Childhood Educational Outcomes of Children Born with Gastroschisis

Abstract

Objective:

To compare third grade literacy and mathematics test proficiency among children born with gastroschisis vs. unaffected controls and identify predictors of academic proficiency among these children.

Study Design:

Infants born with gastroschisis (2000–2005) were identified from the Arkansas Reproductive Health Monitoring System. For each case, two controls were randomly selected from birth certificates and matched for hospital and month of birth, sex, and race/ethnicity. Data on re-hospitalization within the first 6 years and payer data were abstracted from the Arkansas Hospital Inpatient Discharge Database. Surviving cases and controls were linked to the Arkansas Department of Education database containing achievement test scores. The primary outcome was proficiency, defined as performance at or above grade level, on third grade achievement tests. Cases and controls who did not attend public schools could not be linked to education records. Multivariable logistic regression models evaluated the association between study characteristics and academic proficiency.

Results:

The final study cohort included 47 cases and 63 controls. There was no statistically significant difference in rate of literacy (68% vs. 81%, P=0.65) or mathematics proficiency (89% vs. 87%, P=0.15) between cases and controls. On multivariable analysis, a complex gastroschisis (defined as atresia, volvulus, necrosis, or perforation of the bowel) was associated with lower proficiency in literacy (AOR=0.1, 95% CI: 0.02, 0.58; P=0.01). No neonatal or maternal characteristics predictive of lower proficiency in mathematics were identified.

Conclusion:

Among children born with gastroschisis, a complex gastroschisis was associated with lower proficiency on third grade literacy achievement tests.

INTRODUCTION

Gastroschisis is a full-thickness defect of the abdominal wall that is characterized by varying degrees of herniation of the intraperitoneal contents.¹ Multiple surveillance reports have documented an increase in the prevalence of this congenital defect with a survival rate of >90%.^2–3 In addition, gastroschisis was identified to have one of the most expensive average neonatal hospital charges among non-cardiac birth defects in the U.S.⁴ The rising birth prevalence of gastroschisis and the high survival rate and healthcare expenditures of infants born with this defect underscore the need to better understand its potential lasting consequences.

Infants born with gastroschisis have various risk factors that may adversely impact their neurodevelopment. Often, they are moderately preterm⁵ and small for gestational age (SGA).^6,7 Additionally, many of these infants are born to young mothers with lower socioeconomic status (SES).⁸ They require surgery soon after birth⁹ and are exposed to several anesthetic and analgesic agents.^10,11 Studies assessing the childhood outcomes of these infants varied widely in their outcome measures and follow-up periods with some describing their overall health status or quality of life^12–15 and others evaluating their motor function^16–19 or behavior.^{16, 20–22} Studies that specifically assessed their neurodevelopmental outcomes invariably relied on questionnaires¹⁹ or different batteries of aptitude tests such as the Wechsler Intelligence Scale,^{16–18, 21–23} which are primarily designed to predict a child’s ability to learn new skills. On the other hand, we aimed to use performance on achievement tests, a measure intended to reflect what children are actually learning,²⁴ to compare the childhood academic performance of infants born with gastroschisis in Arkansas (2000–2005) vs. unaffected controls and then identify neonatal and maternal predictors of academic proficiency among these children. Our hypothesis, a priori, was that children born with gastroschisis (given their biological risk factors) would have worse academic performance compared to unaffected controls.

METHODS

This retrospective cohort study was approved by the Arkansas Department of Health (ADH) Scientific Advisory Committee and the University of Arkansas for Medical Sciences (UAMS) Institutional Review Board.

Arkansas Reproductive Health Monitoring System (ARHMS)

Children born with gastroschisis, from January 1, 2000 through December 31, 2005, were identified through ARHMS, a statewide, population-based birth defects surveillance program that uses active case ascertainment to monitor approximately 40,000 births. Trained Health Information Systems staff abstract medical records at 46 birthing facilities, including all delivering hospitals, the state’s only pediatric specialty-care hospital and associated clinics, and high-risk pregnancy and prenatal diagnosis centers. ARHMS codes birth defects using the six-digit British Pediatric Association extension of the International Classification of Diseases, Ninth Edition Clinical Modification (ICD-9-CM) coding system, as modified by the Division of Birth Defects and Developmental Disabilities of the Centers for Disease Control and Prevention and by ARHMS. Eligibility for ARHMS requires Arkansas residency of the mother at the time of delivery and diagnosis of a structural congenital malformation during pregnancy, after stillbirth or termination of pregnancy, after live birth, and up to 2 years of age.

