There is a lack of current, multi-institutional data regarding hospital outcomes of infants with interrupted aortic arch (IAA). We analyzed the Pediatric Health Information System database to identify infants with IAA who underwent repair during 2004 to 2019. We classified patients as simple or complex based on associated heart defects. We evaluated factors associated with hospital mortality and complications related to 22q11.2 deletion syndrome (22q11.2del) using mixed logistic regression, accounting for hospital clustering. In 1,283 infants included (904 simple, 379 complex), mortality was higher in the complex group (11.7% vs 4.4%, p <0.001). Factors associated with mortality in the simple group were low birth weight (adjusted risk ratio [aRR] 3.77, 95% confidence interval [95% CI] 1.83 to 7.77), non-22q11.2del genetic conditions (aRR 6.44, 95% CI 1.73 to 23.96), and gastrointestinal anomalies (aRR 8.47, 95% CI 3.12 to 22.95), whereas surgery between 2012 and 2015 (aRR 0.36, 95% CI 0.13 to 0.99) was protective. In the complex group, factors associated with mortality were male (aRR 2.32, 95% CI 1.10 to 4.24) and central nervous system anomalies (aRR 3.73, 95% CI 1.62 to 8.59). Compared with their nonsyndromic counterparts, infants with simple IAA and 22q11.2del were at higher risk of sepsis (aRR 1.63, 95% CI 1.02 to 2.39) and gastrostomy tube placement (aRR 3.18, 95% CI 2.13 to 4.74), and infants with complex IAA and 22q11.2del were at higher risk of gastrostomy tube placement (aRR 2.42, 95% CI 1.20 to 4.88). In conclusion, presence of complex cardiac lesions is associated with increased mortality after IAA repair. The co-occurrence of extracardiac congenital anomalies and non-22q11.2del genetic conditions elevates mortality risk. Presence of 22q11.2del is associated with hospital complications.
Interrupted aortic arch (IAA) is a conotruncal defect characterized by a lack of continuity between the ascending and descending segments of the thoracic aorta. It is a rare congenital heart defect with a prevalence of 0.62 per 10,000 live births. Approximately 27% to 45% of patients with IAA have 22q11.2 deletion syndrome (22q11.2del). IAA presents in conjunction with simple and complex heart defects. Based on single-institution studies, the hospital mortality of infants who underwent repair of IAA is 3% to 5% in the presence of simple associated heart defects and 7% to 16% in the presence of complex associated heart defects. , , Publications from multi-institutional studies are limited to patients operated before 1997. , We undertook this study to examine the present multi-institutional hospital outcomes of infants born with IAA. We aimed to determine factors associated with hospital mortality in patients with IAA with and without complex heart defects and to investigate the association between 22q11.2del and hospital complications. We used a large multi-institutional database with data collected between 2004 and 2019. We hypothesized that hospital mortality in patients with IAA and complex associated heart defects continues to be higher than in infants with simple associated heart defects, and 22q11.2del is associated with increased mortality and risk of complications in patients with IAA.
We performed a retrospective review of hospitalizations contained in the Pediatric Health Information System (PHIS) database. PHIS is an administrative database that contains inpatient, emergency department, ambulatory surgery, and observation encounter-level data from over 50 not-for-profit, tertiary care, pediatric hospitals in the United States. These hospitals are affiliated with the Children’s Hospital Association (Lenexa, Kansas). Data quality and reliability are assured through a joint effort between the Children’s Hospital Association and participating hospitals. For the purposes of external benchmarking, participating hospitals provide discharge/encounter data, including demographics, diagnoses, and procedures. Nearly all of these hospitals also submit resource utilization data (e.g., pharmaceuticals, imaging, and laboratory) into PHIS. Data are deidentified at the time of data submission, and data are subjected to several reliability and validity checks before being included in the database. For this study, data from 45 hospitals were included. The use of the PHIS database is not considered human subject research, and it is exempted from review by the Baylor College Medicine Institutional Review Board. Data from January 2004 through December 2019 were used for this study.
