Risk factors for complications of necrotizing enterocolitis in term neonates with congenital heart disease at a referral hospital in Colombia: A retrospective cohort study

Abstract

Introduction:

Necrotizing enterocolitis (NEC) is one of the most severe complications in neonates and infants. However, term infants with congenital heart disease (CHD) represent a scenario of multiple complications due to pathophysiological mechanisms that differ from the classic pathogenesis. The objective of the study was to determine factors associated with complications of NEC in term infants hospitalized for CHD in a pediatric intensive care unit of a referral hospital in Colombia between 2010 and 2023.

Materials and Methods:

A retrospective cohort study was conducted, including 148 term infants with CHD who developed NEC during their hospitalization. Demographic and clinical characteristics were analyzed. Descriptive and bivariate analyses were performed using the Chi-square and Mann–Whitney U-tests. A multivariate analysis was conducted using robust Poisson regression with Stata 18.0® software.

Results:

The most frequent CHD were hypoplastic left heart syndrome (HLHS), pulmonary atresia, and persistent ductus arteriosus. The most common NEC stages were IIa (38.5%) and IIb (23.7%), which were associated with complications such as short bowel syndrome (18.8%) and post-surgical sepsis (37.5%). Among factors related to complications from NEC, medical-surgical management increased the risk (adjusted relative risk: 10.14 and 95% confidence interval: 1.95–48.72), while cardiac surgery reduced the likelihood of complications.

Conclusions:

NEC in infants with CHD is associated with multiple factors currently under investigation, which may explain a pathophysiological mechanism distinct from that of classic NEC.

INTRODUCTION

Necrotizing enterocolitis (NEC) is a potentially life-threatening condition affecting neonates.[1] In this population, mortality rates range from 20 to 30%, particularly among those requiring surgical management.[2] While NEC is more frequently observed in preterm neonates with a gestational age between 29 and 32 weeks, full-term neonates (≥37 weeks of gestation) with risk factors such as low Apgar scores, prolonged rupture of membranes, exchange transfusions, neural tube defects, and congenital heart disease (CHD) may also develop the condition.[3]

This pathology results in severe intestinal injury and long-term sequelae, including intestinal stenosis, short bowel syndrome, and even neurodevelopmental alterations.[1] The overall incidence rates of NEC are estimated at 3%–15%, with approximately 7%–20% of cases occurring in full-term infants.[2]

There are notable differences in clinical presentation, primarily between full-term infants and preterm neonates. It should be noted that NEC in full-term infants presents with distinct pathophysiological mechanisms, including systemic blood flow compromise, a diastolic “steal” phenomenon, and greater susceptibility of the colon to ischemia due to inferior collateral blood supply compared to other intestinal regions.[4]

Children with CHD are estimated to have a 3.3%–11% risk of developing NEC.[3] The CHDs most strongly associated with NEC are those in which systemic or pulmonary flow depends on the patency of the ductus arteriosus. These conditions carry a heightened risk of mesenteric hypoperfusion and ischemia due to a diastolic “steal” phenomenon.[5]

However, optimal treatment remains a challenge and requires interdisciplinary management. Decisions regarding the appropriate empirical antibiotic regimen, strict bowel rest with gastric decompression, and the timing of surgery must be made on an individual basis.[6] In 20%–40% of all cases, surgical intervention is indicated if there is a lack of clinical response to conservative therapy.[5] In cases of pneumoperitoneum or other scenarios of intestinal perforation, surgical intervention is generally indicated as a last resort for management.[7]

Consequently, it is important to determine the factors associated with complications from NEC in children born at term with CHD, given the poor characterization of this population, where the outcome may be more deleterious than in other children with different medical conditions.

MATERIALS AND METHODS

Population and study design

A retrospective cohort study was conducted among children under 12 months of age admitted to the neonatal and pediatric intensive care units of a referral hospital in Colombia between January 1, 2010, and December 31, 2023. The participants were children diagnosed with CHD, born at term (>37 weeks of gestation), who developed NEC during their hospitalization.

Eligibility criteria

Inclusion criteria

The study included all infants (under 12 months of age) with CHD, born at term (>37 weeks of gestation), who were diagnosed with NEC based on clinical and radiological criteria defined in the modified Bell scale[7] and evaluated by pediatric surgery during their hospitalization in the pediatric cardiovascular intensive care unit.

