Impact of COVID‐19 on Prenatal Diagnosis and Surgical Outcomes of Congenital Heart Disease: Fetal Heart Society and Society of Thoracic Surgeons Collaborative Study

Abstract

Background

Fetal echocardiography is the mainstay of prenatal diagnosis of congenital heart disease. The COVID‐19 pandemic led to shifts in triage of prenatal services. Our objective was to evaluate the impact of COVID‐19 restrictions on prenatal diagnosis, surgical outcomes, and disparities in neonatal critical congenital heart disease (CCHD) management in the United States during the pandemic's first year.

Methods and Results

A multi‐institutional retrospective cohort study compared neonatal CCHD outcomes (requiring surgery within 60 days of birth) 1 year prior (prepandemic era) and during the peak pandemic era, supplemented by a Fetal Heart Society survey assessing regional practice changes. Data on prenatal diagnosis, demographics, outcomes, and 2020 state Area Deprivation Index were analyzed using Wilcoxon rank sum and χ² tests. The survey, completed by 72 fetal cardiologists from 9 US census regions, showed 75% of institutions implemented restrictions by March 2020, affecting triage, referrals, and number of prenatal cardiology visits. Compared with CCHD neonates born prepandemic (n=4637), those born during the pandemic (n=1806) had a higher proportion of prenatal diagnosis (66% versus 63%, P<0.05). There were no significant differences in complications or mortality, but pandemic‐era neonates had longer hospital stays. During the pandemic, CCHD neonates had a more disadvantaged Area Deprivation Index and had surgery at hospitals located in more advantaged regions.

Conclusions

Although pandemic‐driven care delivery adjustments affected perinatal cardiology referrals and triage, prenatal diagnosis, perioperative outcomes, and survival remained robust. The management of CCHD demonstrates health care resilience, maintaining core prenatal and perioperative care. Regional variations highlight the need for targeted strategies to address disparities during health care crises.

Nonstandard Abbreviations and Acronyms

ADI

Area Deprivation Index

CCHD

critical congenital heart disease

d‐TGA

D‐transposition of the great arteries

Clinical Perspective.

What Is New?

• This is the first study to examine the COVID‐19 pandemic's impact on prenatal critical congenital heart disease diagnosis in the United States, with unexpected stability in diagnosis rates despite health care disruptions during the pandemic.

• This stability is attributed to adaptive strategies, including the expansion of telehealth, which may become a permanent feature of medical practice. CCHD detection in disadvantaged socioeconomic populations was not adversely affected during the pandemic despite regional differences in triage and access to prenatal care.

What Are the Clinical Implications?

• The study demonstrates the importance of flexible, equitable health care delivery models and emphasizes addressing systemic inequities for future crisis preparedness ensuring high‐quality care for all.

In the United States, the standard of care for prenatal diagnosis of congenital heart disease (CHD) involves a detailed fetal echocardiogram and consultation with a fetal cardiologist. Accurate prenatal identification of CHD enhances medical team coordination and parental readiness for delivery and neonatal management. ¹ , ² However, the onset of the COVID‐19 pandemic significantly strained this system. The pandemic's impact led to a shift in triage of prenatal services, including fetal cardiac evaluations, with clinics and hospitals across the United States postponing or canceling appointments. Despite the rapid dissemination of triage and clinical management guidelines specific to the pandemic, ¹ , ³ variability existed in how centers nationwide adapted prenatal diagnosis and counseling practices for CHD. ⁴ , ⁵ , ⁶ Such variability likely occurred due to regional differences in COVID‐19 outbreak severity, local responses, and available resources for implementing safety measures like social distancing and telehealth services.

The implications of ad hoc triage practices and the effect on prenatal CHD screening are not fully understood. A potential shift in obstetric screening practices could delay the identification of critical CHD (CCHD) necessitating cardiac interventions or surgery within the first 2 months of life. This delay could significantly affect delivery planning and perinatal management, potentially resulting in adverse outcomes for neonates receiving delayed care. Moreover, absence of early and accurate prenatal diagnosis of CHD eliminates informed pregnancy decision‐making and diminishes parental preparedness for managing a child with CHD.

Future crisis preparation requires investigating response and triage decisions made during the pandemic and their impact on timely CCHD diagnosis and management. Data evaluating triage strategy effectiveness provide critical information for developing best practices under extraordinary circumstances. This study aimed to explore the effects of COVID‐19‐induced changes in maternal care access on the prenatal diagnosis of CCHD and the implications for neonatal outcomes, focusing on cases requiring surgical intervention. Our hypotheses include:

1. Regional policy changes during the pandemic led to alterations in triage processes and referrals for prenatal cardiology evaluation.

2. Implementation of pandemic‐related triage strategies would be associated with a reduction of prenatal CHD diagnoses compared with the year before the pandemic, with a greater decrease in diagnoses in areas with increased social deprivation and those farther from tertiary care centers.