We used the International Classification of Disease Ninth Revision (ICD-9) and International Classification of Disease Tenth Revision (ICD-10) diagnostic and procedure codes to identify participants for this study. The inclusion criteria were hospitalizations with diagnostic codes for IAA (ICD-9 code 747.11 or ICD-10 code Q25.21) and congenital heart defects typically associated with IAA (ventricular septal defect, truncus arteriosus, or aortopulmonary window), hospital admission within the first 28 days of life, and procedure codes indicating aortic arch repair during the index hospitalization. We limited the cohort to patients with ventricular septal defect, truncus arteriosus, or aortopulmonary window to keep patients with conotruncal-type IAA and not contaminate the cohort with patients with severe coarctation or aortic arch atresia that may have been coded as IAA.
The exclusion criteria were nonsurgical hospitalizations, and hospitalizations with International Classification of Disease (ICD) diagnostic codes corresponding to pulmonary atresia, double aortic arch, or more complex single ventricle anatomy (hypoplastic left heart syndrome [ICD-9 code 746.01, ICD-10 Code Q23.4], hypoplastic right heart syndrome [ICD-9 code 746.1, ICD-10 codes Q22.4 and Q22.6], common ventricle [ICD-9 codes 745.3 and 745.7], double inlet left ventricle [ICD-10 codes Q20.4]). We excluded hospitalizations without mortality data from the mortality analyses (n = 4). The infants were classified based on the complexity of their heart defects; the simple group included infants with ventricular septal defect and/or atrial septal defect only, and the complex group comprised infants with other heart defects.
Demographic characteristics were extracted from the PHIS database, including date of birth, gender, birth weight, race/ethnicity, gestational age, insurance type, region of the United States where care was provided, and year of birth. We used deidentified hospital codes included in the PHIS database to determine the hospital for each patient in the sample. Low birth weight was defined as birth weight <2,500 g. Prematurity was defined as gestational age at birth <37 weeks. We used Fenton growth charts to calculate weight for gestational age z score and percentiles to define small for gestational age (<10th percentile). For patients without available gestational age or birth weight, we used ICD diagnostic codes for prematurity, small for gestational age, and low birth weight to determine if they had these characteristics (Supplementary Table 1). Demographic data were complete for the entire study sample except for birth weight, gestational age, weight for age z score, and race/ethnicity (available in 94%, 75%, 74%, and 68% of the study sample, respectively).
Heart defects, genetic conditions, extracardiac congenital anomalies, and complications were identified using ICD codes (Supplementary Table 1). We used ICD diagnostic codes 279.11 and 758.32, and ICD-10 codes D82.1 and Q93.81 to define presence of 22q11.2del. A nonsyndromic patient was defined as a patient without a codable genetic condition listed in the Supplementary Table 1. Cardiac and noncardiac procedures were identified using ICD procedure codes (Supplementary Tables 2 and 3). We also used proxy associations of ICD codes to infer that a specific surgical repair took place (as long as the dates of service of the procedure codes were the same). For example, we inferred that a patient underwent Yasui procedure if the patient had ICD procedure codes for arch repair, ventricular septal defect repair, and right ventricle to pulmonary artery conduit on the same date (Supplementary Table 2). If an initial cardiac surgery included ventricular septal defect closure and aortic arch repair, it was considered as single-stage 2-ventricle repair.
To calculate hospital surgical volumes, we performed a separate query in the PHIS database that included hospitalizations with ICD procedure codes for aortic arch repair, ventricular septal defect repair, truncus arteriosus repair, arterial switch operation, atrioventricular septal defect repair, Tetralogy of Fallot repair, shunt placement, and cardiopulmonary bypass (Supplementary Table 2). For each hospital, we calculated surgical volume as the average of the annual number of hospitalizations of infants <28 days of life who underwent any of the procedures listed above between 2004 and 2019.
