EBNEO Commentary: Long-Term Outcomes following Necrotizing Enterocolitis and Spontaneous Intestinal Perforation

March 15, 2023

Manuscript citation:

Vaidya R, Yi JX, O’Shea TM, Jensen ET, Joseph RM, Shenberger J, Gogcu S, Wagner K, Msall ME, Thompson AL, Frazier JA, Fry R, Singh R; ELGAN-ECHO Study Investigators. Long-Term Outcome of Necrotizing Enterocolitis and Spontaneous Intestinal Perforation. Pediatrics. 2022 Nov 1;150(5):e2022056445. doi: 10.1542/peds.2022-056445. PMID: 36200375;

Reviewed by:

Jessica Jakubowicz, MD
Assistant Professor
Department of Pediatrics, Division of Neonatology
University of Arkansas for Medical Sciences/Arkansas Children’s Hospital


In (P) surviving former extremely low gestational age (GA) newborns (23 to 27 weeks GA) with a (I) history of medical necrotizing enterocolitis (NEC), surgical NEC, or spontaneous intestinal perforation (SIP) compared to (C) former extremely low GA newborns with neither NEC nor SIP, is there a (O) difference in growth and/or neurodevelopmental outcomes at (T) ages 10 and 15 years?


  • Design: This was a multicenter prospective cohort study. Data were derived from the extremely low gestational age newborns (ELGAN) study.
  • Allocation: This study did not involve allocation. For the current study, children were categorized into 4 groups: medical NEC, surgical NEC, SIP, and no NEC/ SIP, on the basis of diagnoses made during the NICU hospitalization.
  • Blinding: Neurodevelopmental assessments were done by examiners who were masked to children’s neonatal histories.
  • Follow-up period: For the purposes of this study, patients from the original ELGAN study were followed at ages 10 and 15 years old.
  • Setting: Between 2002 and 2004, 1506 infants born in 14 participating institutions across 5 states in the United States were enrolled as part of the observational cohort ELGAN study.
  • Patients:
    • 1222 ELGAN (23 to 27 weeks GA) infants survived until discharge
      • 966 were eligible for follow-up assessment at age 10 years (on the basis of availability of data on levels of protein biomarkers in the first 2 postnatal weeks)
        • 889 (92%) participated in comprehensive ND, neurobehavioral, and growth assessments at 10 years
        • 694 (71.8%) participated in comprehensive ND, neurobehavioral, and growth assessments at 15 years
      • Inclusion criteria:
        • All infants born between 23 weeks and 27 6/7 weeks of gestation, between March 2002 and August 2004, at 1 of the 14 institutions
        • No exclusion criteria were specified
      • Exposure:
        • Children were categorized into 4 groups on the basis of diagnoses made during the NICU hospitalization:
          • Medical NEC
          • Surgical NEC
          • SIP
          • No NEC/ SIP
        • For infants with NEC, a Bell’s classification stated in the medical record was recorded.
        • If an infant had NEC but no Bell’s classification was listed, the research assistant assigned a classification using a manual containing descriptions of stages of NEC on the basis of modified Bell’s classification.
        • The study manual specified that infants without NEC be classified as having SIP if they “had a gastrointestinal perforation documented on radiograph and an isolated intestinal perforation was confirmed at the time of exploratory laparotomy or strongly suspected clinically (if managed with Penrose drain and exploratory laparotomy was not done).”
      • Outcomes:
        • Compare growth and ND outcomes at 10- and 15-years follow-up for school-aged children born extremely preterm who developed NEC or SIP during their NICU stay with controls who had neither NEC nor SIP.
          • Growth outcomes
            • Height, weight, and head circumference (HC) were collected by a research assistant after all outer garments were removed.
            • If children were unable to stand unsupported, either a wheelchair scale or the difference of the parent’s weight, plus child’s weight, and the parent’s weight alone was used for weight measurements. The child’s length while lying down was used as a substitute.
          • Neurodevelopmental outcomes
            • Motor function was assessed with the Gross Motor Function Classification System (GMFCS) at age 10 and 15 years.
            • Cognitive outcomes
              • At age 10,cognitive outcomes were assessed with the School-Age Differential Ability Scales-II (DAS-II) Verbal and Nonverbal Reasoning subscales, which were averaged to create a full-scale IQ composite.
              • At age 15, full-scale IQ was assessed with Wechsler Abbreviated Scale of Intelligence-II.
              • Latent profile analyses were done in addition at both 10 and 15 years of age
            • Attention-deficit/hyperactivity (ADHD) outcomes
              • At age 10, participants were included in the ADHD symptom group if they met criteria in any 2 of the 3 following contexts:
                • Parent or caregiver completed the Child Symptom Inventory, Fourth Edition (CSI-4)
                • The child’s current teacher was asked to complete the Child Symptom Inventory, Fourth Edition, Teacher Checklist.
                • Information on the basis of the parent’s indication of the child having been diagnosed previously by a clinician to have ADHD.
              • At age 15, ADHD was assessed with the Mini International Neuropsychiatric Interview for Children and Adolescents.
            • Analysis:
              • Multivariable analysis was performed, adjusting for birth weight z-score, GA, and sex, when comparing outcomes between no NEC/SIP group and medical NEC, surgical NEC, and SIP.
              • Inverse probability weighting was conducted to adjust for the dropout of subjects with necrotizing enterocolitis.
                • For growth outcomes, weights were dependent on maternal education, maternal single status, public insurance, and severe chronic lung disease.
                • For ND outcomes, weights were dependent on maternal education, maternal single status, public insurance, severe chronic lung disease, white matter damage, and severe retinopathy of prematurity.
              • Differences were considered statistically significant if the corresponding P values were <.05. Confidence intervals (CIs) were constructed using alpha = .05.
            • Patient follow-up:
              • 1506 infants were enrolled in the original ELGAN study.
                • 7% with medical NEC
                • 3% with surgical NEC
                • 4% with SIP
              • 1222 infants were discharged alive from the NICU and 1198 survived to 10 years.
              • Follow-up data were available from 889 children at 10 years and 694 at 15 years.
                • At 10 year visit, 138 (15.5%) had medical NEC, 33 (3.7%) had surgical NEC, 29 (3.3%) had SIP, and 689 (77.5%) had no NEC/SIP.
                • At 15 year visit, 105 (15.1%) had medical NEC, 28 (4.0%) had surgical NEC, 21 (3.0%) had SIP, and 540 (77.8%) had no NEC/ SIP.