Exclusion criteria

Infants with a history of gastrointestinal malformations and/or prematurity (<37 weeks of gestation), or NEC diagnosed before admission were excluded.

Data collection

Data were collected retrospectively from the medical records of children diagnosed with NEC, obtained from the institutional database. All patients assigned the ICD-10 codes P77 (NEC of the newborn) and K55.30 (NEC, unspecified) were identified, and the pre-established inclusion and exclusion criteria were applied. The data were recorded in a form developed in Google Forms® and subsequently exported to an Excel® file for statistical analysis.

Statistical analysis

Descriptive analysis was performed to summarize the data, including proportions for qualitative variables and measures of central tendency and dispersion for quantitative variables. Bivariate analysis evaluated the association between NEC complications and clinical-demographic variables. For quantitative variables with non-normal distribution, the Mann–Whitney U-test was applied after conducting the Shapiro–Wilk test; for qualitative variables, the Chi-square test was used. P value < 0.05 was considered statistically significant for establishing associations. Variables with P ≤ 0.05 or biological relevance were included in a multivariate robust Poisson regression model. All analyses were conducted using STATA 18.0 software (College Station, TX, USA).

Ethical considerations

No direct intervention was performed with the participants, and the study was classified as a low-risk investigation. The principles of confidentiality and privacy outlined in the Declaration of Helsinki were strictly followed, ensuring the protection of personal and sensitive data. The research protocol was reviewed and approved by the hospital’s research ethics committee.

RESULTS

Between 2010 and 2023, 372 children with CHD diagnosed with NEC were admitted to the hospital, with 264 participants excluded due to gastrointestinal malformations and a history of prematurity. Of the 148 remaining children, 38 required medical-surgical management, and 110 received conservative medical management [Figure 1].

Figure 1.

Study flow diagram

With respect to the demographic characteristics of the population, males predominated over females (61.5% vs. 38.5%). Eighty-seven percent (87.2%) of the participants had a birth weight >2500 g. The most common stage of NEC in children with CHD at term was IIa (38.5%), followed by IIb (23.7%), according to the modified Bell classification [Table 1].

Table 1.

Characteristics of the study population

Demographic characteristic	n=148, n (%)
Sex
Male	91 (61.5)
Female	57 (38.5)
Birth weight (g), median (IQR)	3000 (2700–3245)
Birth weight (g)
<2500	19 (12.8)
>2500	129 (87.2)
Age at diagnosis of NEC
<1 month	88 (59.5)
>1 month	60 (40.5)
Weight at diagnosis of NEC (g), median (IQR)	3220 (2835–3700)
Weight at diagnosis of NEC (g)
<2500	13 (8.8)
>2500–3500	81 (54.7)
>3500	54 (36.5)
Classification of CHD based on pulmonary blood flow
Increased pulmonary flow	100 (67.7)
Decreased pulmonary flow	48 (32.3)
Severity classification of NEC
Stage IA	2 (1.3)
Stage IB	17 (11.5)
Stage IIA	57 (38.5)
Stage IIB	35 (23.7)
Stage IIIA	17 (11.5)
Stage IIIB	20 (13.5)

IQR: Interquartile range, NEC: Necrotizing enterocolitis, CHD: Congenital heart disease

In 67.7% of cases, NEC occurred in children with CHD and increased pulmonary blood flow. The most common defects associated with NEC were hypoplastic left heart syndrome (HLHS), pulmonary atresia, persistent ductus arteriosus (PDA), and coarctation of the aorta [Tables 2 and 3].

Table 2.

Distribution of enterocolitis by type of Congenital heart disease

	n=148, n (%)
CHD with increased pulmonary flow
Hypoplastic left heart syndrome	19 (12.8)
Patent ductus arteriosus	15 (10.1)
Aortic coarctation	14 (9.5)
Ventricular septal defect	13 (8.8)
Interrupted aortic arch	12 (8.1)
Truncus arteriosus	7 (4.7)
Atrioventricular canal	6 (4.1)
Atrial septal defect	6 (4.1)
Anomalous venous return	5 (3.4)
Aortic valvular stenosis	2 (1.4)
Mitral valvular stenosis	1 (0.7)
CHD with decreased pulmonary flow
Pulmonary valvular atresia	19 (12.8)
Transposition of the great arteries	11 (7.3)
Tetralogy of Fallot	6 (4.1)
Ebstein’s anomaly	4 (2.7)
Pulmonary valvular stenosis	4 (2.7)
Tricuspid atresia	4 (2.7)

CHD: Congenital heart disease

Table 3.