The primary outcome was hospital mortality. The secondary outcomes were length of stay (LOS) and hospital costs. Hospital mortality was defined as death before hospital discharge. Hospital costs were calculated by multiplying the adjusted billing charges reported in the PHIS database by the center- and year-specific cost-to-charge ratios to account for cost variations across the hospitals included in the sample. The hospital costs were further standardized to $2,019 using the United States Bureau of Labor Statistics Consumer Price Index Research Series ( http://www.bls.gov/cpi/cpiurs.htm ). We used ICD diagnostic and procedure codes to identify hospital complications (Supplementary Table 1).
We compared the infants in the simple and complex groups by demographic and clinical characteristics. Categoric variables were compared using chi-square test or Fisher’s exact test, as appropriate. Continuous variables were compared using Wilcoxon sum-rank test. We then performed bivariate analyses to compare demographic and clinical characteristics between survivors and nonsurvivors in the entire patient cohort and within the individual IAA groups to identify factors associated with hospital mortality. Variables with p <0.1 in the bivariate analyses were included in multivariate analyses to evaluate adjusted associations with hospital mortality. If there was collinearity between variables, only one of the collinear variables was included in the multivariate analyses. We used Spearmen correlation coefficients (ρ) to evaluate the relation between surgical volume and hospital mortality. Mixed-effects log-binomial multivariate regression models with backward elimination and hospital as a random effect were used to identify factors associated with hospital mortality in the entire study cohort. An interaction term between group (simple vs complex) and year of hospitalization was tested and was found not to be statistically significant. Subsequently, mixed-effects log-binomial multivariate regression with backward elimination and hospital as a random effect was performed within each IAA group (simple and complex). Study era (2004 to 2007, 2008 to 2011, 2012 to 2015, and 2016 to 2019) was included in all the multivariate analyses for mortality regardless of significance. Hospital mortality rates for the IAA groups were then calculated by era and displayed in graphs. A subanalysis was performed to determine if patients with complex IAA grouped by era were different based on demographic and clinical characteristics using chi-square test and Kruskal-Wallis test. We compared secondary outcomes (LOS and hospital costs) between the infants in the simple and complex groups using Wilcoxon sum-rank test.
Bivariate subanalyses were performed within each IAA group to compare hospital outcomes between infants with 22q11.2del and nonsyndromic infants using chi-square test and Wilcoxon sum-rank test for categoric and continuous variables, respectively. As mortality was found to be similar in patients with 22q11.2del and nonsyndromic infants, we evaluated if 22q11.2del was associated with increased risk of hospital complications. Mixed-effects log-binomial multivariate regression analyses accounting for age at the time of initial cardiac surgery >14 days, low birth weight, gender, and insurance type as fixed effects, and hospital as a random effect were used to calculate risk ratios for hospital complications in infants with 22q11.2del within each IAA subgroup, using nonsyndromic infants as the reference group. These analyses were adjusted for demographic characteristics. Statistical analyses were performed using SAS University Edition 3.8 (SAS Institute Inc. Cary, North Carolina), R version 3.5.1 (University of Auckland, Auckland, New Zealand), and R Studio Version 1.1.423 (Boston, Massachusetts). Type-I error rate for all statistical tests was set at 5%.
We identified a total of 1,283 infants with IAA that met inclusion criteria and underwent surgical repair from 2004 to 2019 (simple n = 904, 70%; complex n = 379, 30%). The characteristics, hospital complications, and outcomes of the simple and complex IAA groups are shown in Table 1 . Birth weight and birth weight z score for gestational age were lower in the simple group compared with the complex group, although prematurity was more common in the complex group. Genetic conditions were coded more frequently in simple IAA compared with complex IAA, although this was almost completely driven by differential presence of 22q11.2del. Approximately a third of the patients within each group had extracardiac congenital anomalies. Median LOS was similar between the simple and complex groups although hospitalization costs were significantly higher in the complex group. There were 88 hospital deaths in this cohort, for a hospital mortality rate of 6.8%. Hospital mortality for the complex group was significantly higher than in the simple group (11.7% vs 4.4%, p <0.001).
