Abstract
Objectives:
To study the demographic and clinical characteristics of preterm infants with BPD to determine what factors are most strongly predictive of outpatient mortality with the goal of identifying those individuals at greatest risk.
Study design:
Demographic and clinical characteristics were retrospectively reviewed for subjects (n=862) recruited from an outpatient bronchopulmonary dysplasia (BPD) clinic. Characteristics of the deceased and living participants were compared using nonparametric analysis. Regression analysis was performed to identify factors associated with mortality.
Results:
Of the 862 subjects, 13 (1.5%) died during follow-up for an overall mortality rate of approximately 15.1 deaths per 1,000 subjects. Two patients died in the post-neonatal period (annual mortality incidence (AMI)=369.9 per 100,000), 9 died between 1 and 4 years of age (AMI=310.2 per 100,000), and 2 died between the ages of 5 and 14 (AMI=71.4 per 100,000). After adjusting for gestational age and BPD severity, mortality was found to be associated with the amount of supplemental oxygen required at NICU discharge (aHR: 4.10, p=0.001), presence of a gastrostomy tube (aHR: 8.13, P = .012), and presence of a CSF shunt (aHR: 4.31, p=0.021).
Conclusions:
The incidence of mortality among preterm infants with BPD is substantially higher than that seen in the general population. Need for greater amounts of home supplemental oxygen, gastrostomy tubes, and CSF shunts were associated with an increased risk of post-discharge mortality. Future studies should focus on clarifying risk factors for the development of severe disease to allow for early identification and treatment of those at highest risk.
Infant and childhood mortality are important public health measures that indicate the health of a population. U.S. infant mortality has declined from a recent high of 6.75 deaths per 1,000 births in 2007 to a historic low of 5.67 in 2018.1,2 Infant mortality is further classified into neonatal mortality, or deaths that occur between 0 and 28 days of life, and post-neonatal mortality, defined as death occurring between 28 and 364 days of age with two-thirds of infant deaths occurring between 0 and 28 days of life.2
Preterm birth, or birth at less than or equal to 37 weeks gestation, is the second leading cause of infant death in the U.S. and the leading cause of death in children under the age of five years worldwide.2,3, Infants born preterm have an increased risk of mortality in both the neonatal and post-neonatal periods compared with infants born full-term, with poorer survival correlated with decreasing gestational age and birth weight.2 Longitudinal studies have further illustrated that infants discharged from neonatal intensive care units (NICUs) have increased rates of mortality that persists through the post-neonatal period, suggesting that the factors contributing to preterm birth continue to play a role in this population as they age.4,5 Additionally, findings from previous studies 6,7,8 performed in high-resource settings suggest that the morbidities associated with prematurity, which result in prolonged hospitalizations, continue to play an important role in the development of subsequent complications and mortality in these infants after discharge.
Bronchopulmonary dysplasia (BPD), the most common cause of chronic lung disease in infancy, is characterized by impaired alveolar growth, airway inflammation, prolonged supplemental oxygen dependence, tachypnea, and airflow obstruction.9 Infants with BPD are at increased risk of significant morbidities such as rehospitalization, cardiovascular sequelae, growth failure, and poor neurodevelopmental outcomes when compared with term and other preterm infants.10–12
Preterm infants, and especially those with additional co-morbidities, are at an increased risk of post-discharge mortality than term infants. However, it is unclear what conditions in early life contribute to an increased risk of mortality in infants with BPD after initial hospital discharge and how far into childhood this risk persists. The aim of this study was to investigate demographic and clinical characteristics of preterm infants with BPD to determine what factors are most strongly predictive of outpatient mortality in this population. These data would help clinicians identify and intervene in those individuals at greatest risk.