Bivariate analysis of demographic factors associated with complications secondary to necrotizing enterocolitis

Characteristics	With complications (n=32), n (%)	Without complications (n=116), n (%)	cRR (95% CI)	P
Sex
Male	22 (24.2)	69 (75.8)	1.37 (0.70–2.69)	0.34
Female	10 (17.5)	47 (82.5)
Birth weight (g)
>2500	29 (22.5)	100 (77.5)	1.42 (0.48–4.22)	0.76
<2500	3 (15.8)	16 (84.2)
Age at diagnosis of NEC
<1 month	19 (21.6)	69 (78.4)	0.99 (0.50–1.86)	0.99
>1 month	13 (21.7)	47 (78.3)
Weight at diagnosis of NEC (g)
<2500	5 (38.5)	8 (61.5)	1.17 (0.91–1.51)	0.20
2500–3500	15 (18.4)	66 (81.6)	0.96 (0.83–1.11)	0.60
>3500	12 (22.2)	42 (77.8)	Ref.	-
Classification of CHD based on pulmonary blood flow
Decreased pulmonary flow	15 (31.2)	33 (68.8)	1.83 (1.00–3.36)	0.048
Increased pulmonary flow	17 (17)	83 (83)
Severity classification of NEC
Stage III	27 (73)	10 (27)	16.2 (6.72–39.01)	<0.001
Stage I–II	5 (4.5)	106 (95.5)

CHD: Congenital heart disease, NEC: Necrotizing enterocolitis, Ref.: Reference category, cRR: Crude relative risk, CI: Confidence interval

Regarding preoperative clinical characteristics, 56.8% (84/148) of children received prostaglandin before the development of NEC. Before NEC onset, 48.7% received total parenteral nutrition (TPN), and 26.3% received enteral feeding.

Among NEC cases related to cardiovascular surgery, 55% (60/109) occurred preoperatively and 45% (49/109) postoperatively. Median time to onset was 18.5 days presurgery (interquartile range [IQR]: 9–34.5) and 13 days postsurgery (IQR: 4–31).

Regarding the time elapsed between the diagnosis of NEC and death in children affected by this condition, a median of 20 days was observed (IQR: 2.5–53). An overall mortality rate of 31.7% (47/148) was estimated in the total population of children affected by NEC [Table 4].

Table 4.

Bivariate analysis of clinical factors associated with necrotizing enterocolitis

Characteristics	With complications (n=32), n (%)	Without complications (n=116), n (%)	cRR (95% CI)	P
Use of prostaglandin before NEC
No	15 (23.4)	49 (76.6)	1.15 (0.62–2.13)	0.63
Yes	17 (20.2)	67 (79.8)
Vasopressor support before NEC
No	15 (26.8)	41 (73.2)	1.45 (0.78–2.66)	0.23
Yes	17 (18.5)	75 (81.5)
Use of mechanical ventilation before NEC
Yes	21 (21.2)	78 (78.8)	0.94 (0.49–1.79)	0.86
No	11 (22.5)	38 (77.5)
Use of oxygen therapy before NEC
<50%	23 (20)	92 (80)	1.03 (0.88–1.32)	0.79
>50%	2 (16.7)	10 (83.3)	Ref.	-
No	7 (33.3)	14 (66.7)	1.18 (0.88–1.32)	0.33
Type of nutritional support before NEC
Parenteral nutrition	13 (18.1)	59 (81.9)	1.05 (0.84–1.32)	0.62
Enteral	12 (30.8)	27 (69.2)	1.20 (0.94–1.52)	0.13
Crystalloids	5 (23.8)	16 (76.2)	1.12 (0.85–1.46)	0.41
Mixed	2 (12.5)	14 (87.5)	Ref.	-
Required cardiac surgery
Yes	19 (17.4)	90 (82.6)	0.52 (0.28–0.95)	0.04
No	13 (33.3)	26 (66.7)
Development of NEC in relation to cardiac surgery
Presurgery	10 (16.7)	50 (83.3)	0.90 (0.40–2.05)	0.81
Postsurgery	9 (18.4)	40 (81.6)
Days between NEC and cardiac surgery, median (IQR)	26 (17.2–43.2)	17 (9–32.5)		0.21
Days between cardiac surgery and NEC, median (IQR)	15 (4–30)	12.5 (3.7–31.7)		0.98
Days of hospitalization before NEC	8 (3–18)	10.5 (3.2–25.7)		0.60
Total hospital duration median (IQR)	45 (9.7–105.3)	62.5 (30.2–111.3)		0.16
Days of hospitalization after NEC, median (IQR)	32.5 (3.7–86.5)	48.5 (16.7–84.2)		0.21
Days of hospitalization between NEC and death, median (IQR)	5 (1–30)	41 (10.5–85)		0.026
Classification of enterocolitis treatment
Medical-surgical	28 (73.7)	10 (26.3)	20.2 (7.6–54)	<0.001
Medical	4 (3.6)	106 (96.4)
Death
Yes	17 (36.2)	30 (63.8)	2.43 (1.33–4.44)	0.003
No	15 (14.8)	86 (85.2)