Main Results:

  • Baseline characteristics
    • Infants with SIP and surgical NEC were more likely male, had lower GA, and lower birth weight compared with infants with no NEC/SIP.
    • Infants with medical NEC had similar baseline demographic characteristics compared with infants with no NEC/SIP.
  • Medical NEC and outcomes
    • Children who had medical NEC had similar weight, BMI, height, and HC compared with children with no NEC/SIP at both age 10 and 15 years.
    • Children who had medical NEC had similar neurodevelopmental outcomes as children with no NEC/SIP.
  • Surgical NEC and outcomes
    • At age 10 years, children who had surgical NEC tended to weigh less and be shorter than children with no NEC/SIP, however these differences were not statistically significant.
    • At age 15 years, children who had surgical NEC had significantly lower average weight z-score, lower BMI z-score, and lower height z-score than children with no NEC/ SIP.
    • After adjusting for confounding variables and sample attrition, no differences were found at either 10 or 15 years of age between the ND outcomes of children who had surgical NEC and children with neither SIP nor NEC.
  • SIP and outcomes
    • At age 10 years, children who had SIP tended to have lower weight and height compared with no NEC/ SIP, however these differences were not statistically significant after adjusting for confounding variables.
    • When adjusting for sample attrition,
      • At age 10 years
        • Children with SIP had lower average weight (effect estimate: -0.63, 95% CI: -1.24 to -0.02) and lower height (effect estimate: -0.57, 95% CI: -0.97 to -0.16) than children with no NEC/SIP at age 10 years.
      • At age 15 years
        • Children who had SIP had similar weight, BMI, and height as children who had no NEC/SIP.
      • Children who had SIP had similar neurodevelopmental outcomes as children with no NEC/SIP.
Adjusted Associations between NEC and 10- and 15-Year Outcomes
No NEC/SIP Medical NEC Surgical NEC SIP
Effect Estimate (b)/OR (95% CI) P Effect Estimate (b)/OR (95% CI) P Effect Estimate (b)/OR (95% CI) P
Age-10 outcome
BMI z-score Referent 0.13

