FormalPara Key Summary Points

Why carry out this study?

Mothers with congenital heart disease (CHD) have increased risk for having children with CHD.

Fetal echocardiograms (F-echo) are recommended for all pregnancies when maternal CHD is present.

The utility of an F-echo after a normal level II ultrasound (LII-US) in pregnancies with maternal CHD has not been specifically studied.

The aim of this study was to evaluate if any diagnosis of a critical CHD was missed in a fetus with maternal CHD who had a normal LII-US.

What was learned from this study?

Critical CHD was not missed on fetuses when there was a normal level II ultrasound performed for maternal CHD indications.

Larger studies are needed to verify these results, but if consistent, the utility of screening F-echoes in fetuses with maternal CHD should be reevaluated if a normal LII-US has been performed.

Introduction

The risk for recurrent congenital heart disease (CHD) is highest when the mother has CHD versus when the father or sibling has CHD [1, 2]. Risk for subsequent CHD in the offspring is highly dependent on the maternal CHD diagnosis and ranges from < 3% to as high as 18% [1,2,3,4,5,6,7]. Because of this risk, both the American Heart Association and the American Society of Echocardiography recommend a fetal echocardiogram (F-echo) if maternal CHD is present [8, 9].

Current guidelines recommend a level II ultrasound (LII-US) for all high-risk pregnancies at 18–22 weeks gestation [10,11,12]. Recent studies suggest that an F-echo may not add significant clinical benefit if second-trimester screening ultrasound had no cardiac concerns [13,14,15,16,17,18,19,20,21]. No study has specifically evaluated the utility of an F-echo after a normal LII-US in pregnancies with maternal CHD, a population that is at increased risk for CHD.

The primary goal of this study was to evaluate if any diagnosis of a critical CHD was missed in a fetus with maternal CHD who had a normal LII-US. The secondary goal was to determine if any CHD was missed postnatally when a LII-US was read as normal.

Methods

This retrospective chart review study involving human participants was in accordance with the ethical standards of the institutional (Study 00002965) and national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The Human Investigation Committee (IRB) approved this study. The Nationwide Children’s Heart Center database was searched for all F-echoes performed between 1/1/2015 and 12/31/2022 that had the referral indication of maternal CHD.

Patients were included if they had a LII-US that had no cardiac concerns and an F-echo. The LII-US is a detailed or targeted ultrasound examination that is performed when there is an increased risk of a fetal anomaly based on maternal history, abnormal screening results, or abnormal findings on a standard or limited ultrasound examination. LII-US included 2D imaging and color Doppler analysis of the heart per guidelines [12]. All F-echoes were performed per guidelines and included 2D, color Doppler, and pulse/continuous wave Doppler analysis of the cardiac structures [8, 9]. Post-natal transthoracic echocardiograms (pTTE) were reviewed if available. Patients were excluded if they did not have a LII-US, if there was no post-natal data available on the infant, anatomic or functional cardiac concerns on the LII-US, or if the LII-US was performed after the F-echo. Functional cardiac concerns included the presence of a pericardial effusion, ≥ moderate valvular regurgitation, or > moderately decreased cardiac function. This study therefore evaluated only patients that had a normal LII-US with no cardiac concerns.

Critical CHD was defined as any CHD that required catheterization or surgical intervention < 1 month of age. This would essentially include all single ventricle physiologies, transposition of the great arteries, any obstructive lesions such as critical coarctation of the aorta, critical semilunar valvar stenosis, or obstructed pulmonary veins. Though critical CHD has been defined as CHD requiring intervention within the first 24 h of life in the past [22], we felt that a time interval of 30 days was clinically more relevant for the practitioner since ductus closure unmasking critical CHD may occur outside the 24-h period. All other CHD diagnoses were also noted on the F-echo and pTTE when present.

Patients included in the study had their LII-US, F-echo, and pTTE reviewed to determine if there were any discrepancies between the reports. The last pTTE performed on the patient was considered to be the reference echocardiogram in comparison to the LII-US. Gestational age at LII-US and F-echo was recorded. Any interventional cardiac procedures, defined as catheterization or surgical, were recorded for any patient noted to have CHD. The overall outcome for any patient diagnosed with a CHD was also detailed.