NEC: Necrotizing enterocolitis, IQR: Interquartile range, cRR: Crude relative risk, CI: Confidence interval, Ref.: Reference category

Bivariate analysis

A bivariate analysis was conducted to identify risk factors for complications from NEC in children with CHD. Sex, weight, and age did not show a significant relationship with complications. However, children with CHD with decreased pulmonary flow physiology had a higher risk of developing complications compared to those with CHD with increased pulmonary physiology [Table 3].

A significant association was found between NEC severity and CHD with increased pulmonary flow (P = 0.009). Stage IIA and IIB of NEC primarily affected children with CHD with increased pulmonary blood flow. In contrast, Stage IIIB NEC was more frequent in children with CHD with decreased pulmonary blood flow [Figure 2].

Figure 2.

Severity of necrotizing enterocolitis according to pulmonary physiology in congenital heart disease

Consequently, pulmonary stenosis and tetralogy of Fallot were the CHD with the highest number of cases of stage IIIB NEC. In contrast, PDA was mainly associated with stage IIB NEC, whereas aortic coarctation was more frequently observed in stage IIA NEC [Figure 3a and b]. Stage III NEC was associated with higher complication risk (crude relative risk [cRR]: 16.2 and 95% confidence interval [CI]: 6.72–39.01), although this decreased after adjustment (adjusted relative risk [aRR]: 1.88 and 95% CI: 0.46–9.64).

Figure 3.

(a) Severity of enterocolitis in congenital heart disease (CHD) with decreased pulmonary flow. (b) Severity of enterocolitis in CHD with increased pulmonary flow. HLHS: Hypoplastic left heart syndrome, PDA: Patent ductus arteriosus, CoA: Aortic coarctation, IAA: Interrupted aortic arch, TA: Truncus arteriosus, VSD: Ventricular septal defect, ASD: Atrial septal defect, AVC: Atrioventricular canal, PAVD: Anomalous pulmonary venous drainage, AS: Aortic valve stenosis, AM: Mitral valve stenosis, PVA: Pulmonary valve atresia, TGA: Transposition of the great arteries, TOF: Tetralogy of Fallot, EA: Ebstein’s anomaly, PVS: Pulmonary valve stenosis, TA: Tricuspid valve atresia

Other factors, such as prostaglandin use, vasopressor support, and mechanical ventilation, were not associated with NEC complications [Table 4]. Patients undergoing cardiac surgery had a 48% lower risk of NEC complications compared with children not undergoing cardiac surgery (cRR: 0.52 and 95% CI 0.28–0.95) [Table 4].

Finally, children who experienced complications from NEC had twice the risk of death compared to patients without NEC complications (cRR: 2.43 and 95% CI: 1.33–4.44).

Treatment and type of complications in children with necrotizing enterocolitis

PDA was the most frequent CHD associated with NEC complications, followed by pulmonary atresia, tetralogy of Fallot, and tricuspid atresia. Complications occurred in 21.6% of children with NEC, with postoperative sepsis being the most common (37.5%), followed by short bowel with loss of the ileocecal valve (12.5%) [Table 5].

Table 5.