METHODS
Population and Study Design:
A retrospective chart review was performed for subjects (n=862) who were recruited from an outpatient BPD clinic between January 2008 and March 2021. All subjects were born at less than or equal to 32 weeks gestation and diagnosed with BPD per the NHLBI consensus statement.13 At the time of enrollment in the study, all subjects were outpatients at a single center for BPD care, but they received their neonatal care at a variety of NICUs across the state of Maryland prior to referral to the outpatient BPD clinic for follow-up. All caregivers were consented per study protocol as approved by the Johns Hopkins University Institutional Review Board (Protocol NA_00051884).
Demographic and Clinical Data:
Demographic and clinical data were obtained through chart review for birth measures, clinical characteristics, and circumstances surrounding death including age at death, location and cause of death, home respiratory support, primary diagnoses, and medications at time of death. Median household income was estimated based on residential zip code and 2019 American Community Survey data from the US Census. Diagnosis of pulmonary hypertension was based on clinically-obtained echocardiogram at 37 weeks corrected age.
Statistical analyses:
Annual mortality incidence was calculated for the study population in three different age groups (post-neonatal, 1-4 years of age, and 5-14 years of age) to compare with the general population (most recent data available were utilized). Nonparametric analyses (Fisher exact and Wilcoxon rank-sum tests) were used to compare demographic and clinical features between the deceased portion of the population (n=13) and the living portion of the population (n=849). Kaplan-Meier time-to-event analysis was used to illustrate time to death in the study population. Cox regression analysis was performed with mortality as the dependent variable and individual predictors that were associated with or trending towards association with mortality in initial analysis as independent variables in univariate models. Cox regression analyses were adjusted for gestational age as well as BPD severity. Multivariate Cox regression analysis was performed by dropping non-significant predictors identified as significant in univariate analysis in a stepwise fashion. STATA IC 15 (StataCorp LP) was used for analysis. P-values <0.05 were considered statistically significant.
RESULTS
Of the 862 subjects in the outpatient registry, 13 (1.5%) died during follow-up after NICU discharge for an overall mortality rate of approximately 15.08 deaths per 1,000 subjects. Median age at time of death was 2.47 years; the median follow-up time for survivors was 7.79 years. A total of two subjects died in the post-neonatal period for an annual mortality incidence of 369.9 deaths per 100,000; 9 subjects died between one and four years of age for an annual mortality incidence of 310.2 deaths per 100,000, and 2 subjects died between the ages of five and fourteen years for an annual mortality incidence of 71.4 deaths per 100,000. As demonstrated in Table I, the annual mortality incidence was increased in the study population in every age group as compared with general population data with an incidence rate ratio of 1.96 for the post-neonatal period, 12.92 for ages one to four years, and 5.37 for ages five to fourteen years.14
Table 1.
Mortality Incidence
| Age Range | BPD Study Population | General Population | Incidence Rate Ratio | |||
|---|---|---|---|---|---|---|
| Number of Deaths after NICU Discharge | Observed Subject Time* (years) | Individuals Included in Analysis (n) | Annual Mortality Incidence (per 100,000) | Annual Mortality Incidence (per 100,000)16 | ||
| 28 - 364 days | 2 | 540.72 | 840 | 369.9 | 189 | 1.96 |
| 1 - 4 years | 9 | 2901.33 | 843 | 310.2 | 24.0 | 12.92 |
| 5 - 14 years | 2 | 2801.61 | 584 | 71.4 | 13.3 | 5.37 |
Time after NICU discharge
The deceased population was 53.9% female and had an average gestational age of 27.5 ± 3.7 weeks. 76.9% of deceased subjects identified as non-white race/ethnicity. The average median household income was $67,000 ± $23,800, and 53.9% had public insurance. Analysis (Table 2) illustrated that the deceased population was not significantly different from the living population in terms of these demographic characteristics.
Table 2.