Arkansas Department of Health

The Health Statistics Branch of the ADH manages data from births and deaths in Arkansas. ARHMS data were linked to ADH Vital Health Statistics. This linking process was performed by securely sending a data file to ADH containing ARHMS key identifiers such as mother and child first and last names. Two unaffected controls were then randomly selected from birth certificate data and matched for each case on hospital and month of birth, sex, and race/ethnicity. The resultant file was subsequently linked to infant death records using birth certificate number and to fetal death records using mother’s name and child’s date of death. This linkage allowed us to evaluate if cases or controls died before completing third grade (approximately 9 years of age). The ADH also provided demographic and socioeconomic characteristics and limited clinical variables (Table I). SGA and large for gestational age (LGA) infants were defined, based on the U.S. sex specific national standard growth curves, as infants with birth weight <10^th or >90^th percentile, respectively.²⁵

Table I.

Description of Maternal and Infant Characteristics for Infants Born with Gastroschisis (Cases) and Infants Born without Structural Birth Defects (Controls) in Arkansas, 2000–2005^a,b

Characteristics	Cases N=47	Controls N=63	P-value
Maternal Characteristics
Age (years), mean (SD)	20 (3)	25 (6)	<0.0001
Parity, n (%)			0.04
1	44 (94)	50 (79)
≥2	3 (6)	13 (21)
Caucasian race, n (%)	42 (89)	55 (87)	0.74
Prenatal care (yes), n (%)	45 (98)	62 (98)	1.00
Missing	1	0
Smoking (yes), n (%)	12 (26)	11 (18)	0.30
Alcohol consumption in pregnancy (yes), n (%)	1 (2)	0 (0)	0.43
Less than high school education, n (%)	13 (28)	12 (19)	0.28
Neonatal Characteristics
Male sex, n (%)	30 (64)	41 (65)	0.89
Birth weight (g), mean (SD)	2618 (467)	3008 (814)	0.0002
Gestational age (weeks), mean (SD)	37 (2)	38 (4)	0.09
Preterm (yes), n (%)	16 (34)	18 (29)	0.54
Size for gestation, n (%)			0.007
SGA	8 (17)	3 (5)
AGA	39 (83)	53 (84)
LGA	0	7 (11)
Singleton, n (%)	47 (100)	59 (94)	1.00
Apgar score at 5 minutes <7, n (%)	6 (13)	1 (2)	0.04
Delivery by cesarean section, n (%)	18 (38)	27 (43)	0.63
Payer type, n (%)c			0.74
Medicaid or self-pay	27 (57)	32 (54)
Private	20 (43)	27 (46)
Missing	0	4

AGA, appropriate for gestational age; LGA, large for gestational age; SD, standard deviation; SGA, small for gestational age.

^aThese are the cases of gastroschisis who were transferred to Arkansas Children’s Hospital following birth

^bBoth cases and controls presented in this table have survived till third grade and were successfully linked to their third grade school standardized achievement tests

^cData on payer type were collected from the Arkansas Inpatient Hospital Discharge Database.

Chi-square/Fisher Exact tests were used for comparison of categorical variables based on sample size; T-test/Wilcoxon Rank Sum test were used for mean and median comparisons, respectively.

Arkansas Children’s Hospital (ACH)

ACH is the only tertiary children’s hospital in Arkansas. Over 90% of infants born with gastroschisis in Arkansas are transferred to ACH shortly after birth. If the medical record number was not available, infant name and date of birth as well as mother’s name were used to match cases with gastroschisis to their medical records at ACH. Clinical data were abstracted by study investigators from the medical records including type of gastroschisis (simple vs. complex [defined as atresia, volvulus, necrosis, or perforation of the bowel]),^26,27 intestinal failure (defined as time to full enteral feeding >6 weeks),²³ and age at full enteral feeding (defined ≥100 kcal/kg/d of enteral feeding) (Table II; online).