3. Decreased prenatal diagnosis of CCHD during the study period would be associated with worse surgical outcomes.

METHODS

Patients and Data Collection

This study was approved by the Seattle Children's Institutional Research Board (study 00003593) on March 23, 2022. Because of the sensitive nature of the data collected for this study, requests to access the data set from qualified researchers trained in human subject confidentiality protocols may be sent to the Society of Thoracic Surgeons Participant User File Research Program (https://www.sts.org/PUFApplication). Survey data are available from the corresponding author upon reasonable request. This research was determined to be exempt research with a waiver of informed consent from the Advarra Institutional Review Board (Mod01760092, version 1.1, approval date July 17, 2023).

Pandemic‐Era Practice Changes Survey

Data on regional practice changes during the pandemic were collected through a prospective, cross‐sectional survey conducted by the Fetal Heart Society. This survey, distributed via the Fetal Heart Society listserv from December 1, 2022 to September 1, 2023, gathered information on clinical practices, timing of regional COVID‐19 restrictions, specific policy implementations, fluctuations in fetal cardiology triage strategies, and referral and visit volumes (Data S1). Respondents (Data S2) providing divergent answers within the same practice were contacted together and were asked to consolidate their responses to ensure consistency in data reported. Using the Society of Thoracic Surgeons regionalization scheme, ⁷ the surveys were then categorized by region.

Retrospective Cohort Study

We conducted a retrospective cohort study using data from the Society of Thoracic Surgeons– Congenital Heart Surgery Database (STS‐CHSD). The STS‐CHSD contains information on patients who have undergone surgical procedures submitted by participating pediatric cardiothoracic surgical programs in the United States. Data collection is focused on specific preoperative factors, intraoperative and postoperative variables, and outcomes; no prenatal data are available in the database, except for a designation that specifies whether a prenatal diagnosis was available. This database excludes patients who have catheter‐based interventions and those who have no cardiothoracic surgeries. Protected health information such as date of birth, patient address, and hospital names are not available for multicenter studies executed under the participant user file mechanism. For this study, the quartile of birth and surgery were provided by the STS‐CCHD.

This study compares patients born with CCHD requiring surgery or intervention by 60 days of age at 2 time periods; the control group included patients born in the prepandemic period (February 1, 2019 to February 1, 2020), and the study group included patients born during the peak of the COVID‐19 pandemic (May 1, 2020 to November 1, 2020). Surgical and outcomes data for patients born during these 2 periods were available through December 31, 2021. The pandemic‐era cohort included neonates with presumed prenatal diagnosis after the March 13, 2020 pandemic declaration and expected to deliver from May 20, 2020 onward (assuming prenatal diagnosis before 30 weeks gestation). We excluded patients with CCHD born from February 1 2020 to May 19, 2020, in which the timing of prenatal diagnosis, prepandemic versus pandemic era, would be difficult to ascertain due to variability in onset of COVID‐19 policy responses. The cutoff date of November 1, 2020 was selected to represent the period of cessation of emergency triaging ⁸ in most areas (Figure 1).

Figure 1. Timeline delineating patients with CCHD born in the prepandemic era (control) versus pandemic era (study) using the STS‐CHSD.

The area from February 1, 2020 to May 1, 2020 depicts a period during which patients were not included, because the timing of prenatal diagnosis could have fallen in the prepandemic or pandemic era (assuming gestational age at diagnosis <30 weeks). CCHD indicates critical congenital heart disease; and STS‐CHSD, Society of Thoracic Surgeons–Congenital Heart Surgery Database.

The STS‐CHSD provided demographic and clinical data. Primary CHD diagnoses and surgeries were subsequently grouped to prevent covariance. For instance, D‐transposition of the great arteries (d‐TGA) diagnosis included neonates with/without ventricular septal defect and with/without outflow tract obstruction.

A state and national Area Deprivation Index (ADI) for 2020 ⁹ , ¹⁰ was calculated for patient and hospital zip code. The ADI is a measure of neighborhood‐disadvantage that could influence health care outcomes. It is a socioeconomic demographic statistical measure that identifies high deprivation areas and includes income, employment, education, household characteristics, and housing type. ¹⁰

We also collected data on primary outcomes including hospital discharge mortality, hospital length of stay, preoperative risk factors (preoperative cardiopulmonary resuscitation, mechanical circulatory support, mechanical ventilation, cardiogenic shock, renal dysfunction, dialysis, mechanical ventilation, neurologic deficit, stroke, intracranial hemorrhage, seizure), and postoperative complications (postoperative cardiac arrest, reintubation, tracheostomy, dialysis, neurological deficit). Exclusion criteria included missing data in the antenatal diagnosis field in the STS‐CHSD and surgeries performed in 2021 quarter 3 and 4, which represent surgeries beyond 60 days of age for patients born by November 1, 2020. Additional subgroup analyses by US region (Figure 2) were performed on both the national survey and STS‐CHSD data.

Figure 2. Map depicting the 9 US regions as delineated by the STS‐CHSD along with patients and survey respondents with the following data by region: number of CCHD born prepandemic era/control; CCHD born pandemic era/study (N).

CCHD indicates critical congenital heart disease; FHS, Fetal Heart Society; and STS‐CHSD, Society of Thoracic Surgeons–Congenital Heart Surgery Database.

