-to-skin contact has been associated with improved neonatal stabilization and breastfeeding rates and has been demonstrated to be feasible and safe even in high-risk cardiac patients [27,28,29]. Increasing literature supports the concept of “zero separation” of neonates from their parents, even in those babies requiring NICU-level care [30]. Mothers of preterm babies kept in a couplet with the mother instead of a separate NICU had less maternal stress and improved mental health [30]. While not examined officially in this study, it is assumed that decreased separation and promotion of a family-integrated care model enhances parental satisfaction.

This study informed our clinical practice and allowed for expansion and implementation with partner hospitals that are a farther distance from our children’s hospital. A portion of the mild concern patients was referred back to deliver at the local hospital if postnatal echocardiogram and monitoring capabilities were present there. For these institutions, the fetal team hosts a regular meeting with the receiving institution to review upcoming patients, postnatal recommendations, and subsequent postnatal courses. Delivery at an institution remote from a children’s hospital could be considered for mild concern patients if the institution is capable of providing postnatal monitoring and management as defined in Table 2 as well as escalation of care for any clinical concern for CoA. Additionally, while not expected in the mild concern patients, families should be counseled about the possibility of escalation of care, the risk of neonatal transfer to a children’s hospital, and subsequent parent-baby separation. A thorough and thoughtful analysis of institutional backup systems would be recommended as well as a strong partnership with a fetal care center to help guide this process.

Gestational age at assessment may impact risk stratification, given the possibility of worsening aortic arch hypoplasia with increasing gestational age. Our prenatal protocol of repeat fetal echocardiograms every 4–6 weeks until 36 weeks gestation with reassessment and update of risk stratification minimizes this limitation, assuming term or near-term delivery. However, fetal risk stratification of preterm infants born below a certain gestational age may be unreliable and necessitate closer monitoring of postnatal clinical status, earlier postnatal echocardiograms, and/or serial surveillance until term-corrected gestational age.

Conclusion

The prenatal classification of coarctation severity using fetal echocardiographic indices provides a personalized, family/baby-friendly approach to postnatal management that can decrease invasive interventions while allowing for close monitoring and promoting newborn care. This tiered strategy is successful in babies who are born at or near term without other prenatally identified abnormalities. Those babies born prematurely or with associated complications will require more individualized triage for NICU admission. A postnatal system for enhanced observation and evaluation of neonates with concern for CoA could allow for individualized escalation of care plans as well as avoidance of unnecessary separation and medical interventions, all while maintaining patient safety and avoiding adverse safety events. In delivery centers with adequate capabilities, consideration should be given to a stepwise postnatal evaluation with escalation of care as indicated while supporting the family-baby unit.