Arkansas Hospital Inpatient Discharge Database (AHIDD)

The AHIDD includes discharges in Arkansas with a hospital length of stay of ≥1 day. Clinical information includes dates of service, length of stay, procedures, discharge diagnoses, and discharge status. The AHIDD includes a unique person identifier that allows the linkage of a person across hospitalizations and across years. Re-hospitalization among infants with gastroschisis typically occurs early in life.¹⁷ Using the birth certificate number, the ADH linked ARHMS surviving cases with gastroschisis to their AHIDD discharge claims to identify if patients were re-hospitalized during the first 6 years. The number of cases who had gastroschisis-related readmissions (e.g., bowel obstruction or feeding intolerance) and the median number and length of stay of such readmissions were determined. The AHIDD also documents payer coverage.

Arkansas Department of Education (ADE)

The file including data from surviving cases and controls was securely transmitted to the ADE. The ADE database includes scores and proficiency designations from Arkansas public school students who took the Augmented Benchmark Examination tests in literacy and mathematics (grades 3–8), that are administered annually. Scores designated as proficient or advanced (“proficient”) represented performance at or above grade level, while scores designated “non-proficient” represented performance below grade level.²⁸ The threshold of proficiency is set by the ADE. Cases and controls were linked to their annual ADE records using their social security numbers. Students who were home schooled, attended private schools, or moved out of Arkansas, did not take the Benchmark examination or could not be matched to their achievement test records and thus were excluded. A unique study identifier for each case and control and another identifier for the unique three-record, case-control matches were created. This final file was anonymized with key identifiers removed and securely transmitted to UAMS researchers for analysis. The primary outcome variables were odds of proficiency on third grade literacy and mathematics achievement tests.

Statistical Analysis

Descriptive summaries were represented by frequency distribution for categorical data and mean/standard deviation or median/interquartile range for continuous data based on data normality. We compared characteristics of cases and controls and proficient and non-proficient cases using T-Test/Mann-Whitney and Chi-square/Fischer’s exact tests for continuous and categorical data, respectively. Bivariate analyses were used to compare academic proficiency between surviving gastroschisis cases and controls and among cases with gastroschisis. Collinearity and correlation between variables were examined using variance inflation and correlation factors, respectively. Variables with P≤0.2 on bivariate analysis were considered as possible covariates in the multivariate logistic regression models, using a stepwise approach. We confirmed our results by re-analyzing the data using Firth logistic regression, a penalized likelihood based method, in order to provide bias-reduction for the small sample size of the study cohort.²⁹ Model validity was measured using the area under the curve of the Receiver Operator Characteristic. Statistical analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC).

RESULTS

Study Cohort

There were 105 pregnancies complicated with gastroschisis between January 2000 and December 2005 (Figure I). A total of 21 cases died (9 were stillbirths and 12 died shortly after delivery or later during childhood) (Table III; online). Eight infants were born in level II neonatal intensive care units (NICUs) near the Arkansas-Missouri border, could not be linked to ACH medical records, and were presumed to be transferred to tertiary neonatal centers in nearby states (Figure I). All remaining 76 cases of gastroschisis who survived until childhood and were transferred to ACH after birth, were successfully linked to their hospital discharge data for the first 6 years after birth. Almost 62% of gastroschisis cases (47/76) and 41% of unaffected controls (63/152) were successfully linked to their third grade literacy and mathematics achievement test scores.

Figure I.

Diagram of the Linkage of Gastroschisis Cases and Controls through the Multiple Databases

Study Cohort Characteristics

The neonatal and maternal characteristics of the study population are summarized in (Table II;online). Overall, 38% were delivered preterm (defined as <37 completed weeks’ gestation) and 22% were SGA. Approximately 20% (15/76) of patients had complex gastroschisis. The frequency of intestinal failure among all cases was 25%. The overall median duration of invasive ventilation and parenteral nutrition was 6 and 27 days, respectively. Approximately 34% of gastroschisis cases were readmitted within the first 6 years after birth. Neonatal and maternal characteristics, including maternal education and payer type, were comparable between gastroschisis cases who could be linked vs. not-linked to their education data.