Statistical Analysis

Statistical analyses were conducted using SAS software, version 9.4 (SAS Institute, Cary, NC). Continuous variables were summarized using mean±SD or median (15th–85th percentile) for skewed distributions. The Wilcoxon rank sum test was applied for comparisons of continuous variables. Categorical data were summarized by frequencies and percentages and compared using the χ² test. Length of stay was compared via Poisson regression, with an offset term to adjust for the different exposure time frames. Statistical significance was set at P<0.05. Due to the CCHD diagnosis and surgical groupings described above to limit covariance, multivariable analysis was not performed.

RESULTS

Survey Responses and Practice Changes

Seventy‐two US fetal cardiologists completed the Fetal Heart Society survey, representing all 9 US regions. Table 1 summarizes survey responses. By March 20, 2020, 56 institutions (79%) altered their fetal cardiology triage strategies due to significant pandemic restrictions. Subsequently, restrictive institutional policies lessened by November 1, 2020 in 30 institutions (43%). A significant proportion of respondents (69%) agreed that the pandemic led to triage changes, resulting in decreased referrals and visits. Moreover, 73% reported adherence to recommended triage guidelines ³ during this period. A notable reduction in clinic visits occurred, with 64 respondents (89%) indicating decreases (Figure 3). Forty‐four percent of respondents adopted telemedicine consultations during the pandemic; fetal cardiac diagnoses seen by telemedicine ranged from normal cardiac anatomy to complex CHD. Postrestriction, 58 respondents (81%) noted a return to prepandemic visit volumes, though only 36% observed a similar rebound in referrals.

Table 1.

Impact of Pandemic Restrictions on Prenatal Cardiology Practices by Regions

Survey question	Response	East South Central, n=3, n (%)	East North Central, n=18, n (%)	West South Central, n=4, n (%)	Middle Atlantic, n=6, n (%)	Mountain, n=7, n (%)	New England, n=6, n (%)	Pacific, n=15, n (%)	South Atlantic, n=9, n (%)	West South Central, n=5, n (%)
Local restrictions resulted in changes in triage of prenatal cardiology visits.	Strongly disagree	0 (0)	2 (11)	0 (0)	0 (0)	3 (43)	0 (0)	1 (7.1)	0 (0)	0 (0)
	Disagree	0 (0)	1 (5.6)	1 (25)	0 (0)	0 (0)	0 (0)	4 (29)	1 (11)	0 (0)
	Neutral	0 (0)	4 (22)	0 (0)	1 (17)	1 (14)	0 (0)	0 (0)	1 (11)	2 (40)
	Agree	2 (67)	6 (33)	3 (75)	4 (67)	3 (43)	4 (67)	6 (43)	2 (22)	1 (20)
	Strongly agree	1 (33)	5 (28)	0 (0)	1 (17)	0 (0)	2 (33)	3 (21)	5 (56)	2 (40)
Local triage changes: followed guidelines.	Yes	3 (100)	15 (83)	4 (100)	6 (100)	3 (43)	2 (33)	9 (60)	8 (89)	3 (60)
Local restrictions resulted in decrease in fetal cardiology referrals.	Strongly disagree	0 (0)	0 (0)	0 (0)	0 (0)	3 (43)	1 (17)	0 (0)	1 (11)	0 (0)
	Disagree	0 (0)	4 (22)	0 (0)	1 (17)	2 (29)	1 (17)	6 (43)	3 (33)	3 (60)
	Neutral	2 (67)	3 (17)	0 (0)	1 (17)	0 (0)	0 (0)	1 (7.1)	2 (22)	1 (20)
	Agree	1 (33)	8 (44)	4 (100)	3 (50)	1 (14)	2 (33)	5 (36)	2 (22)	0 (0)
	Strongly agree	0 (0)	3 (17)	0 (0)	1 (17)	1 (14)	2 (33)	2 (14)	1 (11)	1 (20)
Did restrictions impact prenatal management and counseling?	Yes	2 (67)	5 (28)	3 (75)	1 (17)	2 (29)	2 (33)	5 (36)	2 (22)	3 (60)
Did restrictions impact delivery of fetuses with cardiac abnormalities?	Yes	0 (0)	3 (17)	2 (50)	1 (17)	0 (0)	0 (0)	1 (7.1)	1 (11)	1 (20)
Was telemedicine used for fetal cardiology visits?	Yes	2 (67)	8 (44)	2 (50)	1 (17)	1 (14)	3 (50)	7 (50)	4 (44)	4 (80)
When reopening processes initiated, fetal cardiology triage returned to pre‐COVID standards.	Strongly disagree	0 (0)	0 (0)	0 (0)	1 (17)	1 (14)	0 (0)	0 (0)	1 (11)	0 (0)
	Disagree	0 (0)	1 (5.6)	0 (0)	0 (0)	0 (0)	1 (17)	1 (7.1)	2 (22)	2 (40)
	Neutral	1 (33)	3 (17)	0 (0)	1 (17)	2 (29)	1 (17)	3 (21)	2 (22)	0 (0)
	Agree	0 (0)	10 (56)	3 (75)	4 (67)	2 (29)	0 (0)	6 (43)	1 (11)	2 (40)
	Strongly agree	2 (67)	4 (22)	1 (25)	0 (0)	2 (29)	4 (67)	4 (29)	3 (33)	1 (20)
Referrals improved with reopening processes.	Strongly disagree	0 (0)	0 (0)	0 (0)	0 (0)	1 (14)	0 (0)	0 (0)	0 (0)	0 (0)
	Disagree	1 (33)	0 (0)	0 (0)	0 (0)	0 (0)	1 (17)	2 (14)	1 (11)	0 (0)
	Neutral	0 (0)	3 (17)	0 (0)	2 (33)	2 (29)	0 (0)	1 (7.1)	6 (67)	1 (20)
	Agree	0 (0)	9 (50)	3 (75)	3 (50)	1 (14)	2 (33)	6 (43)	1 (11)	3 (60)
	Strongly agree	2 (67)	6 (33)	1 (25)	1 (17)	3 (43)	3 (50)	5 (36)	1 (11)	1 (20)
Visits improved with reopening processes.	Strongly disagree	0 (0)	0 (0)	0 (0)	0 (0)	1 (14)	0 (0)	0 (0)	1 (11)	0 (0)
	Disagree	1 (33)	0 (0)	0 (0)	0 (0)	0 (0)	1 (17)	2 (14)	0 (0)	0 (0)
	Neutral	0 (0)	2 (11)	0 (0)	2 (33)	2 (29)	0 (0)	1 (7.1)	1 (11)	0 (0)
	Agree	0 (0)	11 (61)	2 (50)	3 (50)	1 (14)	2 (33)	5 (36)	6 (67)	3 (60)
	Strongly agree	2 (67)	5 (28)	2 (50)	1 (17)	3 (43)	3 (50)	6 (43)	1 (11)	2 (40)

