Highlights
- •
Infants with BPD receiving higher levels of respiratory support on discharge from hospital is associated with poorer NDI.
- •
Understanding the relative risks of NDI may facilitate better-informed counselling of parents regarding outcomes at age 3 years.
- •
Subclassifying BPD infants based on levels of respiratory support received at home is a pragmatic way to predict outcome.
Educational Aims
The reader will come to appreciate that:
- •
The likelihood of infants with BPD developing poorer childhood neurodevelopmental outcomes is associated with the level of respiratory support provided on discharge from hospital.
- •
Understanding the relative risks of neurodevelopmental impairment may facilitate better-informed counselling of parents regarding clinical outcomes at age 3 years.
Abstract
Objective
To investigate the neurodevelopmental outcomes for preterm infants born < 29 weeks gestation with/without bronchopulmonary dysplasia (BPD).
Study design
Preterm infants < 29 weeks’ gestation born 2007–2018 in New South Wales and the Australian Capital Territory, Australia, were included. Infants who died < 36 weeks’ postmenstrual age and those with major congenital anomalies were excluded. Subjects were assessed at 18–42 months corrected age using the Bayley Scales of Infant Development, 3rd edition.
Results
1436 infants without BPD (non-BPD) and 1189 infants with BPD were followed. The BPD group, 69 % infants were discharged without respiratory support (BPD1), 29 % on oxygen (BPD2) and 2 % on pressure support/tracheostomy (BPD3). Moderate neurodevelopmental impairment (NDI) was evident in 5.7 % of non-BPD infants, 11 % BPD1, 15 % BPD2, 15 % BPD3 infants. Severe NDI was seen in 1.7 % non-BPD infants, 3.4 % BPD1, 7.3 % BPD2, 35 % BPD3 infants. After adjusting for confounders, infants with BPD2 (OR 2.24, 99.9 % CI 1.25 to 5.77) or BPD3 (OR 5.99, 99.9 % CI 1.27 to 46.77) were more likely to have moderate-severe NDI compared to non-BPD infants.
Conclusion
The majority of infants with BPD were discharged home without respiratory support and had better neurocognitive outcomes in early childhood compared to those that required home-based oxygen or respiratory support.
Introduction
Infants with bronchopulmonary dysplasia (BPD) have more intermittent hypoxaemia with feeding, sleeping and activity which places them at risk of hypoxic-related injuries 1 . A study revealed BPD infants experienced a median of 23 episodes of low oxygen saturation (<90 %) over 8 h of monitoring . Infants with increasing severity of BPD have an increased risk of cerebral palsy and neurosensory problems such as blindness, deafness and cognitive delay . Recent data suggested that after accounting for known confounding variables, only those with severe BPD had an increased risk of neurodevelopmental impairment (NDI) compared to non-BPD infants despite an increase in absolute percentage of NDI with each level of BPD severity assessed at 2 and 5 years .
There has been an expanding use of home-based non-invasive ventilation (NIV) treatment in paediatric patients facilitated by an emphasis on earlier discharge, improvements in technology and a growing acceptance as a long-term respiratory therapy . However, few studies have exclusively focused on its use in extremely preterm neonates with respiratory failure associated with BPD. It is possible that wider access to NIV treatment could improve the respiratory status in severe BPD infants which could translate to improved gas exchange and developmental outcomes by limiting the risk of hypoxic-related injuries .
Given the increasing numbers of extremely preterm survivors, and the paucity of data comparing medium-term neurocognitive outcomes for BPD infants discharged with or without home-based respiratory support, we explored their neurodevelopmental outcomes up to age 3.5 years.
Methods
Study type
A retrospective, multicentre cohort study from the Neonatal Intensive Care Units (NICU) database of New South Wales (NSW) and the Australian Capital Territory (ACT), Australia. Ethics approval: the Sydney Children’s Hospital Network Human Research Ethics Committee (2020/ETH03066).