Characteristics of treatment for necrotizing enterocolitis and complications

Clinical characteristics	n=148, n (%)
Classification of enterocolitis treatment
Medical	110 (74.3)
Medical-surgical	38 (25.7)
Classification of surgical treatment
Bowel resection with anastomosis	25 (77.7)
Laparotomy	7 (8.1)
Bowel resection with diversion	6 (14.2)
Complications associated with NEC
No	116 (78.4)
Yes	32 (21.6)
Classification of complications secondary to NEC
Postsurgical sepsis	12 (37.5)
Presurgical sepsis	3 (9.4)
Short bowel	6 (18.8)
Short bowel + loss of the ileocecal valve	4 (12.5)
Loss of the ileocecal valve without Short Bowel	4 (12.5)
Bowel obstruction	2 (6.2)
Enterocutaneous fistula	1 (3.1)

NEC: Necrotizing enterocolitis

Postoperative sepsis was the most common in children with VSD and PDA. Short bowel occurred in CHD with decreased pulmonary flow, such as tetralogy of Fallot, pulmonary atresia, and tricuspid atresia. Short bowel with valve loss, the most severe complication, occurred only in children with tricuspid atresia, PDA, and aortic coarctation [Figure 4].

Figure 4.

Distribution of necrotizing enterocolitis complications according to congenital heart disease. PDA: Patent ductus arteriosus, PVA: Pulmonary valve atresia, TOF: Tetralogy of Fallot, TA: Tricuspid valve atresia, PAVD: Anomalous pulmonary venous drainage, VSD: Ventricular septal defect, HLHS: Hypoplastic left heart syndrome, IAA: Interrupted aortic arch, AVC: Atrioventricular canal, EA: Ebstein’s anomaly, CoA: Aortic coarctation, TGA: Transposition of the great arteries

Regarding NEC treatment, 74.3% of children received medical management, while 25.7% required surgical intervention. Among surgeries, intestinal resection with anastomosis was most common (77.7%), followed by resection with diversion (14.2%).

The risk of complications from NEC was significantly higher in children requiring surgical management (cRR: 20.2 and 95% CI: 7.6–54) than in those receiving medical management.

Multivariate analysis

In the analysis of sociodemographic variables, no significant association was found between sex or birth weight of 2500 g or more and the occurrence of complications. In the bivariate analysis, CHD with decreased pulmonary blood flow was linked to a higher risk of NEC-related complications (cRR: 1.83 and 95% CI: 1.00–3.36). However, after model adjustment, the risk appeared overestimated, with a nonsignificant adjusted relative risk (aRR: 1.17 and 95% CI: 0.60–2.28) [Table 6].

Table 6.

Multivariable analysis of factors associated with complications due to necrotizing enterocolitis

Characteristic	cRR (95% CI)	P	aRR (95% CI)	P
Sex
Male	1.37 (0.70–2.69)	0.34	1.09 (0.60–2.05)	0.79
Female	Ref.	Ref.	Ref.	Ref.
Birth weight (g)
>2500	1.42 (0.48–4.22)	0.76	1.16 (0.45–3.57)	0.77
<2500	Ref.	Ref.	Ref.	Ref.
Classification of CHD based on pulmonary blood flow
Decreased pulmonary flow	1.83 (1.00–3.36)	0.048	1.17 (0.60–2.28)	0.63
Increased pulmonary flow	Ref.	Ref.	Ref.	Ref.
Severity classification of NEC
Stage III	16.2 (6.72–39)	<0.001	1.88 (0.46–9.64)	0.41
Stage I–II	Ref.	Ref.	Ref.	Ref.
Classification of NEC treatment
Medical-surgical	20.2 (7.6–54)	<0.001	10.14 (1.95–48.72)	0.005
Medical	Ref.	Ref.	Ref.	Ref.
Required cardiac surgery
Yes	0.52 (0.28–0.95)	0.04	0.58 (0.32–1.00)	0.07
No	Ref.	Ref.	Ref.	Ref.
Use of prostaglandin before NEC
Yes	1.15 (0.62–2.13)	0.63	1.27 (0.60–2.77)	0.529
No	Ref.	Ref.	Ref.	Ref.
Vasopressor support before NEC
No	1.45 (0.78–2.66)	0.23	1.20 (0.54–2.52)	0.63
Yes	Ref.	Ref.	Ref.	Ref.
Enteral nutrition before NEC
Yes	2.46 (0.78–14.6)	0.20	1.34 (0.60–2.98)	0.47
No	Ref.	Ref.	Ref.	Ref.