Percentages indicate the degree of agreement with the statement posed in the question. The n (%) indicates the number (percentage) who responded as to whether the practice was adopted or impacted as described.

Figure 3. Summary of FHS fetal cardiology survey nationally by quarter of the calendar year.

Bar plots depicting trends in prenatal cardiology referral and visit changes during 2020 to 2021. FHS indicates Fetal Heart Society; NA, not applicable; and Q, quarter.

Comparative Analysis of Prepandemic and Pandemic Era Births

Table 2 includes a descriptive comparison of the STS‐CHSD control (N=4637) and study group (N=1806); prenatal diagnosis status was unknown in 119 (1.9%) subjects who were therefore excluded. Table 3 shows the diagnoses, surgeries, and outcomes for these 2 groups. Analysis of STS‐CHSD data revealed a comparative increase in prenatal diagnoses during the pandemic compared with prepandemic (N=1184, 63% versus N=2915, 66%; P=0.043). Although overall diagnosis of CCHD was stable, there was variability in specific diagnoses, namely a decreased proportion of d‐TGA (13% versus 11%; P=0.01) during the pandemic. Although extracardiac and genetic abnormalities were similar between groups, the diagnosis of 22q11.2 deletion was increased in the study group (3% versus 4%; P=0.006). We demonstrated increased use of hybrid stage 1 for left heart hypoplasia (0.95% versus 1.6%; P=0.04) in the study group. No difference in Norwood procedures between groups existed.

Table 2.

STS‐CHSD Characteristics of Patients Born in Prepandemic (Control) and Pandemic Era (Study)

Characteristic	Prepandemic era (n=4637)	Pandemic era (n=1806)	P value
	N (%) or median (15th–85th percentile)
Demographics
Prenatal diagnosis	2915 (63)	1184 (66)	0.043
Sex (female)	2020 (44)	767 (42)	0.43
Premature birth (<37 wk gestation)	1037 (22)	430 (24)	0.081
Birth weight (kg)	3.04 (2.17–3.64)	3.05 (2.09–3.67)	0.68
Weight at surgery (kg)	3.5 (2.59–6.4)	3.48 (2.5–6.15)	0.32
No preoperative risk factors*	1885 (41)	698 (39)	0.14
STAT mortality score	0.4 (0.2–1.3)	0.4 (0.2–1.4)	0.31
Race
Asian	223 (5.1)	91 (5.4)	0.64
Black	761 (17)	289 (17)	0.79
Caucasian	3065 (70)	1148 (68)	0.14
American Indian/Alaskan Native	83 (1.9)	55 (3.3)	0.0015
Native Hawaiian/Pacific Islander	60 (1.4)	33 (2)	0.098
Other	458 (10)	212 (13)	0.021
Ethnicity
Hispanic and Latino	976 (21)	400 (22)	0.48
Geographic/socioeconomics
Distance from patient to hospital zip code (miles)	7.25/29.7/119	7.21/31.8/118	0.69
Patient median state ADI ranks for 2020†	2.5/5.5/9	3/6/9	0.017
Patient median national ADI ranks for 2020‡	25.5/58/83.5	28/59/84.3	0.016

ADI indicates Area Deprivation Index; STAT mortality score: tool to analyze mortality risk associated with congenital heart surgery and ranges from Category 1 to Category 5; and STS‐CHSD, Society of Thoracic Surgeons–Congenital Heart Surgery Database.