Study Population
| Mean ± S.D. [Range] | Study Population (n = 862) | Living (n = 849) | Deceased (n = 13) | P value | ||
|---|---|---|---|---|---|---|
| Demographic Characteristics | Sex (% female) | 42.8% | 42.6% | 53.9% | 0.57 | |
| Gestational age (weeks) | 26.8 ± 2.6 [22.3, 36.9] | 26.8 ± 2.6 [22.3, 36.9] | 27.5 ± 3.7 [24.0, 36.0] | 0.77 | ||
| Race/ethnicity (% non-white) | 64.0% | 63.8% | 76.9% | 0.40 | ||
| Median household income ($’000s) | 76.3 ± 26.4 [28.5, 186.6] | 76.4 ± 26.4 [28.5, 186.6] | 67.0 ± 23.8 [31.9, 125.6] | 0.16 | ||
| Public insurance (% yes) | 55.9% | 56.0% | 53.9% | 1.00 | ||
| Clinical Characteristics | Supplemental oxygen at NICU discharge (% yes) | 42.7% | 42.2% | 76.9% | 0.020 | |
| Oxygen amount at NICU discharge if on oxygen (LPM) | 0.38 ± 0.37 [0.03, 2.00] (n = 368) | 0.36 ± 0.35 [0.03, 2.00] (n = 358) | 0.87 ± 0.75 [0.06, 2.00] (n = 10) | 0.039 | ||
| Home ventilator (% yes) | 4.1% | 3.7% | 30.8% | 0.001 | ||
| Severity of BPD | Mild | 13.6% | 13.8% | 0.0% | 0.36 | |
| Moderate | 34.8% | 34.9% | 30.8% | |||
| Severe | 47.0% | 46.8% | 61.5% | |||
| Unknown | 4.6% | 4.6% | 7.8% | |||
| Pulmonary hypertension (% yes) | 16.9% | 16.6% | 38.5% | 0.053 | ||
| Age at NICU discharge (months) | 4.4 ± 2.7 [0.1, 26.5] (n =861) | 4.4 ± 2.6 [0.1, 26.5] (n = 848) | 7.9 ± 6.1 [3.1, 24.4] | 0.007 | ||
| Surgical Procedures | Tracheostomy (% yes) | 5.2% | 4.7% | 38.5% | <0.001 | |
| Gastrostomy tube (% yes) | 30.2% | 29.3% | 84.6% | <0.001 | ||
| Nissen fundoplication (% yes) | 18.2% | 17.9% | 38.5% | 0.07 | ||
| CSF shunt (% yes) | 7.9% | 7.5% | 30.8% | 0.015 | ||
| PDA Procedures (% yes) | 19.9% (n = 860) | 20.0% (n = 847) | 15.4% | 1.00 | ||
| Meds | Diuretics (% yes) | 62.8% (n = 860) | 62.8% (n = 847) | 61.5% | 1.00 | |
| Inhaled steroids (% yes) | 81.2% | 81.2% | 84.6% | 1.00 | ||
| Pulmonary anti-hypertensives (% yes) | 5.3% | 5.3% | 7.7% | 0.51 | ||
| Growth Measures | Birth weight (grams) | 923 ± 407 [380, 4200] (n = 850) | 922 ± 404 [380, 4200] (n = 837) | 989 ± 552 [510, 2290] | 0.75 | |
| Birth weight percentile (%) | 42 ± 25 [1, 99] (n = 850) | 42 ± 25 [1, 99] (n = 837) | 33 ± 26 [2, 79] | 0.18 | ||
Post-discharge mortality was associated with an increased need for supplemental oxygen at the time of NICU discharge (76.9% in the deceased population versus 42.2% in the living population; p=0.02) as well as an increased amount of supplemental oxygen required (0.87 ± 0.75 LPM in the deceased population versus 0.36 ± 0.35 LPM in the living population; p=0.039). The deceased population was more likely to have had a tracheostomy (38.5% compared with 4.7% in the living population; p<0.001), a gastrostomy tube (84.6% versus 29.3% in the living population; p<0.001), and to require home ventilation (30.8% versus 3.7% in the living population; p=0.001). Cerebrospinal fluid (CSF) shunts were more common in the deceased population (30.8% vs. 7.5% in the living population; p=0.015). The deceased population was older at the time of NICU discharge (7.9 ± 6.1 months) compared with the living population (4.4 ± 2.6 months; p=0.007). Although BPD severity was not found to be significantly associated with mortality, none of the individuals in the deceased population had mild disease.