All data are presented as means and standard deviations unless otherwise stated. The negative predictive value of a normal LII-US for critical CHD and any type of CHD was calculated.

Results

During this time period, a total of 296 fetuses were referred for an F-echo due maternal CHD. A total of 121 fetuses were excluded (no LII-US, n = 89; no post-natal data available, n = 23; cardiac concerns on the LII-US, n = 5, LII-US performed after the F-echo, n = 4). Thus, 175 patients were included in this study. Maternal CHD diagnosis was heterogenous, with most of the diagnosis consisting of septal or valvular abnormalities (Table 1).

Table 1 Maternal diagnosis
Full size table

LII-US was performed at 19.8 ± 2.9 weeks gestational age. F-echo was performed at 24.2 ± 2.8 weeks gestational age. No patient with a normal LII-US had a diagnosis of a critical CHD by F-echo (n = 175, negative predictive value = 100%). A total of 33 fetuses had a pTTE performed.

Postnatally, ten patients were diagnosed with a CHD (negative predictive value = 94.3%). The majority of defects missed were valvular or septal defects (Table 2). F-echo correctly diagnosed two of the ten missed LII-US CHD (mild aortic root dilation, n = 1; vascular ring, n = 1). F-echo had nine false-positive results compared to the pTTE (ventricular septal defect, n = 4; coarctation, n = 3; mild mitral regurgitation, n = 1, dilated ascending aorta, n = 1). F-echo for fetuses with maternal CHD and a prior normal LII-US had a sensitivity of 20.0%, specificity of 94.5%, positive predictive value of 18.2%, negative predictive value of 95.1%, and accuracy of 90.3% for a correct pTTE diagnosis.

Table 2 Outcomes
Full size table

Clinically, all the patients are doing well with either a minor cardiac defect that is being clinically followed or has resolved (n = 9) or had catheterization or surgical intervention at an appropriate time interval (n = 1). This latter patient had a genetic proven diagnosis of Noonan syndrome (Table 2). There were no mortalities recorded.

Discussion

Current guidelines recommend an F-echo for mothers with CHD. The guidelines do not take into account if a LII-US was felt to be normal [8, 9]. There is increasing data suggesting that a normal second-trimester screening ultrasound is unlikely to miss critical CHD [13,14,15,16,17,18,19,20,21]. There are minimal data evaluating the utility of F-echo in mothers with CHD when a LII-US read was normal. In this study, a normal LII-US did not miss any critical CHD. LII-US did miss mainly septal defects and minor valve issues, but this was also true of the F-echo.

These findings are consistent with multiple recent studies documenting the accuracy of second-trimester screening ultrasounds when compared to F-echo in diagnosing CHD and most importantly critical CHD [13,14,15,16,17,18,19,20,21]. These previous studies have noted an approximately 2–3% missed CHD diagnosis rate on the second-trimester screening ultrasound when compared to an F-echo in the current era [13, 18]. The possibility of a missed critical CHD when there was a prior normal second-trimester screening ultrasound was < 0.5% in large population studies [13, 19, 20]. The estimated number of F-echo needed to be performed to detect a missed critical CHD ranged from 249 to 750 [13, 20].

Despite the fact that no critical CHD was missed on LII-US, there still remained patients that were diagnosed with CHD in the postnatal period. The sensitivity for F-echo may range as low as 20% to as high as 80% for revealing minor CHD [23,24,25]. Most of the missed CHD in this study dealt with septal defects or minor valvular abnormalities, which is consistent with previous F-echo studies [23,24,25,26]. There are also certain types of CHD that are difficult or impossible to diagnose on F-echo such as a patent ductus arteriosus, coarctation of the aorta, and septal defects. These are known limitations to F-echo.

We acknowledge that the purpose of F-echo is not only for diagnosis, but also for subsequent counseling of families when a CHD is diagnosed. In this study, two CHD diagnoses were correctly made that the LII-US missed, a vascular ring and mild aortic root dilation. However, these prenatal diagnoses did not alter the overall postnatal cardiac care. Both patients are asymptomatic and are being followed conservatively. Conversely, this must be weighed against the reassurance of a false-negative F-echo or the anxiety that may be induced for a false-positive F-echo report. We concede that the vascular ring may never have been diagnosed postnatally, unless the patient developed symptoms. Treatment options are variable for asymptomatic vascular rings with nonintervention being one possibility, as is the case for this patient [27,28,29,30]. In addition, the patient with aortic root dilation would have had a pTTE regardless of the F-echo because of the family history per guidelines, so this did not affect overall clinical management [31].

