Health, Development, and Quality of Life in Children with Severe Bronchopulmonary Dysplasia

Abstract

Objectives:

The primary objective of this study was to profile the childhood health, development, and health-related quality of life (HR QoL) for children with severe bronchopulmonary dysplasia (BPD) discharged from a quaternary neonatal intensive care unit-based program. The secondary objectives were (1) to evaluate which outcomes correlated with parent-reported overall health and HR QoL, and (2) to evaluate how outcomes varied across age.

Study Design:

We collected cross-sectional data through telephone interviews with 282 families of children ages 18 months to 11 years. We used regression analysis to evaluate correlations between child outcomes and parent-reported health, quality of life, and child’s age.

Results:

Respiratory morbidities were near universal, with 42% of children ever having required a tracheostomy and 77% having experienced wheezing. Respiratory morbidities were strongly correlated with parent-reported health and QoL. Most respiratory morbidities decreased with increasing child age. Developmental morbidities were also marked: 20% carried a diagnosis of autism and 97% required an individualized educational plan. Developmental morbidities had a more prominent impact on quality of life in older children than in younger children.

Conclusions:

Respiratory and developmental morbidities in this referral cohort are on numerous counts more severe than any reported in the literature. While respiratory morbidities and overall health improve over time, developmental morbidities become increasingly prominent, resulting in lower quality of life.

Introduction

Bronchopulmonary dysplasia (BPD), the most common chronic morbidity of prematurity, is associated with a wide range of sequelae that affect children in the years to come.[1,2] BPD affects respiratory health, development, and quality of life.[3–5] The severity of disease in infancy correlates with the extent of long-term sequelae.[6] As treatments improve, children with increasingly severe BPD are surviving.[7] In more mild disease, children generally wean off respiratory support by the time of hospital discharge, and both respiratory and non-respiratory effects of BPD generally improve over time.[3,5] This population of children with BPD may experience more severe reactions to childhood illness or reveal neurodevelopmental disabilities as they age, but in many respects morbidities improve through childhood. In contrast, children with severe BPD are more likely to require a prolonged index hospitalization and frequently have chronic medical needs. For these children, BPD is more likely to impact their daily lives for years after neonatal intensive care unit (NICU) discharge. The extent of this impact and time course of potential improvements remain unclear.

There have been variable and generally consensus-based approaches in defining BPD and associated severity levels.[8] Recently an evidence-based definition utilized the levels of respiratory support at the diagnosis of BPD that were associated with long-term sequelae, for example discharge on respiratory support or rehospitalizations, to develop a severity based grading system. However, all grades are defined based on what clinical features in the neonatal period predict particular long-term sequelae, for example discharge on respiratory support or rehospitalizations.[9] As a result, stratifying sequelae by grade of BPD is circular and leads to detection of the outcomes upon which the definitions were originally based. There are limited data about whether these are the outcomes important to families.[10,11] Further, BPD, particularly in its most severe form, affects many organ systems, which has a multi-dimensional impact on the lives of children and families. Thus, measuring a narrow subset of respiratory outcomes may not capture the reality of living with BPD.[11] Some work on severe BPD focuses only on children who have tracheostomies, which represents only a small fraction of patients and eliminates many children who have other manifestations that may be considered equally severe by both families and clinicians. Finally, outcome studies frequently lack a longitudinal dimension.[12] For many parents facing difficult clinical realities, “will this get better or worse?” and “how long will this last” are also important questions.[13]

In this study, we sought to overcome some of these knowledge gaps by evaluating respiratory and non-respiratory childhood outcomes of patients discharged from a quaternary-level program for the inpatient care of infants with BPD. We report cross-sectional data on numerous childhood outcomes for this population: measures of respiratory technology and function, development and learning measures, parent-reported overall health, and a health-related quality of life (HR QoL), or the impact of a child’s health on their ability to live a fulfilling life.[14] We compare outcomes in different age children, ranging from age 18 months to 11 years. A secondary objective was to evaluate correlates between outcomes and parent-reported health or HR QoL.

Methods

We contacted the families of children who received inpatient care from the Newborn and Infant Chronic Lung Disease (NeoCLD) Program at Children’s Hospital of Philadelphia from its inception in May 2010 through November 2020. This multidisciplinary program focuses on the care of infants with the most severe forms of BPD and receives 50 – 70 patients annually, referred from across the United States.