Details surrounding the deaths of the deceased population are detailed in Table 3 and include the age of the subject at time of death, location of death, cause of death (if known), home respiratory support at time of death, selected diagnoses at time of death, and medications at time of death. Kaplan-Meier time-to-event analysis shows a median age at time of death of 2.47 years (Figure; available at www.jpeds.com). Five subjects died while readmitted to the hospital after the initial NICU admission, including three in the pediatric intensive care unit and two on the inpatient floor. Seven subjects died in the home setting including three in the emergency department after suffering a cardiopulmonary arrest at home. Cardiopulmonary arrest was identified as the cause of death in seven or 54% of subjects. A neurologic cause of death, such as ventriculoperitoneal shunt malfunction or head trauma, was identified in three subjects. Other causes of death included progression of underlying disease and septic shock. At the time of death, four subjects were on a ventilator, one subject received supplemental oxygen via tracheostomy collar, five patients required supplemental oxygen via nasal cannula, and three subjects required no additional respiratory support.
Table 3.
Deceased subjects
| Patient | Age Group at death (years) | Location | Cause | Home respiratory support at time of death | Selected diagnoses at time of death | Medications |
|---|---|---|---|---|---|---|
| 1 | 5-14 | PICU | Septic shock prior to admission (Withdrawal of care) | Trach collar (21%) | Cerebral palsy, HIE, seizures, CLD, OSA, and laryngotracheomalacia | Levetiracetam, baclofen, fluticasone, albuterol, ipratropium, glycopyrrolate, lansoprazole, erythromycin |
| 2 | <1 | Car | Arrest | NC oxygen | CLD | Chlorothiazide, spironolactone, ranitidine |
| 3 | 1-4 | Home with hospice | Shunt malfunction | NC oxygen | Hydrocephalus with shunt, CLD | Fluticasone, albuterol, glycopyrrolate, lansoprazole, erythromycin, levetiracetam |
| 4 | 5-14 | Inpatient floor | Head trauma (Withdrawal of care) | Room air | Chromosomal abnormality, ASD, VSD, pHTN with stenting of pulmonary vein, CLD | Fluticasone, montelukast, sildenafil, bosentan, clopidogrel, acetylsalicylic acid |
| 5 | 1-4 | Unknown | Unknown | Room air | pHTN, hydrocephalus with shunt, HSV encephalitis, seizure disorder, cortical blindness, CLD | Fluticasone, albuterol, erythromycin, levetiracetam, phenobarbital, ranitidine, valacyclovir |
| 6 | 1-4 | Home | Arrest | Ventilator | CLD | Fluticasone, albuterol, ipratropium, montelukast, chlorothiazide, spironolactone, lansoprazole |
| 7 | <1 | ED | Cardiac arrest (home) | NC oxygen | CLD | Ranitidine |
| 8 | 1-4 | Home with hospice | Multi-organ progressive disease (Withdrawal of care) | NC oxygen | HIE, hydrocephalus with shunt, TPN dependence, chronic pancreatitis, seizures, CLD, OSA | |
| 9 | 1-4 | ED | Respiratory failure/cardiac arrest (home) | Ventilator | CLD, Pierre-Robin sequence, severe trismus | Beclomethasone, levalbuterol, ranitidine |
| 10 | 1-4 | PICU | Hydrocephalus progression (Withdrawal of care) | NC oxygen | Hydrocephalus with shunt, seizure disorder, CLD | Beclomethasone, albuterol, chlorothiazide, levetiracetam, topiramate, ranitidine |
| 11 | 1-4 | Inpatient floor | Respiratory failure; multi-organ progressive disease (Withdrawal of care) | Ventilator | Diffuse cystic PVL & hydrocephalus s/p ventriculostomy, seizures, CLD, PDA, right atrial thrombus | Baclofen, diazepam, gabapentin, levetiracetam, phenobarbital, topiramate, fluticasone, levalbuterol |
| 12 | 1-4 | ED | Arrest (home) | Room air | CLD, G6PD with hyperbilirubinemia, seizure disorder | Fluticasone, albuterol, levetiracetam, phenobarbital, erythromycin, cyproheptadine, omeprazole |
| 13 | 1-4 | PICU | Trach dislodgement (Home) (Withdrawal of care) | Ventilator | Hydrocephalus, CLD, with trach/vent, pHTN, adrenal insufficiency | Budesonide, levalbuterol, sildenafil, spironolactone, chlorothiazide, azithromycin |
Figure 1.