Six patients had false-positive results on the F-echo, five patients in this study had normal F-echo, but were diagnosed with CHD in the postnatal period, and three patients had CHD concerns on the F-echo, but the pTTE diagnosed a different CHD. Of note, the referral indication was a murmur for all five of the patients with a normal F-echo. False-positive results are a known limitation of F-echo [16]. Increased maternal anxiety due to a fetal diagnosis of CHD is well reported and thus a false-positive F-echo would be a negative experience for the family [32,33,34,35].

Another factor to consider whenever ordering additional ultrasound testing is bioeffects and safety. Even though ultrasound emits minimal bioeffects, all ultrasound studies should follow the “as low as reasonably achievable (ALARA)” principle [10]. This principle affects both evaluating imaging modality and what exams are absolutely necessary but also length of scans. If CHD can be adequately ruled out with a LII-US, then by following ALARA, an F-echo might not be deemed necessary.

Due to these issues, an alternative strategy may be to reassure the family that it is extremely unlikely that a critical CHD is present if the LII-US was normal. There is no question that if there are cardiac concerns or even uncertainty on the LII-US, an F-echo should be performed [8, 9]. If there are cardiac concerns after the fetus is born, then a pTTE may be performed to reassess the situation [36]. This occurred in the five patients that had a normal F-echo, but subsequently had a CHD diagnosis postnatally. This possible change in strategy is also reflected in the slight changes present from the initial F-echo guidelines published in 2014 versus the updated guidelines published in 2023 [8, 9]. Some previous indications that were listed as may be indicated or probably indicated have been changed to not indicated or may be indicated, respectively. As stated by the most recent guidelines, “Some conclude that a well-executed obstetric scan is cost-effective compared with referring these higher risk pregnant patients directly for fetal echocardiography. However, this conclusion depends heavily on the quality of the screening ultrasound and its interpretation, and the cost–benefit ratio of screening is related to the sensitivity, specificity, and posttest probability as described earlier.” [9]. As technology and comfort level with evaluating cardiac issues improves over time in the obstetric and maternal–fetal field, one would assume the LII-US and possibly the regular second-trimester screening ultrasound will continue to improve in detecting CHD [18].

There are multiple limitations of this study. This was a retrospective study with all the inherent shortcomings of such a design. The study assessed a single center, and the overall numbers were thus small, despite evaluating an 8-year time period. Maternal CHD diagnosis was heterogeneous. Twenty-three patients were unaccounted for postnatally. Of those patients, 22 had normal F-echo and one had concerns for a small muscular ventricular septal defect, so it is extremely unlikely critical CHD was missed, though it is possible that other minor CHD diagnoses were overlooked. LII-US were from a variety of clinics, but all located within a major metropolitan area and thus the generalizability of these results to other settings needs to be evaluated. The findings are all based on LII-US reports since access to the images were unable to be obtained, hence this study could not delineate if LII-US accuracy was due to image quality or interpretation. Similarly, body mass indexes were not available for the mothers, so no comment can be made if diagnoses were possibly missed due to difficult images due to maternal body habitus. The educational degree and experience level of the sonographer or physician performing the LII-US is unknown, thus no comment can be made if better or worse detection of CHD would have occurred based on these factors. This study did not evaluate the accuracy of an LII-US if a cardiac concern was present since this was not the purpose of this study. This study did not evaluate the accuracy of a standard second-trimester obstetric ultrasound that is performed for non-high-risk pregnancies so no comment can be made in that regard.

Conclusions

Critical CHD was not missed with a normal LII-US in this at-risk population. F-echo also missed the majority of CHD when an LII-US was read as normal, however the majority of these were minor types of CHD. The incremental cost–benefit ratio, financially and emotionally, of screening F-echo in fetuses with maternal CHD should be further evaluated if a normal LII-US has been performed. Larger studies are needed to determine if these findings remain consistent.