NEC: Necrotizing Enterocolitis, Ref.: Reference category, cRR: Crude relative risk, aRR: Adjusted relative risk, CI: Confidence interval

Similarly, in the bivariate analysis, Stage III NEC was associated with a 16-fold higher risk of complications than Stage I–II NEC. However, after model adjustment, the risk decreased to an aRR of 1.88 (95% CI: 0.46–9.64), indicating a weaker association. For medical-surgical management, the unadjusted model showed a cRR of 20.2. After adjusting for sex and age, the risk was notably reduced, with an aRR of 10.14 (95% CI: 1.95–48.72).

Regarding cardiovascular surgery, the results suggested a protective effect, with a reduced risk of NEC-related complications in children who underwent cardiac surgery (aRR: 0.58 and 95% CI: 0.32–1.00) compared to those who did not. Finally, the use of prostaglandin, vasopressors, and the type of nutrition before NEC onset were not significantly associated with complications.

DISCUSSION

NEC in term infants with CHD remains a pathology with unfavorable outcomes, often leading to severe complications, reduced quality of life, and even death. Despite advances in CHD management, various demographic and clinical factors affect the gastrointestinal tract. These interactions are the focus of recent studies aimed at identifying interventions to reduce the risk of NEC.[8]

The pathophysiology of classic NEC has been under debate, leading to the use of the term “cardiac enterocolitis,” which encompasses pathogenic mechanisms specific to non-premature infants with CHD. First described in 1976, its incidence ranges from 3% to 7.1%. Factors such as retrograde diastolic flow, mesenteric hypoperfusion, postsurgical ischemia/reperfusion, the adaptive response of the intestinal microvasculature to chronic hypoxia, hyperviscosity, and intestinal dysbiosis have been associated with this condition.[8,9,10]

Among immunological factors, toll-like receptor 4 activation by stimuli such as ischemia plays a key role in NEC pathogenesis in term neonates with CHD. Additionally, two SIGIRR gene variants (p.S80Y and c. 102_121dup) have been identified exclusively in NEC cases, suggesting a potential genetic predisposition.[11]

This retrospective study included a cohort of 148 full-term infants who developed NEC over 13 years at a tertiary care institution specializing in CHD in Colombia. The study aimed to identify factors associated with secondary complications due to NEC.

Regarding demographic characteristics, sex was not significantly associated with the complications, despite males comprising 61.5% of the cohort. A majority (87.2%) had a birth weight over 2500 g, contrasting with premature neonates, in whom low birth weight is a factor associated with the presentation of NEC and poor outcomes.[12]

Most infants (59.5%) were younger than 1 month of age, but this was not associated with increased complications, consistent with Bubberman et al., who found that NEC was most frequent between 22 and 25 days of age in infants with CHD.[13]

However, CHD with decreased pulmonary blood flow seemed to increase the likelihood of complications compared to heart disease with increased pulmonary blood flow (aRR: 1.17 and 95% CI: 0.60–2.28). The most common diagnoses included HLHS and pulmonary atresia, which were the most frequent conditions, followed by PDA, aortic coarctation, VSD, and transposition of the great arteries (TGA), consistent with previous studies in term neonates.[14,15,16,17] These findings may be explained by diastolic steal in ductus-dependent CHD, in which systemic flow through the ductus arteriosus reduces mesenteric perfusion.[17,18]

The use of prostaglandin and inotropic support before NEC onset was not associated with complications. Some studies suggest that prostaglandin may reduce the risk of NEC, although high doses or prolonged use can be harmful.[18] Inotropes, especially when used for more than 7 days, may increase NEC risk.[19]

Similarly, enteral nutritional support before NEC was also not associated with complications. While 48.7% received TPN, 26.3% were fed enterally, most often among infants with VSD. Nutritional practices in CHD remain debated. Although breast milk is beneficial for maintaining mucosal integrity and immunity,[9,20] enteral feeding increases intestinal blood flow demands. However, studies show that enteral stimulation of 20–30 ml/kg/day does not increase NEC risk in children during the preoperative cardiovascular phase.[21,22]

Recent studies indicate that corrective or palliative cardiac surgery reduces mortality in children with CHD. Similarly, our findings showed a 42% lower risk of NEC-related complications in surgical patients (aRR: 0.58 and 95% CI: 0.32–1.00), which could decrease mortality.[23]

In contrast, 55% of the children developed NEC before cardiac surgery, with a median time of 18.5 days between admission and NEC (IQR: 9–34.5). Aortic coarctation, pulmonary atresia, HLHS, and TGA were the most frequent pathologies to debut with preoperative NEC. These findings differ from those of Lau et al., who reported a higher incidence of preoperative NEC in nonductus-dependent CHD and postoperative NEC in ductus-dependent lesions.[24]