^*Areas with greater socioeconomic disadvantages are ranked higher.

^† Median state ADI rank: ranks the neighborhood compared with other neighborhoods in the same state.

^‡ Median national ADI rank: ranks the neighborhood compared with other neighborhoods in the same country.

Table 3.

STS‐CHSD Diagnoses and Primary Surgeries Patients Born in Prepandemic (Control) and Pandemic Era (Study)

Characteristic	Prepandemic era (n=4637)	Pandemic era (n=1806)	P value
	N (%) or median (15th–85th percentile)
Primary congenital heart disease
Aortic arch hypoplasia	883 (19)	308 (17)	0.062
Aortopulmonary window	28 (0.61)	14 (0.78)	0.44
Atrial septal defect	3 (0.065)	5 (0.28)	0.03
Atrioventricular septal defect	112 (2.4)	69 (3.8)	0.0022
Corrected transposition of the great arteries	45 (0.97)	21 (1.2)	0.49
Ebstein/tricuspid valve anomaly	21 (0.45)	16 (0.89)	0.039
Hypoplastic left heart syndrome	648 (15)	282 (16)	0.39
Mitral valve abnormality	7 (0.15)	5 (0.28)	0.29
Interrupted aortic arch	4 (0.087)	3 (0.17)	0.38
Single ventricle (not hypoplastic left heart syndrome)	543 (12)	214 (12)	0.89
Pulmonary atresia/intact ventricular septum	77 (1.7)	27 (1.5)	0.63
Pulmonary atresia/ventricular septal defect	168 (3.6)	54 (3)	0.21
Pulmonary atresia/ventricular septal defect/multiple aortopulmonary collaterals	16 (0.35)	2 (0.11)	0.11
D‐transposition of the great arteries	620 (13)	199 (11)	0.01
Total anomalous pulmonary venous connection	197 (4.3)	79 (4.4)	0.83
Tetralogy of Fallot	510 (11)	190 (11)	0.57
Truncus arteriosus	108 (2.3)	51 (2.8)	0.25
Other	647 (13.9)	267 (14.8)	0.10
Genetic abnormalities
22q11.2 deletion	141 (3)	80 (4.4)	0.0059
Trisomy 21	149 (3.2)	60 (3.3)	0.82
Trisomy 18	11 (0.24)	6 (0.33)	0.5
Trisomy 13	5 (0.11)	2 (0.11)	0.97
Heterotaxy syndrome	43 (0.93)	11 (0.61)	0.21
Primary cardiac surgical repair
Aortic arch repair	849 (18)	313 (17)	0.36
Arterial switch operation	500 (11)	170 (9.4)	0.11
Stage 1 hybrid procedure for hypoplastic left heart syndrome	44 (0.95)	28 (1.6)	0.039
Norwood procedure	436 (9.4)	175 (9.5)	0.72
Pulmonary artery banding	305 (6.6)	133 (7.4)	0.26
Aortopulmonary shunt	316 (6.8)	108 (6)	0.22
Tetralogy of Fallot repair	313 (6.8)	127 (7)	0.69
Truncus arteriosus repair	103 (2.2)	45 (2.5)	0.52
Atrioventricular septal defect repair	0	0
Outcomes
Mortality or major postoperative complication†	839 (22)	341 (22)	0.7
Mortality at hospital discharge	235 (5.1)	91 (5)	0.96
Total length of stay (d)	24 (8–83)	26 (9–91)	0.0008
Postoperative length of stay (d)	16 (6–66)	18 (7–73)	0.0037
Readmission within 30 d	630 (15)	229 (14)	0.5

STS‐CHSD indicates Society of Thoracic Surgeons–Congenital Heart Surgery Database.

^† Postoperative cardiac arrest, reintubation, tracheostomy, dialysis, and neurological deficit, and hospital length of stay.

Premature delivery, preoperative risk factors, and postoperative complications were similar between the 2 groups. CCHD neonates were from more disadvantaged (higher) ADI neighborhoods during the pandemic era (P=0.017). The Society of Thoracic Surgeons database identified 69 unique hospitals performing CCHD surgery, and there was no difference in hospitals between the control and study cohort. No hospitals halted surgery, and no new hospitals initiated surgical care during the pandemic. The American Indian/Alaska Native population and those identified as other race experienced increased prenatal CCHD diagnoses during the pandemic (P=0.002 and 0.021, respectively). No differences by distance from patient zip code to surgical hospital zip code occurred.

Impact of COVID‐19 on Neonatal Surgical Metrics and Outcomes

To assess the specific impact of COVID‐19 on neonatal surgical outcomes, we compared patients born in the prepandemic era who underwent surgery in the same timeframe (2019 quarter 1–quarter 4) to those born during the pandemic era who had surgery within the corresponding period (2020 quarter 2–2021 quarter 2). Table 3 depicts the comparisons between the groups. A delay in time to surgery was noted in the pandemic group; the median age at surgery in the pandemic group was older (12 [4–51] versus 17 [4–140] days; P<0.0001). These delays were greatest in patients undergoing arterial switch operation (6 [3–11] versus 7 [4–15] days; P=0.03), aortopulmonary shunts (10 [5–28] versus 12 [6–45] days; P=0.02), tetralogy of Fallot repair (37 [8–123] versus 48 [13–244] days; P=0.01) and ventricular septal defect repair (44 [24–89] versus 53 [30–171] days; P=0.002). Likely related, the pandemic group demonstrated increased weight at time of surgery (mean 3.89 kg versus 3.46 kg; P<0.0001). The pandemic cohort experienced a statistically significant increase in hospital length of stay (18 [7–73] versus 24 [8–83] days; P<0.001). However, mortality until hospital discharge remained constant.