Study population
Infants < 29 weeks’ gestation admitted to a NICU between 1 January 2007 and 31 December 2018. Infants who died prior to 36 weeks postmenstrual age (PMA), those with major congenital malformations or with incomplete data were excluded. Bronchopulmonary dysplasia (BPD) is defined as the dependence on any form of respiratory therapy [O 2 , NIV, tracheostomy] at 36 weeks PMA in babies born < 32 weeks’ gestation 1, . Each infant was categorised into a BPD group or a non-BPD group. Within the BPD group: BPD1- infants discharged without any respiratory support, BPD2- discharged on continuous low flow oxygen and BPD3- discharged on other respiratory support including NIV, tracheostomy or invasive ventilation.
Data source
Data were obtained from The Neonatal Intensive Care Units’ Data Registry (NICUS), which is an ongoing prospective statewide audit of infants admitted to the ten neonatal intensive care units (NICU) perinatal centres and two children’s hospitals). Definitions used and data quality have been described elsewhere .
Outcome
Outcomes were defined according to NICUS Data Collection Manual . Neurodevelopmental outcome was assessed at different ages according to individual NICU follow-up policies (18–42 months corrected age) by a multidisciplinary team using Bayley Scales of Infant Development, 3rd edition (BSID-III). NDI is defined as the presence of one or more of the following: Cerebral palsy with impairment classification of Gross Motor Function Classification System (GMFCS) ≥ 1; Motor and/or language and/or cognitive score > 1 standard deviation [SD] below the mean on the BSID-III; severe visual impairment with visual acuity < 6/60 in the better eye; hearing impairment with hearing aids and/or cochlear implant .
Mild NDI was defined as ambulant cerebral palsy (GMFCS level 1 or 2), no hearing or vision impairment, or a BSID composite score of 1–2 SD below the mean (70–85). Moderate NDI was defined as cerebral palsy (GMFCS level 3), hearing loss requiring bilateral hearing aids or unilateral/bilateral cochlear implants or a BSID composite score of 2–3 SD below the mean (55–69). Severe NDI was defined as cerebral palsy (GMFCS levels 4–5), bilateral blindness or a BSID composite score > 3 SD below the mean (<55).
Statistical analysis
Antenatal and perinatal characteristics and major morbidities were compared across all infants at review. Neurodevelopmental outcomes were compared between non-BPD infants and infants with BPD severity groups. Comparisons were made using independent sample t-tests, ANOVA tests or Kruskal-Wallis tests for continuous characteristics, log-rank tests for time-to-event characteristics, and chi-squared tests (or Fisher’s exact tests where any sub-group with n < 5) for categorical characteristics.
Logistic regression was used to investigate whether the diagnosis of BPD [all severities] or the need for home oxygen or positive pressure support [NIV/tracheostomy] was independently associated with adverse neurodevelopmental outcomes at follow-up, after adjusting for suspected confounders. Potential confounders of the relationship between BPD and the neurodevelopmental outcomes were identified using univariate logistic regression models, and final models were selected using backwards step-wise selection via p-values and the Akaike information criteria. The area under the curve of the ROC curve and Akaike Information Criteria was calculated for the models. As many associations with BPD were investigated and no specific pre-planned hypotheses were specified prior to the analysis, a Bonferroni adjustment to the statistical significance level was performed to control the Type I error rate. Therefore, an adjusted significance cut-off of p < 0.001 was used for all analyses. All analyses were carried out using R v4.1.2 .
Results
Infants born < 29 weeks’ gestation 2007–2018 who survived to 36 weeks PMA totalled 4100 infants of whom 1842 (45 %) had BPD (57 % male) and 2258 (55 %; 51 % male) did not have BPD ( Fig. 1 ).
Perinatal characteristics
Infants with BPD were born earlier (25.97 +/- 1.39 weeks vs 27.05 +/- 1.09 weeks; P <.001), and of lower birthweights (854 +/- 208 g vs 1055.94 +/- 211.83 g; P <.001) compared to the non-BPD group.