| Variable | Simple IAA (n = 904) | Complex IAA (n = 379) | p | ||
|---|---|---|---|---|---|
| Age at initial cardiac surgery (days) | 11 | ± 13 | 10 | ± 17 | 0.019 |
| Male | 475 | (52.5%) | 210 | (55.4%) | 0.531 |
| * Gestational age (weeks) | 39 | ± 2 | 38 | ± 2 | 0.248 |
| (n = 686) | (n = 277) | ||||
| * Birth weight (grams) | 2970 | ± 624 | 3070 | ± 671 | 0.014 |
| (n = 851) | (n= 353) | ||||
| * Weight for gestational age (z score) | -0.29 | ± 1.23 | -0.03 | ± 1.17 | 0.001 |
| (n = 677) | (n = 271) | ||||
| Low birth weight | 176 | (19.4%) | 62 | (16.4%) | 0.191 |
| Prematurity | 338 | (37.4%) | 168 | (44.3%) | 0.020 |
| Small for gestational age | 132 | (14.6%) | 42 | (11.1%) | 0.093 |
| * Race – ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Others Missing | 319 100 163 32 290 | (35.3%) (11.1%) (18.0%) (3.5%) (32.1%) | 139 30 68 22 120 | (36.7%) (7.9%) (17.9%) (5.8%) (31.7%) | 0.197 |
| * Insurance type Public Other | 487 417 | (54.2%) (46.4%) | 204 175 | (54.1%) (46.4%) | 0.952 |
| (n = 898) | (n = 377) | ||||
| US region Midwest Northeast Southeast Southwest West | 222 134 162 115 271 | (24.6%) (14.8%) (17.9%) (12.7%) (30.0%) | 104 54 76 51 94 | (27.4%) (14.2%) (20.1%) (13.5%) (24.8%) | 0.378 |
| Hospital surgical volume (n) | 83 | ± 35 | 82 | ± 34 | 0.281 |
| Study era 2004 – 2007 2008 – 2011 2012 – 2015 2016 – 2019 | 200 241 228 215 | (22.1%) (26.7%) (25.2%) (23.8%) | 95 88 108 88 | (25.1%) (23.2%) (28.5%) (23.2%) | 0.490 |
| Genetic conditions | |||||
| Any genetic condition | 437 | (48.3%) | 95 | (25.1%) | <0.001 |
| 22q11.2 deletion syndrome | 404 | (44.7%) | 72 | (19.0%) | <0.001 |
| Trisomy 21 | 3 | (0.3%) | 11 | (2.9%) | <0.001 |
| Turner syndrome | 3 | (0.3%) | 1 | (0.3%) | 0.842 |
| Other unspecified genetic anomalies | 29 | (3.2%) | 11 | (2.9%) | 0.774 |
| Congenital anomalies | |||||
| Any extracardiac congenital anomaly | 265 | (29.3) | 126 | (33.2) | 0.163 |
| Central nervous system anomalies | 48 | (5.3%) | 39 | (10.3%) | <0.001 |
| Respiratory system anomalies | 118 | (13.1%) | 46 | (12.1%) | 0.654 |
| Gastrointestinal system anomalies | 33 | (13.1%) | 15 | (4.0%) | 0.791 |
| Genitourinary system anomalies | 128 | (14.2%) | 54 | (14.2%) | 0.967 |
| Hospital complications | |||||
| Cardiac arrest | 28 | (3.1%) | 20 | (5.3%) | 0.061 |
| Arrhythmia | 168 | (18.6%) | 55 | (14.5%) | 0.079 |
| Extracorporeal membrane oxygenation | 67 | (7.4%) | 48 | (12.7%) | 0.003 |
| Neurologic complications | 92 | (10.2%) | 41 | (10.8%) | 0.731 |
| Renal complications | 191 | (21.1%) | 101 | (26.6%) | 0.031 |
| Gastrointestinal complications | 74 | (8.2%) | 32 | (8.4%) | 0.879 |
| Sepsis | 155 | (17.1%) | 61 | (16.1%) | 0.646 |
| Hospital outcomes | |||||
| Length of stay (days) | 27 | ( 18 – 46) | 28 | (18 – 49) | 0.523 |
| Hospital costs, 2019 (US dollars) | 238,363 | ± 259,208 | 271,458 | ± 277,435 | 0.029 |
| * Hospital mortality | 40 | (4.4%) | 44 | (11.7%) | <0.001 |
| (n = 903) | (n = 376) | ||||
⁎ For variables with missing data the corresponding n is shown within parentheses under the reported values. p ≤0.05 is statistically significant.