The questionnaires were selected and developed by a multidisciplinary group, including physicians, nurses, social workers, and physical therapists, all with a focus on clinical care for and research to improve outcomes of patients with BPD. As a group, we selected instruments that are valid across a wide age range, in children with a wide range of abilities, and appropriate for telephone administration. Because HR QoL does not always correlate neatly with health outcomes, nor parents’ perceptions of overall health, we included a question directly eliciting parent’s perception of their child’s health.[15] The final list of questionnaires included:

1. A questionnaire about family demographic characteristics

2. A questionnaire about the child’s health since NICU discharge, including the question “In general, how would you describe your child’s health” on a 5-point Likert scale from poor to excellent

3. The International Study of Asthma and Allergies in Childhood (ISAAC) standardized questions to assess respiratory symptoms[16,17]

4. The Adaptive Behavior Assessment System (ABAS), 3^rd Edition, a questionnaire about the child’s development with a focus on activities of daily living across the lifespan[18]

5. The PedsQL Inventory, which is a standardized metric of HR QoL. The PedsQL includes a total score and domain scores relating to physical, emotional, social, and school functioning, each on a 100-point scale. There are versions adapted to age <2, 2-4, 5-7, and 8-12 years.[19]

The NeoCLD Program has maintained a patient registry since its establishment. The study team reviewed the medical records of each child in the registry. Patients who died during the hospitalization or after discharge were not contacted, but the age of death and cause of death, if available in the electronic medical record, were recorded. Parents who did not speak English were included and assisted in completing questionnaires using a medical interpreter. We called the legal parent(s) or guardian(s) of all eligible children to offer participation in this virtual follow-up study. We made up to three attempts to reach parents. When families did not answer the phone or the numbers in the medical record were incorrect, an invitation letter was mailed to the most recent address on record. Parents who consented were assisted in completing questionnaires over the phone. Most data were gathered from parent report, including diagnoses of developmental delay, autism, cerebral palsy, poor weight gain which were asked as, “Has a doctor ever expressed concern that your child has [condition]?” Parents were asked whether their child received outpatient therapies including physical therapy, occupational therapy, speech therapy, behavioral therapy, and autism-related services. The study team also reviewed the medical record to fill in missing data wherever possible, but parents’ reports were honored above the medical record in disagreements. We recorded BPD grade based on respiratory support at 36 weeks in accordance with the 2019 NRN definition.[9] All data were maintained in REDCap.[20,21]

We analyzed data using Stata 18.0 (College Station, TX). For all variables, we calculated the percentage of children affected, mean values, and standard deviations (SD). Chronological age was analyzed as a continuous variable, except in analyzing the PedsQL HR QoL, which has slightly modified questions for the 2-4, 5-7, and 8-12 year age groups. We prospectively designated the following variables as primary outcomes for which additional analyses were conducted: current respiratory support (yes or no), tracheostomy (ever placed), current medicines (number), current respiratory medicines (number), wheezing in the past 12 months, rehospitalizations (lifetime number), hearing or vision impairment, poor weight gain, walking alone, cerebral palsy (yes or no), developmental delay, autism, outpatient therapies (number of services), parent-reported health, HR QoL (total and all subscores). We assessed changes in the primary outcome variables by child age using linear regression analysis and created graphical plots of variables that changed significantly with age. We restricted this analysis to variables that may change over a child’s life.

Next, we re-scaled the Likert rating of overall health to a 100-point scale to parallel the HR QoL metric and assessed the correlations between the primary outcome measures and the outcomes of HR QoL or parent-reported health. For these correlations we used Wilcoxon rank sum or Spearman’s rank correlation coefficient, as appropriate, to assess for significant relationships and then used regression analysis to assess the magnitude of the correlations. We also compared patient characteristics between respondents and non-respondents using t-tests. We did not correct for multiple comparisons as these analyses are exploratory in nature. This study received institutional review board approval from Children’s Hospital of Philadelphia, and all parents included provided informed consent.

Results

Between June of 2021 and July of 2022, 568 children received care from the NeoCLD Program and survived to discharge. Of these, 89 children (16%) died after NICU discharge so were excluded from the current study. The mean age of death was 1.0 year (SD 1.5). The cause of death could be determined from the medical record for 48 children and was the result of respiratory disease in 70% of these (n=34). The study team called 479 eligible families and a majority participated in this study (n=282, 59% overall, and 67% of those with whom we successfully established contact, Figure 1). Respondents did not differ from non-respondents with regard to child’s gestational age, birth weight, age at NICU discharge, or age at follow-up call (Table 1).