More BPD infants received surfactant therapy (92.7 % vs 79.1 %; P <.001) and more doses of surfactant (1.67 +/- 0.76 vs 1.35 doses +/- 0.57; P <.001) compared to non-BPD infants. BPD infants were more likely to receive postnatal steroids (41 % vs 5.2 %; P <.001) and more days of steroids (17.78 +/- 25.70 vs 8.64 +/-7.10 days; P <.001) compared to non-BPD infants. BPD infants were more likely to receive longer periods of intermittent mandatory ventilation, high frequency oscillatory ventilation, continuous positive airway pressure (CPAP), nasal high flow oxygen and low flow oxygen compared to non-BPD infants ( Table 1 ). BPD infants were more likely to have spontaneous air leak requiring drainage (3.9 % vs 1.8 %; P <.001) and require a tracheostomy (0.4 % vs 0 %; P <.001).
| Characteristic | Total (n = 4100) | BPD group (n = 1842) | Non BPD group (n = 2258) | P value |
|---|---|---|---|---|
| Maternal features | ||||
| Maternal age, years | 30.38 (6.25) | 30.25 (6.14) | 30.49 (6.34) | 0.2 |
| Pregnancy induced hypertension | 16 % (6 7 2) | 18 % (3 2 4) | 15 % (3 4 8) | <0.001 |
| Pre-labour rupture of membranes, hours | 81.72 (237.69) | 101.56 (292.59) | 65.53 (179.42) | <0.001 |
| Gestational diabetes | 9.8 % (4 0 0) | 8.7 % (1 6 0) | 11 % (24) | 0.04 |
| Chorioamnionitis (proven) | 27 % (1098) | 28 % (5 1 1) | 26 % (5 8 7) | 0.2 |
| Antenatal magnesium sulphate given | 41 % (1675) | 45 % (8 2 8) | 38 % (8 4 7) | <0.001 |
| Antenatal steroid completed | 49 % (1994) | 50 % (9 1 2) | 48 % (1082) | 0.3 |
| Preterm labour | 70 % (2880) | 69 % (1266) | 71 % (1614) | 0.06 |
| Caesarean section | 60 % (2472) | 59 % (1089) | 61 % (1383) | 0.2 |
| Birth features | ||||
| Gestational age, week | 26.57 (1.35) | 25.97 (1.39) | 27.05 (1.09) | <0.001 |
| Male gender | 54 % (2199) | 57 % (1054) | 51 % (1145) | <0.001 |
| Birth weight, gram | 965.26 (232.88) | 854.10 (208.05) | 1055.94 (211.83) | <0.001 |
| Small for gestation age (<10 centile) | 6.6 % (2 7 1) | 12 % (2 1 5) | 2.5 % (56) | <0.001 |
| Head circumference, cm | 24.72 (2.08) | 23.92 (2.05) | 25.38 (1.88) | <0.001 |
| Head circumference (<10 centile) | 2.7 % (1 1 1) | 4.2 % (78) | 1.5 % (33) | <0.001 |
| Outborn | 9.0 % (3 6 8) | 7.7 % (1 4 2) | 10 % (2 2 6) | 0.01 |
| Apgar score < 7 at 5 min | 26 % (1046) | 33 % (6 0 5) | 20 % (4 4 1) | <0.001 |
| Respiratory features | ||||
| Surfactant given | 85 % (3495) | 93 % (1708) | 79 % (1787) | <0.001 |
| Time surfactant given – Median (IQR) | 23 (58) | 22 (46) | 25 (70) | 0.01 |
| Postnatal steroids | 21 % (8 6 8) | 41 % (7 5 0) | 5.2 % (1 1 8) | <0.001 |