(-0.16 to 0.42)

0.38 -0.17

(-0.64 to 0.31)

0.49 -0.32

(-0.91 to 0.26)

Weight z-score Referent 0.15

(-0.09 to 0.40)

0.22 -0.20

(-0.64 to 0.24)

0.38 -0.44

(-1.07 to 0.20)

Height z-score Referent 0.06

(-0.16 to 0.29)

0.57 -0.14

(-0.47 to 0.19)

0.40 -0.31

(-0.75 to 0.12)

DAS-II FSIQ Referent 0.21

(-3.32 to 3.73)

0.91 -2.04

(-9.13 to 5.05)

0.57 -3.57

(-11.8 to 4.65)

LPA Referent 1.13


0.49 1.46


0.27 1.43


Age-15 outcome
BMI z-score Referent 0.24

(-0.04 to 0.53)

0.10 -0.55

(-1.09 to -0.01)

0.04* 0.08

(-0.60 to 0.75)

Weight z-score Referent 0.19

(-0.15 to 0.53)

0.28 -0.75

(-1.25 to -0.25)

<0.01* 0.09

(-0.76 to 0.93)

Height z-score Referent -0.08

(-0.34 to 0.18)

0.55 -0.65

(-1.16 to -0.14)

0.01* 0.03

(-0.53 to 0.59)

WASI-II FSIQ Referent -1.62

(-5.65 to 2.41)

0.43 -6.51

(-14.2 to 1.14)

0.10 -0.07

(-8.50 to 8.36)

LPA Referent 0.89


0.60 0.91


0.81 0.90



*Adjusted for Fenton birth weight z-score, GA, and sex. Table recreated from original paper

Conclusion: Surgical NEC and SIP were found to be associated with growth impairment in late childhood. ND outcomes among school-aged children born extremely preterm with any NEC or SIP were found to be no different from children without NEC/SIP.


Necrotizing enterocolitis (NEC) and spontaneous intestinal perforation (SIP) remain serious gastrointestinal complications of premature infants. NEC typically presents with systemic inflammation and mucosal or transmucosal necrosis of the intestinal wall which can be extensive. SIP in comparison tends to be localized and without systemic inflammation. Neonatal mortality has been reported to be higher in infants with NEC (particularly surgical NEC) and SIP when compared to infants with no history of NEC or SIP (1). When specifically comparing NEC and SIP mortality outcomes, NEC has been associated with a higher neonatal mortality (2).

In recent years there has been a focus on long-term outcomes of extremely premature infants. Most studies describe outcomes for infants with NEC or SIP in the earlier years (i.e., 18-30 months corrected age) and neurodevelopmental outcomes specifically. Infants with NEC or SIP have been reported to have an increased odds of death or significant neurodevelopmental impairment (NDI) at 18-30 months of corrected age (3). There has also been a reported increased incidence of NDI comparing surgical NEC to medical NEC (4).

Growth outcomes in the literature for infants with NEC or SIP are similar in limited follow-up age and varying in results. Some studies have shown no difference in growth outcomes with respect to NEC alone (5) and others demonstrate initial growth failure at discharge in infants with NEC but no difference in growth failure rates when compared to controls during follow-up (6).