Kaplan-Meier Survival Graph
Of the clinical characteristics included in univariate regression analysis (Table 4), when adjusting for gestational age and BPD severity, a higher supplemental oxygen requirement at the time of NICU discharge (adjusted hazard ratio (aHR): 6.25; p<0.001), home ventilator use (aHR: 9.17; p=0.002), and age at NICU discharge (aHR: 1.18; p<0.001) were associated with mortality. Tracheostomy placement (aHR: 9.61; p=0.001), gastrostomy placement (aHR: 12.22; p=0.002), and presence of a CSF shunt (aHR: 5.75; p=0.006) were also associated with mortality. There was no significant association between requirement for supplemental oxygen at NICU discharge (p=0.06), diagnosis of pulmonary hypertension (p=0.08), or presence of a Nissen fundoplication (p=0.22) and mortality.
Table 4.
Predictors of Mortality
| Adjusted Hazard Ratio ± S.E. [95% C.I.] | Univariate Cox Regression (n = 862) | P value | Final Multivariate Cox Regression (n = 862) | P value | |
|---|---|---|---|---|---|
| Clinical Characteristics | Supplemental oxygen at NICU discharge (yes=1; no=0) | 3.47 ± 2.30 [0.94, 12.74] | 0.06 | - | - |
| Oxygen amount at NICU discharge if on oxygen (LPM) | 6.25 ± 2.72 [2.66, 14.68] | <0.001 | 4.10 ± 1.78 [1.75, 9.62] | 0.001 | |
| Home ventilator (yes=1; no=0) | 9.17 ± 6.65 [2.21, 38.02] | 0.002 | - | - | |
| Pulmonary hypertension (yes=1; no=0) | 2.94 ± 1.83 [0.87, 9.93] | 0.08 | - | - | |
| Age at NICU discharge (months) | 1.18 ± 0.05 [1.08, 1.29] (n = 861) | <0.001 | - | - | |
| Surgical Procedures | Tracheostomy (yes=1; no=0) | 9.61 ± 6.35 [2.63, 35.10] | 0.001 | - | - |
| Gastrostomy tube (yes=1; no=0) | 12.22 ± 9.79 [2.55, 58.71] | 0.002 | 8.13 ± 6.77 [1.59, 41.54] | 0.012 | |
| Nissen fundoplication (yes=1; no=0) | 2.09 ± 1.26 [0.64, 6.81] | 0.22 | - | - | |
| CSF shunt (yes=1; no=0) | 5.75 ± 3.65 [1.65, 19.97] | 0.006 | 4.31 ± 2.73 [1.25, 14.93] | 0.021 | |
N.B. Factors associated or trending towards association in Table 1 were tested as predictors of outpatient mortality in Cox regressions. Initial oxygen amount was coded as 0 if the subject was not on supplemental oxygen. All regressions were adjusted for gestation (weeks) and BPD severity (dummy categorical variable).