Regional Disparities in Response and Outcomes

We examined regional practice variability during the pandemic by performing combined in‐depth analysis of the Fetal Heart Society survey and STS‐CHSD data (Figure 2). Survey responses indicated >75% of respondents from the East South Central, West South Central, Middle Atlantic, New England, and Pacific regions agreed that pandemic restrictions significantly impacted fetal cardiology triage, referrals, and visits. This contrasted with the Mountain, South Atlantic, and East North Central regions, where <50% of respondents agreed.

Regional differences in type of diagnosis in the control and study group exist. There was a significant decrease in prenatal diagnosis of d‐TGA in the study group in the West South Central (47/240 [19.5%] versus 16/163 [9.8%]; P=0.01) and Mountain (11/117 [9.5%] versus 1/77 [1.3%]; P=0.02) regions during the pandemic but not in other regions. Similarly, there was increased left heart hypoplasia diagnosis in the South Atlantic region (46/328 [15%] versus 51/201 [26%]; P=0.002), and 22q11.2 deletion in the Mountain region (2/117 [1.7%] versus 7/77 [9.1%]; P=0.02) and New England region (3/176 [1.7%] versus 9/134 [6.7%]; P=0.02).

Surgical approaches also varied regionally. The East South Central (5/328 [1.5%] versus 12/201 [5.5%]; P=0.005), Pacific (1/217 [0.46%] versus 7/135 [5.2%]; P=0.004), and South Atlantic (7/528 [1.3%] versus 12/207 [5.8%]; P=0.0006) regions reported increased use of hybrid procedures for left heart hypoplasia during the pandemic, differing from practices in other regions. Preoperative factors and postoperative outcomes did not exhibit significant regional differences. Length of hospital stay varied by region, with Middle Atlantic (22 [8–88] versus 26 [9–99]) and New England (18 [7–54] versus 23 [9–58]) regions experiencing the greatest increase in the pandemic group compared with prepandemic (P<0.05).

DISCUSSION

Our investigation demonstrates that prenatal cardiology care experienced significant and regionally variable institutional restrictions and conservative triage strategies during the COVID‐19 pandemic. Despite restrictions, prenatal diagnosis of CCHD increased slightly during the pandemic era compared with the prepandemic era, and the CCHD neonates diagnosed during the pandemic originated from more disadvantaged neighborhoods and American Indian/Alaska Native populations than prepandemic neonates. Surgical mortality, preoperative risks, and postoperative complications also remained unchanged between the 2 eras, though length of hospital stay was longer and age at surgery was older in the pandemic era.

Our findings are consistent with other population‐level studies demonstrating stable or increased use of prenatal care services during the pandemic ¹¹ Our data and prior studies suggest that expeditious shifts to telehealth may have contributed to maintenance of prenatal access. ¹² , ¹³ Even in COVID‐19 epicenters, no decreases in the rates of mandatory in‐person testing, such as second‐trimester ultrasound, ¹⁴ , ¹⁵ existed, a key test needed to screen for CCHD during prenatal care ¹⁶ and initiate referral. Finally, there were no differences in postnatal maternal and infant outcomes during the pandemic and the prepandemic era. ¹² , ¹⁷

Our analysis demonstrates higher deprivation in patients with CCHD during the pandemic era, suggesting improved detection of CCHD in disadvantaged areas during the pandemic. The ADI is a socioeconomic demographic statistical measure that identifies high deprivation areas and includes 17 items across 6 domains: income, employment, education, housing, household characteristics, and housing type. ¹⁰ It has been validated using several known neighborhood disadvantage‐linked outcomes. ¹⁸ We found a similar significant increase in neonatal CCHD detection in the American Indian/Native Alaskan population in the pandemic era. One interpretation is that strategies adopted by centers for access and triage were successful in maximizing detection of significant fetal CHD for those with socioeconomically disadvantaged backgrounds. Several institutions used guidelines ¹ , ³ to optimize resource use and telehealth to improve access. Another possibility is that due to pandemic surges and limited health care providers during the pandemic, there was decreased availability for subspecialty prenatal care at smaller, community hospitals and practices. As a result, patients in these populations may have been referred to higher volume hospitals or practices where a prenatal diagnosis of CCHD was made. The more robust experience of these larger centers may have continued to enhance prenatal diagnosis rate. Alternatively, the observed increase in CCHD diagnosis among live births in lower socioeconomic status populations could represent regional disparities in pregnancy decision‐making counseling and access to termination of pregnancy or disproportionately lower rates of surgery for CCHD during the pandemic. Multiple studies have demonstrated significant impact of socioeconomic, demographic, ethnic, and racial impacts on prenatal diagnosis, mortality, and morbidity. ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴

Telehealth emerged as a critical tool in sustaining prenatal care continuity in the pandemic. The technology facilitated ongoing patient engagement and care management, mitigating the impact of restrictions on in‐person health care services. This adaptation underscores the potential of telehealth to enhance health care delivery, not just in pandemic settings but as a permanent feature of medical practice. Our findings corroborate the literature suggesting telehealth can maintain, if not improve, care quality for pregnant women and their fetuses, particularly in monitoring for conditions like CHD. ¹² , ¹³ Additionally, complete detailed level II anatomy scans performed by obstetricians rarely miss critical CHD diagnoses and result in referral to tertiary centers for delivery and neonatal cardiac care. ²⁵ , ²⁶ The fact that the detailed scans seem to have still picked up disease suggests that there was no significant inappropriate triaging of obstetric care and that patients still sought care despite the emergency. Although we cannot conclude this from our study data, it is a potential conclusion that deserves further research. We conjecture that even when fetal echocardiography was not available to patients during the pandemic, these detailed ultrasounds may have resulted in increased telehealth with fetal cardiologists and cardiac surgical centers. ⁵ Telehealth, in partnership with the primary obstetric team, remains a valuable tool in the postpandemic era, but recent studies suggest that it may exacerbate inequities in health care if not executed with consideration of socioeconomic challenges. ²⁷ , ²⁸

The American Society of Echocardiography guidelines for fetal cardiology triage recommended pausing screening visits for low‐risk indications and case‐by‐case considerations for moderate risk patients. ³ In our study, 73% of respondents followed these recommendations. Given appointment consolidation and triage practices during the pandemic did not detrimentally impact prenatal diagnosis or outcomes, a natural conclusion might be to adopt these practices in the postpandemic era ¹⁷ or when major crises disrupting health care systems occur. ²⁹ Certain low‐risk screening indications for fetal cardiovascular disease have recently been revised to a may consider rather than indicated in a recent update to indication for fetal echocardiogram guidelines. ³⁰ More nuanced analyses have shown that despite stable to increased levels of prenatal care use, there was disproportionately worse access to prenatal care for patients of color, those with public insurance, and those with language barriers during the pandemic. ²⁹ , ³¹ , ³² In addition, lack of knowledge about impact of COVID‐19 on pregnant individuals may have contributed to provider and patient behavior changes resulting in increased prenatal care use and resilience to overcoming prenatal care barriers. ¹³ , ³³ , ³⁴ Thus, future studies focused on understanding the mechanisms behind the inequities in access to prenatal care during the pandemic would directly inform and improve future protocols for triaged care.

Age at surgery was significantly increased during the pandemic. This difference was most notable in arterial switch operation, aortopulmonary shunts, tetralogy of Fallot, and ventricular septal defect repairs. The arterial switch operation and aortopulmonary artery shunts are typically performed within the first days to weeks of life, and the delay during the pandemic may have been due to increased hospital restrictions and decreased capacity, or alternatively related to COVID‐19 infection. For tetralogy of Fallot and ventricular septal defect, the discrepancy in age likely represents noncritical variants of the diagnoses that tolerated surgical intervention outside of the first 60 days of life. Despite this increase in surgical age, there was no increase in surgical mortality, suggesting that management of infants with CHD was sustained during this short period of time without negative impact. These data are not granular enough to understand the reasons for delay in surgery but are potentially related to public health restrictions and access to care such as routine doctor appointments, hospital operational changes which limited the number of surgeries occurring, limitations in staffing, and the uncertainty about the time course of the pandemic in the early phases. Although these data are limited, it may be helpful when considering future scenarios when there are limitations in providing standard CHD surgical care due to epidemics or other reasons (eg, natural disasters, cyberattack). Future work will need to address the role for regionalization of care when there are limitations only in certain regions. Additionally, it is critical to understand how available care is distributed in the population to mitigate disparities that might occur.

There were specific CCHD types that demonstrated difference in the prepandemic to pandemic era. Namely, the decline in the proportion of patients with d‐TGA was an unexpected result from our subanalysis. Although this may be due to usual year‐to‐year variability, we must consider pandemic effects. d‐TGA diagnosis decreased from 13% to 11% in the pandemic group. An important limitation of this study is the reliance on a surgical data set, which means any mortality before receiving surgery would be missed. Prior research investigating survival benefit of prenatal diagnosis in d‐TGA have had mixed results. One study identified infants with d‐TGA who died before presentation, whereas 2 found no difference in preoperative mortality in pre‐ and postnatal diagnosis groups. ³⁵ , ³⁶ , ³⁷ Lower prenatal diagnosis rate of d‐TGA is associated with lower socioeconomic quartile and rural residence. ³⁸ Diagnoses reliant on the outflow tract are more susceptible to a missed diagnosis than studies only requiring a 4‐chamber view, and the difference is more pronounced in rural areas. ³⁹ It is possible that the pandemic restrictions resulted in higher preoperative mortality in patients with d‐TGA in these specific regions, but that it could not be captured due to case ascertainment methodology. If so, one would expect similar declines in ductal‐dependent CCHD diagnoses that are difficult to diagnose prenatally. Although not statistically significant, there was a similar effective decrease in representation of the arch hypoplasia group, from 19% to 17% (P=0.06). There was no change in the rate of proportion of patients with total anomalous pulmonary venous return, which one is of the most frequently missed prenatal diagnoses. ⁴⁰ Further study is warranted to clarify if there was a pandemic‐related increase in preoperative d‐TGA mortality, using additional methods of case ascertainment such as birth defect registries.