| Duration postnatal steroids, days | 16.84 (24.61) | 17.78 (25.70) | 8.64 (7.10) | <0.001 |
| Early postnatal steroids (<7 days old) Late postnatal steroids (>7 days old) | 85 % (7 4 1) 2.4 % (21) | 89 % (6 6 9) 2.0 % (15) | 61 % (72) 5.1 % (6) | 0.02 0.02 |
| IMV, hours | 132.88 (681.15) | 246.24 (998.89) | 40.39 (98.54) | <0.001 |
| HFOV, hours | 27.78 (89.55) | 55.87 (123.24) | 4.87 (31.75) | <0.001 |
| Nasal IMV, hours | 69.35 (184.11) | 127.28 (243.07) | 22.10 (91.63) | <0.001 |
| CPAP, hours | 784.17 (512.15) | 1039.19 (539.57) | 576.13 (377.52) | <0.001 |
| Nasal HFlow, hours | 237.63 (325.00) | 327.46 (393.65) | 164.35 (231.24) | <0.001 |
| Low flow O2, hours | 672.64 (2093.50) | 1351.33 (2871.12) | 118.99 (743.70) | <0.001 |
| Tracheostomy | 0.20 % (8) | 0.43 % (8) | 0.0 % (0) | <0.001 |
| Nitric oxide, hours | 5.84 (36.25) | 11.47 (51.98) | 1.25 (11.66) | <0.001 |
| Postnatal complications | ||||
| Spontaneous air leak requiring drainage | 2.8 % (1 1 3) | 3.9 % (72) | 1.8 % (41) | <0.001 |
| PDA requiring surgical treatment | 4.3 % (1 7 5) | 7.6 % (1 4 0) | 1.6 % (35) | <0.001 |
| NEC requiring surgery | 2.4 % (98) | 3.6 % (66) | 1.4 % (32) | <0.001 |
| IVH grade 3 or 4 | 3.9 % (1 5 8) | 4.7 % (86) | 3.2 % (72) | 0.02 |
| ROP requiring treatment | 9.0 % (3 6 9) | 17 % (3 1 0) | 2.6 % (59) | <0.001 |
| Proven systemic infection – late onset | 29 % (1207) | 42 % (7 6 8) | 19 % (4 3 9) | <0.001 |
| Seizures | 1.4 % (58) | 2.2 % (40) | 0.80 % (18) | <0.001 |
| Length of hospital stay | 91.06 (36.88) | 110.40 (43.84) | 75.29 (18.67) | <0.001 |
| Died after 36 weeks without leaving hospital | 0.93 % (38) | 2.0 % (36) | 0.090 % (2) | <0.001 |
BPD infants were more likely to have seizures (2.2 % vs 0.80 %; P <.001) and late onset sepsis (>72 h of age) (42 % vs 19 %; P <.001) compared to non-BPD infants.
BPD infants required longer hospitalisation (110.40 days +/- 43.84 vs 75.29 days +/- 18.67; P <.001) and were more likely to die after 36 weeks without leaving hospital (2.0 % vs 0.091 %; P <.001) compared to non-BPD infants.
Non-participants
Of the 1842 babies with BPD, 361 had missing data and were excluded from the study. 1481 babies with BPD were discharged from the hospital. A total of 1189 BPD infants were followed up (65 %) and 292 were non-participants. Those followed up were more likely to be born earlier (25.95 weeks +/- 1.38 vs 26.26 +/- 1.40; P <.001) and have lower birth weight (840.3 g +/- 202.5 vs 917.3 g +/- 212.6; P <.001) when compared to non-participants ( Table 2 ).