Figure 1. Enrollment Flowchart.

This figure depicts the total number of eligible families and details the outcome of each contact.

Table 1. Characteristics of respondents versus non-respondents.

This table compares characteristics of respondents versus non-respondents, define as families who were eligible and reached but did not complete the study. P-values reflect comparisons using t-tests. PMA= Post menstrual age; ISAAC= International Study of Asthma and Allergies in Childhood

Characteristic	Respondents Mean (SD) N=282	Non-Respondents Mean (SD) N=138*	p-value
Birth weight (g)	839.2 (458.6)	830.7 (477.1)	0.42
Gestational age at birth (weeks)	26.0 (2.7)	26.1 (2.9)	0.42
Postmenstrual age at discharge (weeks)	61.6 (14.3)	60.3 (12.9)	0.17
Age at contact (years)	5.7 (2.9)	6.0 (3.0)	0.17

The final study cohort had early risk factors for and indicators of BPD, with low gestational age and birth weight. The majority had Grade 2 or 3 BPD. Discharge was, on average, at 61.6 (SD 14.3) weeks post menstrual age, and a majority of children were discharged on respiratory support (Table 2). Length of NICU stay did not change with year of discharge (R² = −0.12, p =0.626). Virtual follow-up occurred when the children were a mean age of 5.7 years (SD 2.9) chronologic age. At the time of follow-up, 95% of children lived at home, with a small minority in long-term care/rehabilitation facilities or hospitalized. The cohort had almost universal long-term sequelae related to respiratory disease. Forty two percent had ever had tracheostomies and, of these, 58% had been decannulated, at a mean age of 3.7 years (SD1.4). The percentage of children not on respiratory support increased across the age range (R² = 0.03, p <0.001, Figure 2), and of children over age 3 at follow up 85% were not on respiratory support.

Table 2. Characteristics of the study sample and outcomes at follow up.

This table includes both characteristics of the study sample and outcome data from follow up.

Characteristic	Percentage or Mean (SD) n=282 (unless noted)
Demographic Characteristics
Female sex	37%
Race
Asian	2%
Black or African American	33%
Multi-Racial	6%
White	36%
Other	19%
Refused	2%
Ethnicity
Hispanic or Latino	12%
Not Hispanic or Latino	85%
Birth/NICU Characteristics
Gestational age (weeks)	26.0 (2.6)
Birthweight (grams)	839.2 (458.9)
Respiratory Support at 36 weeks PMA
No respiratory support (no BPD)	2%
≤2 L nasal cannula (Grade 1 BPD)	6%
Noninvasive support > 2 L (Grade 2 BPD)	38%
Intubated (Grade 3 BPD)	54%
Intraventricular hemorrhage
Grade I/II	23%
Grade III/IV	13%
Periventricular leukomalacia	10%
PMA at NICU discharge (weeks)	61.6 (14.3)
Discharge to medical institution	15%
Respiratory support at discharge
No respiratory support	18%
≤2 L nasal cannula	31%
Noninvasive support > 2 L nasal cannula	10%
Tracheostomy	42%
Follow-Up
Age at follow-up (years)	5.7 (2.9)
Where child lives
Home	95%
Long term care/rehabilitation facility	3%
Long-term hospitalization	2%
Respiratory Health at Follow-Up
Tracheostomy
Ever placed	42%
If tracheostomy: age at follow up	5.6 (2.8)
If had tracheostomy: Decannulated	58%
Mean age at decannulation (years)	3.7 (1.4)
Respiratory support at follow-up
Nasal cannula	4%
CPAP or BiPAP	2%
Tracheostomy with T piece/no support	9%
Tracheostomy with CPAP/support	2%
Tracheostomy with mechanical ventilation	4%
Bronchoscopy, in the past year	13%
Laryngoscopy, in the past year	8.6%
Number of daily medications
All medications	2.2 (1.7)
Respiratory medications only	1.3 (1.1)
Respiratory Symptoms (ISAAC)
Wheezing, ever	77%
Wheezing, past year	51%
Asthma diagnosis	45%
Sleep disturbed by wheezing, past year	52%
Wheezing with exercise, past year	15%
Wheezing limits speech (if speaks), past year	21%
Dry cough at night, past year	18%
Rehospitalization, ever	76%
If yes: Number of lifetime rehospitalizations	4.1 (2.3)
Percentage of rehospitalizations related to respiratory disease	33%
Emergency room for respiratory reason, past year	20%
If yes: Number of visits	1.8 (1.3)
Doctor’s office for respiratory reason, past year	27%
If yes: Number of visits	2.2 (1.2)
Home nursing, ever	37%
Other Health at Follow-Up
Poor weight gain, doctor ever said	48%
Any use of feeding tube, past year	55%
Parent rating of health (Likert scale, 0-100)	74.5 (23.4)
Development and Learning at Follow-Up
Hearing loss	4%
Vision loss	37%
Walking alone, if age >1 year (n=275)	81%
Cerebral palsy	11%
Developmental delay	45%
Attention or behavior problems	27%
Autism	20%
Outpatient therapy, all types, ever	87%
If yes: mean number	3.4 (1.3)
Physical therapy	68%
Occupational therapy	71%
Speech therapy	73%
Feeding therapy	35%
Behavioral therapy or counseling	14%
Enrolled in school, if age >3	83%
If in school: has an IEP	95%
Quality of Life (PedsQL Inventory) at Follow-Up
Total Score	84.4 (11.5)
Domain: Physical Functioning	86.9 (13.6)
Domain: Emotional Functioning	85.4 (15.0)
Domain: Social Functioning	83.0 (16.0)
Domain: School Functioning	79.2 (17.0)