Single-stage 2-ventricle repair was the initial surgical approach utilized in 750 infants (83%) with simple IAA and 305 infants (80%) with complex IAA ( Table 2 ). For infants in the simple group, those that underwent single-stage 2-ventricle repair had lower hospital mortality than those that did not (3.3% vs 9.7%, p = 0.001). For infants in the complex group, hospital mortality was the same regardless of undergoing single-stage 2-ventricle repair or other intervention as the first cardiac surgery (11.8% vs 10.8%, p = 0.811).
| First cardiac surgery by IAA group | Total patients | Total deaths | * Mortality |
|---|---|---|---|
| IAA with simple associated heart defects | (n = 904) | (n = 40) | |
| Single-stage two-ventricle repair | 750 (83.0%) | 25 (62.5%) | 3.3% |
| Other | 154 (17.0%) | 15 (37.5%) | 9.7% |
| IAA with complex associated heart defects | (n = 379) | (n = 44) | |
| Single-stage two-ventricle repair | 305 (80.5%) | 36 (81.8%) | 11.8% |
| Other | 74 (19.5%) | 8 (18.2%) | 10.8% |
| First cardiac surgery by specific diagnosis: | |||
|---|---|---|---|
| IAA and VSD: | (n = 904) | (n = 40) | 4.4% |
| Single-stage two-ventricle repair: | |||
| Arch repair + VSD repair | 717 (79.3%) | 22 (55%) | |
| Yasui procedure | 33 (3.7%) | 1 (2.5%) | |
| Norwood procedure | 59 (6.5%) | 5 (12.5%) | |
| Other palliation or partial repair: | |||
| Arch repair | 71 (7.9%) | 9 (22.5%) | |
| Arch repair and ASD repair | 17 (1.9%) | 1 (2.5%) | |
| Arch repair and pulmonary artery banding | 4 (0.4%) | 1 (2.5%) | |
| Pulmonary artery banding | 2 (0.2%) | 0 (0.0%) | |
| Pulmonary artery banding and PDA stent | 1 (0.1%) | 1 (2.5%) | |
| IAA and truncus arteriosus | (n = 183) | (n = 22) | 12.1% |
| Arch repair + truncus arteriosus repair | 181 (98.9%) | 22 (100%) | |
| Pulmonary artery banding | 2 (1.1%) | 0 (0.0%) | |
| IAA and DORV | (n = 44) | (n = 2) | 4.6% |
| Single-stage two-ventricle repair: | |||
| Arch repair + VSD repair | 20 (44.5%) | 1 (50%) | |
| Yasui procedure | 1 (2.3%) | 0 (0.0%) | |
| Norwood procedure | 9 (20.5%) | 1 (50%) | |
| Other palliation or partial repair: | |||
| Arch repair | 13 (29.5%) | 0 (0.0%) | |
| Arch repair + pulmonary artery banding | 1 (2.3%) | 0 (0.0%) | |
| IAA and Taussig-Bing anomaly | (n = 46) | (n = 4) | 8.7% |
| Single-stage two-ventricle repair: | |||
| Arch repair + arterial switch + VSD repair | 32 (69.6%) | 1 (25%) | |
| Yasui procedure | 1 (2.2%) | 0 (0.0%) | |
| Arch repair + VSD repair | 1 (2.2%) | 1 (25%) | |