Multivariate analysis was performed using the variables found to have association with mortality in univariate analysis. After adjusting for gestational age and BPD severity, a positive association was seen between mortality and the following characteristics: the amount of supplemental oxygen required at NICU discharge (aHR: 4.10; p=0.001), presence of a gastrostomy tube (aHR: 8.13; p=0.012), and presence of a CSF shunt (aHR: 4.31; p=0.021).
DISCUSSION
This study explores the rates and potential predictors of outpatient mortality in a group of premature patients with BPD. Following initial hospital discharge, we found that children in this population have a higher rate of mortality than the general population in all age groups including in the post-neonatal period (incidence rate ratio of 1.96), between the ages of one and four years (incidence rate ratio of 12.92), and between the ages of five and fourteen years (incidence rate ratio of 5.37). In addition, compared with the U.S. population of preterm infants and children, the mortality rate observed in our population of infants with BPD is increased. Specifically, in 2018, the rate of infant mortality was 115 deaths per 100,000 preterm infants (<34 weeks gestation) within the first year of life, including deaths occurring during the initial admission, and we observed mortality rates of 370 and 310 within our population from 28-364 days and 1-4 years, respectively, not including mortality occurring with the initial hospital admission.2 Although other studies have illustrated that infants with BPD are at increased risk for long-term morbidity including readmission to the hospital and multi-system complications such as pulmonary hypertension and feeding difficulties, our study suggests that they may also be at an increased risk of mortality.6,7,8
Several previous studies as well as current U.S. data have illustrated racial and ethnic disparities in mortality among full and preterm infants.2, 15, 16 However, similar to a large population-based study by Kempe et al,17 our study did not find any demographic characteristics that increased a patient’s risk of mortality, which may be due to the overall small number of deceased patients in our population. This suggests that it may be difficult to identify infants at highest risk based on demographic characteristics alone.
We did identify several clinical features that were associated with an increased risk of mortality including the amount of supplemental oxygen required at the time of NICU discharge, the need for home ventilation, and the presence of a tracheostomy, which suggests that those with more severe lung injury are at increased risk of mortality after discharge. Edwards et al18 demonstrated an 80% five-year survival rate for a cohort of children requiring chronic home ventilation. Although our results demonstrated a lower risk of mortality among subjects requiring home ventilation (11%), these patients are at increased risk of outpatient mortality compared with children who do not require home ventilation. Therefore, it is important to continue to explore interventions such as increased caregiver education and hospitalization in a subacute facility prior to discharge home to help mitigate this increased risk. Given that the entire study population had some degree of chronic lung disease of prematurity at baseline, we noted that it was not just respiratory symptoms or respiratory support in the home setting that placed these infants at higher risk of mortality. In fact, in multivariate analysis, when controlling for severity of BPD, the need for a gastrostomy tube and CSF shunt as well as the amount of oxygen required at NICU discharge were still significantly associated with mortality. Interestingly, mortality was not associated with BPD severity, which may suggest that the NHLBI grading system (based on status at 36 weeks corrected age) does not necessarily identify those at greatest risk for mortality. Although other studies have similarly demonstrated an increased risk of mortality in infants with a diagnosis of BPD as compared with those with no lung disease,8 we were unable to make any conclusions about the relationship between the severity of disease and risk of mortality. Additionally, we did not observe any association between mortality and low birth weight as has been seen in several prior studies;10–11, 19 this may be related to our population having a mean birth weight of less than 1,000 grams and a higher risk of comorbidities, as the study population was limited to those infants with BPD. Older age at the time of discharge was associated with an increased risk of mortality which suggests that infants with more prenatal and perinatal complications, requiring more interventions and subsequent longer hospitalization, are at continued risk of mortality following discharge and require close monitoring.