| Characteristic | Total (n = 1481) | Followed up (n = 1189) | Not followed up (n = 292) | P value |
|---|---|---|---|---|
| Maternal features | ||||
| Maternal age, years | 30.15 (6.15) | 30.39 (5.06) | 29.19 (6.45) | 0.003 |
| Pregnancy induced hypertension | 20 % (2 9 0) | 20 % (2 4 3) | 16 % (47) | 0.1 |
| Pre-labour rupture of membranes, hours | 103.76 (301.63) | 95.79 (269.86) | 136.22 (405.08) | 0.05 |
| Gestational diabetes | 8.1 % (1 2 0) | 8.4 % (1 0 0) | 6.9 % (20) | 0.5 |
| Chorioamnionitis (proven) | 26 % (3 9 0) | 28 % (3 2 7) | 22 % (63) | 0.05 |
| Antenatal magnesium sulphate given | 38 % (5 5 6) | 38 % (4 4 6) | 38 % (1 1 0) | 0.8 |
| Antenatal steroid completed | 48 % (7 1 6) | 47 % (5 6 4) | 52 % (1 5 2) | 0.2 |
| Preterm labour | 69 % 1015 | 67 % (7 9 4) | 76 % (2 2 1) | 0.004 |
| Caesarean section | 59 % (8 6 7) | 59 % (6 9 9) | 58 % (1 6 8) | 0.7 |
| Birth features | ||||
| Gestational age, week | 26.01 (1.39) | 25.95 (1.38) | 26.26 (1.40) | <0.001 |
| Male gender | 56 % (8 2 8) | 56 % (6 6 5) | 56 % (1 6 3) | 1.0 |
| Birth weight, gram | 855.50 (206.72) | 840.33 (202.48) | 917.25 (212.57) | <0.001 |
| Small for gestation age (<10 centile) | 12 % (1 7 1) | 13 % (1 5 5) | 5.5 % (16) | <0.001 |
| Head circumference, cm | 23.94 (2.02) | 23.84 (1.98) | 24.38 (2.12) | <0.001 |
| Head circumference (<10 centile) | 4.0 % (57) | 4.3 % (51) | 2.1 % (6) | 0.1 |
| Outborn | 7.8 % (1 1 6) | 2.8 % (33) | 28 % (83) | 0.02 |
| Apgar score < 7 at 5 min | 32 % (4 8 0) | 33 % (3 9 3) | 30 % (87) | 0.3 |
| Respiratory Features | ||||
| Surfactant given | 93 % (1375) | 93 % (1102) | 93 % (2 7 3) | 0.7 |
| Time surfactant given – Median (IQR) | 22 (45) | 21 (39) | 28 (62) | 0.03 |
| Postnatal steroids | 38 % (5 6 7) | 39 % (4 6 1) | 106 (36) | 0.5 |
| Duration postnatal steroids, days | 17.45 (27.29) | 17.75 (29.49) | 16.12 (13.84) | 0.9 |
| Early postnatal steroids (<7 days old) Late postnatal steroids (>7 days old) | 88 % (4 9 9) 2.1 % (12) | 88 % (4 0 6) 2.4 % (11) | 88 % (93) 0.94 % (1) | 0.7 |
| IMV, hours | 264.24 (1104.65) | 278.33 (1224.60) | 206.84 (282.36) | 0.8 |
| HFOV, hours | 52.01 (116.17) | 52.84 (116.94) | 48.64 (113.09) | 0.5 |
| Nasal IMV, hours | 106.80 (219.57) | 111.09 (228.96) | 89.31 (175.50) | 0.08 |
| CPAP, hours | 1027.99 (496.08) | 1040.68 (495.58) | 976.31 (495.62) | 0.03 |
| Nasal HFlow, hours | 310.50 (384.74) | 313.35 (392.92) | 298.87 (349.84) | 0.7 |
| Low flow O2, hours | 1479.77 (3015.19) | 1573.96 (3128.07) | 1096.24 (2471.39) | 0.002 |
| Tracheostomy | 0.47 % (7) | 0.42 % (5) | 0.68 % (2) | 0.6 |
| Nitric oxide, hours | 8.65 (36.62) | 8.66 (37.37) | 8.58 (33.43) | 0.7 |
| Postnatal complications | ||||
| Spontaneous air leak requiring drainage | 4.0 % (59) | 4.0 % (48) | 3.8 % (11) | 1.0 |
| PDA requiring surgical treatment | 8.2 % (1 2 1) | 8.3 % (99) | 7.6 % (22) | 0.7 |
| NEC requiring surgery | 3.4 % (50) | 2.7 % (44) | 2.1 % (6) | 0.2 |
| IVH grade 3 or 4 | 4.2 % (62) | 3.9 % (46) | 5.5 % (16) | 0.3 |
| ROP requiring treatment | 16 % (2 4 4) | 17 % (2 0 1) | 15 % (43) | 0.4 |
| Proven systemic infection – late onset | 43 % (6 3 3) | 43 % (5 1 3) | 41 % (1 2 0) | 0.6 |
| Seizures | 2.3 % (34) | 2.5 % (30) | 1.4 % (4) | 0.6 |
| Length of hospital stay | 110.15 (45.06) | 110.81 (43.77) | 107.49 (49.98) | 0.1 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