| Norwood procedure | 2 (4.3%) | 0 (0.0%) | |
| Other palliation or partial repair: | |||
| Arch repair and pulmonary artery banding | 5 (10.9%) | 1 (25%) | |
| Arch repair | 4 (8.7%) | 0 (0.0%) | |
| Pulmonary artery banding | 1 (2.2%) | 1 (25%) | |
| IAA and AVSD | (n = 40) | (n = 7) | 17.9% |
| Single-stage two-ventricle repair: | |||
| Arch repair + AVSD repair | 29 (72.5%) | 5 (71.4%) | |
| Other palliation or partial repair: | |||
| Arch repair | 8 (20.0%) | 0 (0.0%) | |
| Arch repair and pulmonary artery banding | 2 (5.0%) | 1 (14.3%) | |
| Pulmonary artery banding and PDA stent | 1 (2.5%) | 1 (14.3%) | |
| IAA, TGA and VSD | (n = 24) | (n = 5) | 20.8% |
| Single-stage two-ventricle repair: | |||
| Arch repair + arterial switch + VSD repair | 15 (62.5%) | 4 (80.0%) | |
| Arch repair + VSD repair | 2 (8.3%) | 0 (0.0%) | |
| Norwood procedure | 4 (16.7%) | 1 (20.0%) | |
| Other palliation or partial repair: | |||
| Arch repair + arterial switch | 2 (8.3%) | 0 (0.0%) | |
| Arch repair and pulmonary artery banding | 1 (4.2%) | 0 (0.0%) | |
| IAA and APW | (n = 12) | (n = 0) | 0% |
| Arch + APW repair + ASD closure | 10 (83.3%) | 0 (0.0%) | |
| Arch + APW repair | 2 (16.7%) | 0 (0.0%) | |
| IAA, DORV and AVSD | (n = 8) | (n = 0) | 0% |
| Arch repair + AVSD repair | 4 (50.0%) | 0 (0.0%) | |
| Norwood procedure | 3 (37.5%) | 0 (0.0%) | |
| Arch repair | 1 (12.5%) | 0 (0.0%) | |
| IAA and TOF | (n = 6) | (n = 0) | 0% |
| Arch repair + TOF repair | 1 (16.7%) | 0 (0.0%) | |
| Arch repair | 5 (83.3%) | 0 (0.0%) | |
| IAA, VSD and PAPVR | (n = 4) | (n = 0) | 0% |
| Arch + ASD repair + VSD repair | 4 (100%) | 0 (0.0%) | |
| IAA, VSD, TAPVR | (n = 3) | (n = 1) | 33.3% |
| Arch repair + VSD repair + TAPVR repair | 2 (66.7%) | 0 (0.0%) | |
| Norwood procedure + TAPVR repair | 1 (33.3%) | 1 (100%) | |
| IAA, CCTGA and VSD | (n = 3) | (n = 1) | 50% |
| Norwood procedure | 2 (66.7%) | 1 (100%) | |
| Arch repair + VSD repair | 1 (33.3%) | 0 (0.0%) | |
| IAA, VSD and Ebstein anomaly | (n = 2) | (n = 1) | 50% |
| Arch repair + VSD repair | 1 (50.0%) | 1 (100%) | |
| Arch repair | 1 (50.0%) | 0 (0.0%) | |
| IAA, truncus arteriosus and AVSD | (n =2) | (n = 1) | 50% |
| Arch repair + truncus arteriosus repair + AVSD repair | 1 (50.0%) | 1 (100%) | |
| Arch repair + truncus repair | 1 (50.0%) | 0 (0.0%) | |
| IAA and DOLV | (n = 2) | (n = 0) | 0% |
| Arch repair + VSD repair | 1 (50.0%) | 0 (0.0%) | |
| Yasui procedure | 1 (50.0%) | 0 (0.0%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