Figure 2. Changes in Outcomes over Age.

This figure displays changes in outcomes over age. It includes only outcomes that changed significantly over age when assessed through linear regression analysis.

Children were prescribed an average of 2.2 medications (SD 1.7), of which 1.2 (SD 1.1) were related to respiratory disease. While 77% of children had ever wheezed, the percentage who had in the past year decreased across age (R² = −0.039, p=0.001, Figure 2). Forty-five percent of children had been diagnosed with asthma. A majority, 76%, had been re-hospitalized.

These children also experienced many non-respiratory sequelae, including both growth and developmental problems. Nearly half had experienced poor weight gain, and 55% currently used a feeding tube. Parents reported they had been told by a doctor that their child had cerebral palsy for 11% of children, developmental delay for 45%, attention or behavior problems for 27%, and autism for 20%. To address these problems, most children were receiving several outpatient therapies, though the number decreased across age (R² = 0.071, p =0.032, Figure 2). At the time of follow-up, 81% of children could walk alone. Of children over 3 years of age (n=217), 83% were enrolled in school and, of those who were in school, 95% had an individualized education plan.

Parent report of the child’s general health was on average 74.5 (SD 23.4), corresponding to “very good,” and this rating increased as child age increased (R² = 1.20, p = 0.013, Figure 2). The mean total score on the PedsQL HR QoL was 84.4 (SD 11.5), with higher scores indicating higher HR QoL. Physical and emotional subscores were higher than social and school subscores. Total scores decreased as child age increased, which was the result of decreases in school and emotional subscores. In the 2 – 4 year group, the lowest scoring questions were: not able to do things other children his/her age can do, keeping up playing with other children, and missing school for doctor/hospital. In the 5 – 7 age group, the lowest scoring questions were: not able to do things other children his/her age can do, paying attention, and keeping up playing. In the 8 – 12 age group, the lowest scoring questions were: paying attention in class, forgetting things, and a tie between keeping up playing and keeping up with schoolwork. The questions with the highest ratings (least concern) across the entire age range were: feeling sad or blue, getting teased by other children, and having low energy level.

In regression analysis, the following outcomes were associated with both parent-reported health and HR QoL total score: numbers of therapies, medications, respiratory medications, and wheezing. Respiratory support, inability to walk, and having cerebral palsy were associated with parent-reported health only (Figure 3). Autism, developmental delay, poor weight gain, and rehospitalizations were associated with HR QoL only. Having ever had a tracheostomy and vision or hearing impairment were not associated with parent-reported health or HR QoL. The correlation coefficients for these associations are provided in Table 3.

Figure 3. Correlations between child outcomes and parent-reported health or HR QoL.

The connecting lines depict significant relationships between outcomes and parent-reported health or HR QoL total score as assessed through regression analysis. Thickness of each line is proportional to the magnitude of the correlation coefficient.

Table 3. Correlations between outcomes vs parent-reported health and HR QoL total.