After adjusting for gestational age and severity of BPD, the presence of a CSF shunt was positively associated with an increased risk of mortality. This association between neurologic sequelae of prematurity and increased risk of mortality has been demonstrated in other studies.7,8 Piecuch et al, identified intracranial hemorrhage (ICH) as the most worrisome mortality risk factor associated with sudden unexpected death in a medically stable cohort of low birth weight infants. Previous studies have suggested that seizure activity can disrupt normal physiologic regulation and control of respiratory and cardiac activity, which could lead to sudden death or apparent life-threatening events.20, 21 However, in Piecuch’s study, the increase in mortality was not associated with advanced grades of ICH as only one infant had a CSF shunt, none had seizures or were on anti-epileptics, and all had a normal neurologic examination prior to death. It was conjectured that the association of neurologic sequelae of prematurity and increased risk of mortality may be due to a compromised premature brain with disorganized control of breathing that is more likely to develop sudden respiratory compromise resulting in acute cardiopulmonary arrest.
Feeding difficulties leading to the placement of a gastrostomy tube is a common comorbidity of prematurity, and the presence of a gastrostomy tube was found to be positively associated with an increased risk of mortality in the study population. This may suggest that aspiration, reflux, or swallowing dysfunction are more common in infants with more severe lung disease or that feeding difficulties contribute to worsening lung disease and an overall increased risk of mortality.
This study was limited by the use of an already recruited study population and clinical information obtained primarily via chart review, some of which may have been missing, especially with regards to deaths that occurred outside of the hospital setting. Patients were recruited from a single outpatient center after discharge from multiple NICUs so the care patients received following discharge was similar based on the standard of practice at that center. We do not have information on parental training prior to discharge or time spent in subacute facilities that may influence outcomes.
In conclusion, although survival in preterm infants has improved significantly over the past several decades, we found that the incidence and risk of death after NICU discharge in infants and children with BPD remains substantially higher than that seen in the general population. Larger multicenter studies may be useful in confirming our findings. Future studies should focus on further clarifying risk factors for the development of severe disease and ultimately increased risk of morbidity and mortality in order to allow for early identification, routine follow-up, and regular treatment of those at highest risk.
Supplementary Material
ACKNOWLEDGEMENTS
The authors wish to thank the families who participated in this study.
Supported by the National Institutes of Health (Bethesda, MD, USA)(K23 ES029985 [to J.R.]; R01 HL114800 [to S.M-M.]), the Johns Hopkins Eudowood Foundation (Fellowship grant to B.A.), Children’s Hospital of Philadelphia (to J.C.), and the Thomas Wilson Foundation (to J.C.).The funding sources had no involvement in the writing of the manuscript or the decision to submit. The authors declare no conflicts of interest.
ABBREVIATIONS
- BPD
bronchopulmonary dysplasia
- AMI
annual mortality index
- aHR
adjusted hazard ratio
- CSF
cerebrospinal fluid
- NHLBI
National Heart, Lung, and Blood Institute
- NICU
neonatal intensive care unit
- ICH
intracranial hemorrhage
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Ely DM, Driscoll AK, Matthews TJ. Infant Mortality by Age at Death in the United States, 2016. NCHS Data Brief 2018:1–8. [PubMed] [Google Scholar]
- 2.Ely DM, Driscoll AK. Infant Mortality in the United States, 2018: Data From the Period Linked Birth/Infant Death File. Natl Vital Stat Rep 2020;69:1–18. [PubMed] [Google Scholar]