The characteristics of gastroschisis cases and controls, who survived to third grade and were successfully linked to their education data, are summarized in Table I. Gastroschisis cases were more likely than controls to have a 5-minute Apgar score <7 (13% vs. 2%, P=0.04). While gestational age was similar, gastroschisis cases had lower birth weight (2618 vs. 3008 g, P=0.0002) and were thus more frequently SGA (17% vs. 5%, P=0.007) than controls. In addition, the mothers of gastroschisis cases, compared to controls, were younger (20 vs. 25 years, P<0.0001) and more frequently primiparous (94% vs. 79%, P=0.04).

Educational Outcomes

Forty-seven gastroschisis cases and 63 controls were successfully linked to their third grade test scores (Table IV). There was no statistically significant difference between cases and controls in unadjusted third grade achievement test literacy (68% vs. 81%, P=0.12) or mathematics proficiency (89% vs. 87%, P=0.74) (or raw scores, results not shown). The results remained unchanged once adjusted for the difference in characteristics between cases and controls.

Table IV.

Comparison of Third Grade Literacy and Mathematics Proficiency between Infants Born with Gastroschisis and Infants Born without Structural Defects in Arkansas, 2000–2005.

Education Data	Cases N=47	Controls N=63	a,bP-value Unadjusted	a,cP-value Adjusted
Proficiency on literacy achievement tests (yes), n (%)	32 (68)	51 (81)	0.12	0.65
Proficiency on mathematics achievement tests (yes), n (%)	42 (89)	55 (87)	0.74	0.15

^aLogistic regression was used to model the proficiency test results

^bUnadjusted P-values were obtained with observations being case/control as a predictor

^cAdjusted P-values were obtained using predictor variables: case/control, maternal age, parity, infant birth weight, size for gestation, and Apgar score at 5 minutes.

Bivariate and Multivariate Analysis of Gastroschisis Cases Based on Academic Proficiency

Approximately 68% (32/47) of patients with gastroschisis were proficient on third grade literacy achievement tests, while89% (42/47) were proficient in mathematics (Table V;online). Gastroschisis patients who were not proficient vs. proficient in literacy received more days of total parenteral nutrition (31 vs. 25 days, P=0.03), more frequently had ≥2 surgical procedures under general anesthesia (93% vs. 56%, P=0.01), and were more likely to have a complex defect (40% vs. 6%, P=0.009). On the other hand, we could not detect any statistically significant differences in characteristics of gastroschisis patients who were proficient vs. non-proficient in mathematics. Variables with P≤0.2 on bivariate analyses are summarized in Table VI. Patients with complex in comparison to simple gastroschisis had 10 times the odds of poor proficiency in literacy (AOR: 10.00, 95% CI: 1.71, 58.43; P=0.01). From multivariable modeling, a complex gastroschisis was the only significant predictor of poor proficiency in literacy (AOR=0.1, 95% CI: 0.02, 0.58; P=0.01) (area under the curve of the Receiver Operator Characteristic=0.78, 95% CI=0.65, 0.90). No neonatal or maternal characteristics predictive of lower proficiency in mathematics were identified. We also assessed the difference in proficiency in literacy in gastroschisis patients vs. controls, based on the complexity of the gastroschisis. The AOR of achieving proficiency in literacy vs. controls was lower in patients with complex (AOR: 0.18, 95% 0.02, 1.05; P=0.08) vs. simple gastroschisis (AOR: 0.64; 95% CI: 0.22, 1.82; P=0.40).

Table VI.