This table provides the R2 and p-value for correlations between outcomes and parent-reported health or HR QoL total score

Outcome	Parent-Reported Health		HR QoL Total
	R2	p-value	R2	p-value
Therapies	−3.54	<0.001	−2.07	<0.001
All medications	−4.83	<0.001	−1.96	<0.001
Respiratory Medications	−5.72	<0.001	−3.25	<0.001
Wheezing	−8.83	0.005	−4.10	0.014
Cerebral palsy	−10.28	0.021	−2.40	0.27
Not walking	−9.62	0.007	−2.24	0.25
On respiratory support	−15.52	<0.001	−0.61	0.74
Rehospitalizations	−0.15	0.77	−0.80	0.003
Poor weight gain	−2.56	0.36	−4.43	0.002
Developmental delay	−4.69	0.19	−6.01	<0.001
Autism	−3.29	0.351	−6.86	<0.001
Tracheostomy ever	−4.75	0.094	−2.14	0.141
Hearing	−6.35	0.38	1.71	0.65
Vision	−4.55	0.118	−2.56	0.081

Discussion

We present parent-reported childhood outcomes for a cohort of children with severe bronchopulmonary dysplasia and substantial long-term morbidity. On numerous measures, this cohort had more sequelae of BPD than commonly reported in the literature. They experienced prolonged initial hospitalization and nearly half received tracheostomies. Respiratory morbidities persisted after discharge with a large majority experiencing wheezing, taking daily medications for breathing, and requiring rehospitalization, most frequently related to respiratory disease. The respiratory outcomes of these children were worse than comparable data, where available, for cohorts born at similar gestational ages, with BPD of any grade, with grade 3 BPD, and who required a tracheostomy for BPD. For instance, this cohort had 4.1 hospitalizations by a mean of 5.7 years, whereas a cohort of all children with BPD (oxygen use at 36 weeks) was reported to have only 1.2 hospitalizations by 11 years.[22] Seventy-seven percent experienced wheezing compared to 10% in one cohort of children with grade 2 or 3 BPD and 24% in another cohort of children with all grades of BPD.[22,23] Forty-two percent of the children in this cohort required tracheostomy placement, compared to 31% in a larger cohort of children with Grade 3 BPD.[9] If a tracheostomy was placed, these children were decannulated later than any cohort reported in the literature.[12] In sum, these data suggest that this referral cohort represents the most extreme manifestations of BPD, as would be expected of the children cared for by a quaternary-level referral program with a wide catchment area. Even in this severe cohort, a large majority of children who survived to NICU discharge survived at least through childhood. Respiratory morbidities did overall improve over childhood as has been reported for the more general population of children with BPD.[24]

The pervasive developmental sequelae in this cohort were consistent with a well-established correlation between severity of lung disease and neurodevelopmental implications in BPD.[6] Almost half of families had been told by a doctor that their child has developmental delay, and most families had at least one outpatient therapy to address delays. By the time children reached school age, nearly all required an individualized educational plan (IEP). Twenty percent of families in this cohort have been told by a doctor that their child has autism, which is higher than rates reported rates for extreme prematurity or Grade 3 BPD. In one cohort study of children born at 24 – 26 weeks of gestation, 3.8% had a diagnosis of autism.[25] In a group of children with Grade 3 BPD, 13% had autism.[6] The developmental outcomes of this cohort are a striking demonstration of the many challenges severe BPD poses for children and families even after NICU discharge. In contrast, other cohorts of children with Grade 3 BPD report that half of children have normal academic achievement.[6] This discrepancy likely reflects both greater neurodevelopmental sequelae in this referral population and our multidimensional approach to measuring these sequelae. children with normal academic achievement may still need additional support in school necessitating an IEP, for example for wheelchair access or behavioral support.

BPD is a multi-systemic disease with wide-ranging effects, so it is not unexpected that an array of medical and developmental outcomes in children with BPD were negatively correlated with both parent-reported health and HR QoL. Wheezing was the outcome with the overall biggest impact on both parent-reported child health and HR QoL, confirming that chronic respiratory morbidity does have a major impact on children and families. Being on respiratory support and not walking were the most significant predictors of poor parent-reported child health. Both of these outcomes have a clear impact on the moment-to-moment care children require, perhaps causing parents to constantly consider health. Autism and developmental delay had the strongest correlation with HR QoL. The correlation between poor weight gain and HR QoL was also notable since this is not an outcome frequently reported for the BPD population. The lack of correlation between ever having a tracheostomy and parent-reported health or HR QoL is significant as it suggests that restricting analysis of the most severe BPD to children with tracheostomies would be undue. This complex web of outcomes overall affirms the importance of measuring many different outcomes to attempt to capture the multidimensional effects of BPD, particularly among those with the most severe form.