Postoperative NEC occurred in 45% of children, without an increased risk of complications, with a median onset of 13 days after surgery (IQR: 3–41). HLHS, pulmonary atresia, and VSD were the most frequently associated pathologies. These findings contrast with those of Kargl et al., who identified tetralogy of Fallot and aortic arch hypoplasia as the most common, with a median onset of 19 days (IQR: 10–29).[25]

Some authors suggest that this phenomenon may be explained by reduced mesenteric perfusion due to diastolic steal in patent ductus arteriosus–dependent heart disease, and by the use of bypass, which affects the permeability of the intestinal mucosa, facilitating bacterial translocation.[26] In addition, hypothermia, reperfusion, systemic blood pressure variability, and postsurgical inflammatory response syndrome are other mechanisms that could explain the development of NEC by triggering an adaptive vasoconstriction imbalance between endothelin 1 and nitric oxide.[23,25,27]

Regarding the severity of NEC, 25% of children presented with Stage III NEC, the most severe form, which was initially associated with more complications than Stages I and II. Pulmonary atresia was the congenital heart defect most frequently associated with this severe form, followed by tetralogy of Fallot. Lau et al. reported similar data regarding NEC distribution, with 26% for Stage III NEC, 43% for Stage II NEC, and 31% for Stage I NEC.[24]

Concerning the management of children with NEC, a recent meta-analysis reported that premature neonates required surgical management for NEC more frequently than children with CHD, who required surgery between 0.5% and 1.1%.[9] In contrast, other authors reported higher rates of surgical intervention in the management of NEC between 15% and 38%,[15] similar to the findings of our study, where 25.7% of the children required medical-surgical management, with intestinal resection, and anastomosis being the most performed procedure (77.7%), followed by intestinal resection with diversion (14.2%).

Although surgical management is the only therapeutic option in cases of NEC with a poor prognosis, it was established that this treatment can increase the risk of NEC complications (aRR: 10.14 and 95% CI: 1.95–48.7) in infants with CHD. However, this finding may be influenced by the number of children who developed complications and by factors related to the management of CHD according to its complexity.

Among the complications seen in children with CHD who underwent surgical management, postoperative sepsis was the most common (37.5%), followed by short bowel with preservation of the ileocecal valve (18.8%) and finally short bowel with loss of the ileocecal valve (12.5%).

Regarding the anatomical regions most affected by cardiac NEC, they appeared to differ from those affected by classic NEC in premature infants, where the primarily immature intestine is mostly affected in the small intestine. In contrast, in children with CHD, the colon and distal ileum are more susceptible to secondary injury due to compromised mesenteric perfusion and sensitivity to hypoxia.[8,28] It is important to note that the intestines of children with CHD lack collateral vascular networks that would improve tissue perfusion.[8]

Finally, Spinner et al. have reported an overall mortality rate of 20%–30%, especially in those children who received surgical management for NEC or who were classified in Stage II or III,[13,14,24] similar to our study, which found that 31.7% of infants under 1 year of age who developed NEC died, and the risk of death increased 2.4 times in children with complications compared to those without complications.

It should be mentioned that no lethality was established for NEC, which would have been important to determine whether the children died primarily from NEC or other causes associated with the management of CHD and its complications, which is a limitation in most studies.

Limitations

Study limitations include its retrospective, single-center design, which constrained the sample size for analyzing NEC-related complications. The absence of hemodynamic and laboratory data limited potential associations and external comparisons.

However, this is one of the few documented studies in the scientific literature from a reference center specializing in the management of complex CHD in Colombia and South America that attempted to identify factors related to NEC complications and the management implemented. This is an unusual clinical scenario that still requires further studies to develop predictive models for early diagnosis, avoid complications, and establish therapeutic guidelines for this special population.

CONCLUSIONS

Surgical treatment in children with NEC appears to be linked to intestinal complications, though other factors may also contribute. CHD-related cardiac pathophysiology influences NEC severity through hemodynamic changes and impaired intestinal perfusion. While corrective or palliative cardiac surgery may lower NEC risk, intraoperative factors could trigger its onset postoperatively. Overall mortality aligned with global rates but cannot be attributed solely to NEC due to potential confounding factors.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.