- 3.Liu L, Oza S, Hogan D, Perin J, Rudan I, Lawn JE, et al. Global, regional, and national causes of child mortality in 2000–13, with projections to inform post-2015 priorities: an updated systematic analysis. Lancet 2015;385:430–40. [DOI] [PubMed] [Google Scholar]
- 4.Sells CJ, Neff TE, Bennett FC, Robinson NM. Mortality in infants discharged from a neonatal intensive care unit. Am J Dis Child 1983;137:44–7. [DOI] [PubMed] [Google Scholar]
- 5.Kulkarni P, Hall RT, Rhodes PG, Sheehan MB. Postneonatal infant mortality in infants to a neonatal intensive care unit. Pediatrics 1978;62:178–83. [PubMed] [Google Scholar]
- 6.Hulsey TC, Hudson MB, Pittard WB 3rd. Predictors of hospital postdischarge infant mortality: implications for high-risk infant follow-up efforts. Journal of perinatology : official journal of the California Perinatal Association 1994;14:219–25. [PubMed] [Google Scholar]
- 7.Piecuch RE, Leonard CH, Clyman RI, Cooper BA. Risk factors associated with infant death among very low birth weight infants after discharge from an intensive care nursery. J Dev Behav Pediatr 1998;19:84–8. [DOI] [PubMed] [Google Scholar]
- 8.Kugelman A, Reichman B, Chistyakov I, Boyko V, Levitski O, Lerner-Geva L, et al. Postdischarge infant mortality among very low birth weight infants: a population-based study. Pediatrics 2007;120:e788–94. [DOI] [PubMed] [Google Scholar]
- 9.Baraldi E, Filippone M. Chronic lung disease after premature birth. The New England journal of medicine 2007;357:1946–55. [DOI] [PubMed] [Google Scholar]
- 10.Bhandari A, Bhandari V. Pitfalls, problems, and progress in bronchopulmonary dysplasia. Pediatrics 2009;123:1562–73. [DOI] [PubMed] [Google Scholar]
- 11.Abman SH, Collaco JM, Shepherd EG, Keszler M, Cuevas-Guaman M, Welty SE, et al. Interdisciplinary Care of Children with Severe Bronchopulmonary Dysplasia. The Journal of pediatrics 2017;181:12–28 e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Anderson PJ, Doyle LW. Neurodevelopmental outcome of bronchopulmonary dysplasia. Seminars in perinatology 2006;30:227–32. [DOI] [PubMed] [Google Scholar]
- 13.Jobe AH, Bancalari E. Bronchopulmonary dysplasia. American journal of respiratory and critical care medicine 2001;163:1723–9. [DOI] [PubMed] [Google Scholar]
- 14.Murphy SL, Xu J, Kochanek KD, Arias E, Tejada-Vera B. Deaths: Final Data for 2018. Natl Vital Stat Rep 2021;69:1–83.15. [PubMed] [Google Scholar]
- 15.Hessol NA, Fuentes-Afflick E. Ethnic differences in neonatal and postneonatal mortality. Pediatrics 2005;115:e44–51. [DOI] [PubMed] [Google Scholar]
- 16.Chen HY, Chauhan SP, Rankins NC, Ananth CV, Siddiqui DS, Vintzileos AM. Racial and ethnic disparities in infant mortality in the United States: the role of gestational age. Am J Perinatol 2013;30:469–75 [DOI] [PubMed] [Google Scholar]
- 17.Kempe A, Wise PH, Wampler NS, Cole FS, Wallace H, Dickinson C, et al. Risk status at discharge and cause of death for postneonatal infant deaths: a total population study. Pediatrics 1997;99:338–44. [DOI] [PubMed] [Google Scholar]
- 18.Edwards JD, Kun SS, Keens TG. Outcomes and causes of death in children on home mechanical ventilation via tracheostomy: an institutional and literature review. J Pediatr. 2010. Dec; 157:955–959.e2. [DOI] [PubMed] [Google Scholar]
- 19.Smith VC, Zupancic JA, McCormick MC, Croen LA, Greene J, Escobar GJ, et al. Trends in severe bronchopulmonary dysplasia rates between 1994 and 2002. The Journal of pediatrics 2005;146:469–73. [DOI] [PubMed] [Google Scholar]
- 20.O’Regan ME, Brown JK. Abnormalities in cardiac and respiratory function observed during seizures in childhood. Dev Med Child Neurol 2005;47:4–9. [DOI] [PubMed] [Google Scholar]
- 21.Hewertson J, Poets CF, Samuels MP, Boyd SG, Neville BG, Southall DP. Epileptic seizure-induced hypoxemia in infants with apparent life-threatening events. Pediatrics 1994;94:148–56. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.