Bivariate Analysis among Children with Gastroschisis in Arkansas Based on Proficiency of Third Grade Literacy and Mathematics Achievement Tests^a

Characteristics	Proficiency on Third Grade Literacy Achievement Tests		Proficiency on Third Grade Mathematics Achievement Tests
	AOR (95% CI)	P-Value	AOR (95% CI)	P-Value
Birth weight	1.00 (0.98, 1.001)	0.51	1.00 (0.99, 1.01)	0.18
Type of defect		0.01		0.15
Complex	Reference		Reference
Simple	10.00 (1.71, 58.43)		4.01 (0.60, 26.94)
Age at full enteral feeding	0.96 (0.91, 1.007)	0.09	0.97 (0.94, 1.00)	0.08
Intestinal failure		0.17		0.26
No	Reference		Reference
Yes	0.37 (0.09, 1.56)		0.35 (0.05, 2.20)
Duration of total parenteral nutrition	0.98 (0.95, 1.00)	0.10	0.97 (0.94, 1.00)	0.07
Procedures under general anesthesia		0.03		0.68
1	Reference		Reference
≥2	0.13 (0.02, 0.86)		0.66 (0.09, 4.88)
Exclusive oral feeding at discharge		0.94		0.08
No	Reference		Reference
Yes	1.08 (0.17, 6.65)		6.11 (0.82, 45.32)
Duration of first hospitalization at ACH	0.98 (0.95, 1.01)	0.14	0.99 (0.95, 1.03)	0.59
Payer type		0.14		0.43
Medicaid or self-pay	Reference		Reference
Private	2.75 (0.72, 10.48)		0.49 (0.08, 2.88)

ACH, Arkansas Children’s Hospital; AOR, adjusted odds ratio; CI, confidence interval; Ref, reference; SGA, small for gestational age.

^aOnly variables with P≤0.2 on bivariate analysis of proficiency in literacy or mathematics are presented in this table.

DISCUSSION

We linked ARHMS data, the Arkansas vital statistics and hospital discharge databases, the state’s only tertiary care children’s hospital medical records, and third grade school achievement test results to examine the academic performance of infants born with gastroschisis in Arkansas over a 6-year period (2000–2005) and identify predictors of their third grade proficiency in literacy and mathematics. Contrary to expectations, third grade academic performance was not different between children born with gastroschisis and controls. It is notable that a complex gastroschisis (defined as atresia, volvulus, necrosis, or perforation of the bowel) was the only factor associated with poor proficiency in literacy among these children. Although not statistically significant, the AOR of achieving proficiency in literacy among children born with gastroschisis vs. controls was lower in patients with complex vs. simple gastroschisis.

The clinical characteristics of patients with gastroschisis have been previously described in the literature.³⁰ Consistent with previous reports, many of the cases were SGA, moderately preterm, and born to Caucasian young primiparous mothers.^30,31 The risk of re-hospitalization among gastroschisis patients varied in the literature.³² In our study, 34% of cases had gastroschisis-related readmissions within the first 6 years with a median of 2 readmissions and a length of stay ranging from 2 to 130 days.

Studies that specifically assessed the neurodevelopmental outcomes of these infants included small sample sizes (n<25),^16,33 combined different types of abdominal wall defects,^17,22,34 assessed non-contemporaneous cohorts,^22,34 or limited their analysis to simple gastroschisis.³⁵ In addition, these studies mostly assessed these infants at pre-school age (>1–≤5 years) ^33,35–37 or compared their neurodevelopment evaluation to norms rather than control populations.^18,20 Some of these reports described similar intelligence^20,21 or need of special education services of infants born with gastroschisis in comparison to their age peers,³⁸ while others reported higher frequency of deficits¹⁸ or parental-reported developmental delay¹⁹ as these children grew older. We could only identify two studies that compared the neurodevelopment of gastroschisis survivors to matched controls at school age. Lap et al examined 16 infants with gastroschisis and matched controls and reported significantly lower scores on verbal and total intelligence, response inhibition, selective visual attention, and sustained auditory attention in the gastroschisis group.¹⁶ Hijkoop et al also reported significantly lower Pediatric Perceived Cognitive Function Scores among 23 children with gastroschisis compared to matched controls.²³ The difference with our findings of comparable academic performance between cases and controls may be attributed to the higher proportion of children born SGA in Lap’s study (22 vs. 44%)¹⁶ or possibly to parents underestimating their children’s cognitive abilities in Hijkoop’s report.²³