There are strengths and limitations to the current study. Strengths include a large, prospective, multicenter cohort using infants with a gestational age as low as 23 weeks and masked ND assessments. Limitations to the study included the small number of infants with NEC or SIP as well as the sample attrition. Sample attrition was high particularly for the children of families of a lower socioeconomic position and inverse probability weighting was used to help correct for this. However, as the authors mention, this could lead to a selection bias thereby weakening results and decreasing broad applicability. Another important facet of this study is that the patient population in question is from nearly two decades ago. The field of neonatology has had many advancements during this time which could make the study population less relevant currently. This, however, will always be a problem when very long-term follow-up is undertaken.

Patient follow-up out to the age of 15 years old is uncommon in the field of neonatology and motivating for future studies. Importantly this current study demonstrates results to suggest that previously noted negative outcomes at earlier ages may not persist to school age, particularly the concern for NDI. This article provides an example of ND outcomes having potential to change over time. The authors suggest this could be affected by early intervention services, family support and schools along with “catch-up” growth. Sample attrition resulting in possible selection bias as mentioned previously could affect these results as well. More information from future long-term follow-up studies will be necessary to see if this potential for improved outcomes subsists.

NEC and SIP continue to be a major cause of short-term and potential long-term morbidity and mortality. Additional research in neurodevelopmental and growth outcomes in these patients is vital. Information regarding outcomes will help guide treatment and parental counseling while in the neonatal intensive care unit and direction for outpatient follow-up particularly in the areas of nutrition and development.

Funding: None

Conflicts of Interest: None


  1. Wadhawan R, Oh W, Hintz SR, Blakely ML, Das A, Bell EF, Saha S, Laptook AR, Shankaran S, Stoll BJ, Walsh MC, Higgins RD; NICHD Neonatal Research Network. Neurodevelopmental outcomes of extremely low birth weight infants with spontaneous intestinal perforation or surgical necrotizing enterocolitis. J Perinatol. 2014 Jan;34(1):64-70. doi: 10.1038/jp.2013.128. Epub 2013 Oct 17. PMID: 24135709; PMCID: PMC3877158.

  2. Shah, J., Singhal, N., da Silva, O. et al.Intestinal perforation in very preterm neonates: risk factors and outcomes. J Perinatol 35, 595–600 (2015). https://doi.org/10.1038/jp.2015.41

  3. Zozaya C, Shah J, Pierro A, Zani A, Synnes A, Lee S, Shah PS; Canadian Neonatal Network (CNN) and the Canadian Neonatal Follow-Up Network (CNFUN) Investigators. Neurodevelopmental and growth outcomes of extremely preterm infants with necrotizing enterocolitis or spontaneous intestinal perforation. J Pediatr Surg. 2021 Feb;56(2):309-316. doi: 10.1016/j.jpedsurg.2020.05.013. Epub 2020 May 15. PMID: 32553453.

  4. Matei A, Montalva L, Goodbaum A, Lauriti G, Zani A. Neurodevelopmental impairment in necrotising enterocolitis survivors: systematic review and meta-analysis. Arch Dis Child Fetal Neonatal Ed. 2020 Jul;105(4):432-439. doi: 10.1136/archdischild-2019-317830. Epub 2019 Dec 4. PMID: 31801792.

  5. Dilli D, Eras Z, Özkan Ulu H, Dilmen U, Durgut Şakrucu E. Does necrotizing enterocolitis affect growth and neurodevelopmental outcome in very low birth weight infants? Pediatr Surg Int. 2012 May;28(5):471-6. doi: 10.1007/s00383-012-3051-4. Epub 2012 Jan 25. PMID: 22274546.

  6. Hong CR, Fullerton BS, Mercier CE, Morrow KA, Edwards EM, Ferrelli KR, Soll RF, Modi BP, Horbar JD, Jaksic T. Growth morbidity in extremely low birth weight survivors of necrotizing enterocolitis at discharge and two-year follow-up. J Pediatr Surg. 2018 Jun;53(6):1197-1202. doi: 10.1016/j.jpedsurg.2018.02.085. Epub 2018 Mar 8. PMID: 29627178.

Leave a comment