Despite these obstacles, parents generally rated their child’s health to be “very good,” and this rating improved across age groups, paralleling declining respiratory morbidities for many. HR QoL, as reflected in the PedsQL total score, was remarkably similar to other studies that have applied the same metric in populations of children with BPD.[22,26] For instance, McGrath-Morrow et al.[26] reported a mean score of 85 in children age 18-24 months while our mean was 86 for patients under age 5, and Ronkainen et al.[22] reported a mean score of 78 for a cohort of 11-year-olds, while our mean score in the 10-12 range was 79. Prior studies have suggested that severity of BPD is negatively associated with QoL,[6] but finding similar QoL scores in this cohort of children with severe disease suggests that there may be some limit to this effect. While most studies of HR QoL in BPD examine a narrow age range,[5] our study enabled comparison across ages 18 months to 11 years. Dishearteningly, HR QoL decreases with increasing age. This is the result of decreasing emotional and school dimensions, which is consistent with the changes in the lowest-scoring questions. A plausible explanation is that although physical health improves, the significant developmental issues associated with severe BPD come to the forefront as children progress in school and are expected to function increasingly independently. Researchers have raised questions about the importance of HR QoL as a summative measure given that it combines numerous domains which may be of varied importance to parents, including parents of children with BPD.[11,13,15] We agree with these concerns and therefore interpret the values of total HR QoL scores cautiously. Parents may find it more meaningful and concrete to learn that concerns shift as respiratory sequelae improve through childhood. Many children come off respiratory support, but developmental effects, such as trouble keeping up with schoolwork, become more prominent. There are also hopeful findings: parents report that most children don’t get teased or feel sad.

We interpret our findings within the confines of our study design. This is a study of children referred to a specialized program at a single center. Given known variation in care for BPD[27] and regional differences in referral patterns, even other referral centers may have different outcomes. Yet, this is also a strength because these data provide an opportunity to study a common condition in its most severe form. Another limitation is that the study is cross-sectional rather than longitudinal, so differences across age groups could also represent changes in care of infants with BPD over time. This seems unlikely given that care for BPD has not changed substantially over the past decade and length of NICU stay did not change over time in our cohort.[28,29] Our data collection overlapped with the beginning of the coronavirus-19 pandemic, though our data largely represent outcomes prior to this period. Additionally, our data on post-discharge diagnoses—asthma, cerebral palsy, and autism— are gathered through parental report without validation through medical records. However, it seems unlikely that parents would misremember such significant diagnoses, particularly in the direction of overestimating such diagnoses.[30] A final limitation is our response rate. The severity of a child’s outcome may have correlated with the likeliness of parents responding to this survey, though the direction of bias incurred by such missingness is unpredictable. Importantly, we did not identify significant differences in baseline characteristics between respondents and non-respondents and, reassuringly, our response rate is better than many similar studies.[23,31] Despite these limitations, we believe that this study provides novel data to augment our understanding of the long-term outcomes for children discharged from a referral BPD program and will be useful to clinicians and families.

Conclusion

The present study captures numerous outcomes after discharge for children with the severest form of BPD. While respiratory morbidities improve with age, problems of development and their impact on quality of live become more prominent. These findings underscore the importance of measuring and targeting the developmental sequelae of BPD. The most direct application of these data is to frame expectations for both families and clinicians who care for referral BPD populations about possible medical and developmental sequelae through childhood. For such families, we can say with a high degree of certainty, for instance, that children will experience wheezing and need an IEP in school. The prevalence of autism in BPD has not been widely reported and the prevalence in this cohort would support additional investigation. Data from this cohort can also guide interventions aimed to improve the lives of children with severe BPD, which should focus particularly on developmental impacts. Finally, an enduring question in the care of children with BPD is how best to predict those at risk for severe outcomes.[32] Which children, even if outside a referral population, have disease so severe that they will ultimately experience outcomes like those of the cohort we present here? Additional predictors are needed to distinguish children who will ultimately experience medical and developmental sequalae during childhood.

Abbreviations

BPD

Bronchopulmonary dysplasia

NICU

Neonatal intensive care unit

HR QoL

Health-related quality of life

IEP

Individualized educational plan