Several risk factors were previously identified to increase the risk of adverse neurodevelopmental outcomes among infants born with gastroschisis. Prematurity has consistently been associated with higher morbidities in this population.^35,36,39,40 Van Manen et al described that the odds of death or early childhood neurodevelopmental impairment was 0.6 times lower for each additional gestational week among 39 gastroschisis survivors.³⁶ Low birth weight,³⁹ associated congenital anomalies,³⁸ intestinal failure,²³ and complex gastroschisis^37,40 were also associated with higher risk of neurodevelopment impairment. Furthermore, major surgery and anesthesia during the first weeks after birth, a time of rapid brain growth, had been linked with developmental delay.⁴¹ In our study we found that a complex gastroschisis was the only significant predictor of poor performance on third grade literacy achievement tests. Previous studies have shown that infants with complex gastroschisis typically have associated gastrointestinal anomalies, are at high risk of intestinal failure, and may require higher number of procedures under general anesthesia.¹

In addition, although reading and mathematics skills share overlapping cognitive processes (e.g., encoding of visual stimuli), they also require distinct abilities (e.g., quantity processing). A study, using magnetic resonance imaging, determined that different cortical gray matter regions are preferentially engaged in reading vs. mathematics.⁴² The cortical fascicles required in one activity vs. the other appear to have structural differences in their myelination.⁴² Thus, it is possible that specific events in childhood may adversely affect various brain circuits and is only associated with impairment in literacy rather than mathematics abilities or vice versa.

Our study has multiple strengths including the examination of detailed neonatal hospital course, socioeconomic markers (payer type and maternal education), and hospital readmission data among cases with gastroschisis. In addition, unlike previous studies that relied on aptitude tests, we used standardized school achievement tests which are measures closely aligned with “real-world” outcomes such as high school graduation and college attendance.²⁴ Limitations, however, remain. Our sample size was small as gastroschisis is a rare birth defect. We only could link 62% of our study cohort to their education data, most likely because some moved out of state, enrolled in private school or were home schooled. However, the characteristics of the linked and unlinked patients were comparable. Although we evaluated in our analysis the total number of procedures requiring general anesthesia, we lacked data on the total amount of anesthetic exposure in these patients. In addition, we could not assess the childhood growth measures of patients with gastroschisis. Nonetheless, our study provides clinicians and families with the largest evidence-based study regarding the academic performance at school age of children with gastroschisis.

CONCLUSION

As survival of infants with birth defects has increased, it is now essential to assess the long-term outcomes of these patients. Many researchers, including investigators from Arkansas, have previously examined the educational outcomes of infants born with different birth defects such as congenital heart disease and cleft palate and lip and have consistently concluded that these children have worse academic performance than their peers ^43,44 On the other hand, we determined that third grade academic performance was not different between children with gastroschisis and controls. In addition, a complex gastroschisis in these children was significantly associated with poor proficiency on literacy achievement tests. Prenatal and pre-discharge counseling of infants with gastroschisis should include academic performance. Future studies should investigate educational outcomes among children with gastroschisis using population-based data with a larger number of gastroschisis cases. In addition, further evaluation is needed to determine factors that might improve the academic performance among patients with complex gastroschisis and other birth defects.

Abbreviations

ACH

Arkansas Children’s Hospital

ADE

Arkansas Department of Education

ADH

Arkansas Department of Health

AHIDD

Arkansas Hospital Inpatient Discharge Database

AGA

appropriate for gestational age

ARHMS

Arkansas Reproductive Health Monitoring System

CI

confidence interval

ICD-CM

International Classification of Diseases-Clinical Modification

IQR

interquartile range

LGA

large for gestational age

NICU

Neonatal Intensive Care Unit

SES

socioeconomic status

SGA

small for gestational age

UAMS

University of Arkansas for Medical Sciences

Data Statement:

All relevant data are within the paper and its supporting files.