There was increased diagnosis of left heart hypoplasia in the South Atlantic region during the pandemic, and this region also demonstrated increased use of hybrid stage 1 in that era. Hybrid stage 1 was increased in the East South Central and Pacific regions as well, whereas Norwood procedures remained unchanged between groups and across regions. The increase in hybrid procedures may represent surgical decision‐making or surgical hospital limitations resulting in selection of a less invasive neonatal procedure.

Finally, it is important to consider our findings relative to national and regional impacts of the pandemic overall. In 2020, the United States experienced a 15% surge in age‐adjusted mortality, with COVID‐19 becoming the third leading cause of death among adults. ⁴¹ Infant mortality rates did not parallel this trend, because COVID‐19 did not emerge as a leading cause of infant deaths. ⁴² Although no change occurred in the total population of women of childbearing age, childbirth rates declined 4% from 2019 to 2020, thus potentially affecting the demand for prenatal visits. Furthermore, US births to foreign‐born women declined during the pandemic due to international travel restrictions, ⁴³ and in the setting of known fetal CHD, some who may have chosen to deliver in the United States could not. Thus, the stability of prenatal CCHD diagnosis demonstrated in our study may reflect a decreased demand for prenatal visits during the pandemic. Alternatively, it could reflect the reallocation of health care resources to prioritize critical prenatal screenings. Our investigation suggests that the maintenance of prenatal CHD diagnosis rates was facilitated by strategic reevaluation of triage protocols, ensuring that critical care remained accessible even as the pandemic altered health care access. Finally, these shifts during the pandemic may in part account for the shifts in apparent disease prevalence, which do not have any plausible biological explanation, such as a decrease in d‐TGA diagnoses.

FUTURE DIRECTIONS

This study provides insights into the resilience and adaptability of prenatal and neonatal care systems in the face of the COVID‐19 pandemic and offers a foundation for future research and policy development. Investigating the mechanisms behind the observed geographic and socioeconomic differences in relation to CCHD diagnosis and subsequent outcomes informs targeted interventions to improve equity in health care. Additionally, the pandemic highlighted the potential of telehealth as a critical component of health care delivery, warranting further exploration into its integration into standard care practices. Understanding the regional variations in CHD diagnosis and management practices leads to more personalized and effective health care strategies, improving outcomes for all patients regardless of their geographic location.

LIMITATIONS

This study has the usual limitations of a retrospective cohort study. The reliance on surgical databases excludes infants who might have died before surgery, had surgery delayed to >60 days, or who died in utero or were spontaneously or therapeutically aborted, possibly skewing the understanding impact of prenatal diagnosis on survival rates. Also, a subset of patients was excluded from February 1, 2020 to May 1, 2020, because classifying timing of prenatal diagnosis as prepandemic or during the pandemic was challenging, because initial policy responses to the pandemic varied across the country. Because date of birth and estimated delivery date were unavailable, the control and study group time periods are subject to fixed cohort bias. At the start of a time‐period, shorter gestation pregnancies may be missed, and at the end of a time period, longer gestation pregnancies may be missed. ⁴⁴ Another limitation of using a national database is reliance on complete and accurate data entry. In this cohort, zip code information was available in 75%, patient ADI in 72%, hospital ADI in 92%, birth location in 88%, and race in 95%. Our methodology may not fully capture the indirect effects of pandemic restrictions on preoperative mortality, particularly for conditions such as d‐TGA, which saw an unexpected decrease in our cohort. Finally, there were a significant number of variables required from the STS database to fully test our hypotheses, including multiple CCHD diagnoses, surgeries, and outcomes.

CONCLUSIONS

Our study highlights the resilience of prenatal and neonatal CCHD care systems during the COVID‐19 pandemic, evidenced by stable prenatal diagnosis rates and adaptable health care practices. Moreover, increased hospital length of stay and age at surgery during the pandemic, without a corresponding increase in surgical mortality, points to a potential area of resilience needing further exploration.

As the health care system continues to navigate the aftermath of the pandemic and prepares for future crises, our findings emphasize the need for flexible, equitable health care delivery models that adapt to changing circumstances while ensuring high‐quality care for all patients, including those from marginalized communities. Regional variations in practice and outcomes, as observed, call for a nuanced approach to health care policy and resource allocation, ensuring every patient has access to timely and effective care regardless of socioeconomic status or geographic location.

Sources of Funding

This study was funded by Seattle Children's Hospital Heart Center and Fetal Center Institutional grants.

Disclosures

None.

Supporting information

Data S1–S2