Methods:
Study Population : A retrospective cohort study was conducted
examining premature neonates born at less than 28 weeks gestation from
2010-2018 at a single center. The study was approved by the Tufts Health
Sciences Institutional Review Board at Tufts Medical Center. In Epoch 1
(2010-2014), the clinical practice in our center had been to routinely
intubate these infants in the delivery room and administer early
surfactant therapy. Following participation in the SUPPORT trial which
demonstrated the benefits of early continuous positive airway pressure
(CPAP) instead of surfactant administration in the delivery room
[6], respiratory care practice changed and early intubation only
occurred with poor respiratory effort or apnea, increased oxygen
requirement on CPAP of 5-6cmH2O, or significant
hypercarbia. To more clearly define this practice change, the clinical
database was queried for all neonates born at less than 28 0/7 weeks
without significant congenital anomalies for each year from 2010-2018.
Between 2014 and 2015, a sharp decrease in intubation/surfactant
instillation was noted in both the delivery room and in the first 24
hours of life (Table 1). Decision was made to limit subject enrollment
to the end of 2018 given that 1) our mode of CPAP delivery had changed
from ventilator-derived and VIASYS Infant Flow driver apparatuses to
bubble CPAP and 2) Common Rule statutes on retrospective human subject
research limited further enrollment. No other significant respiratory
practice changes were made during this period, including caffeine use
and oxygen saturation targets, nor did we use Vitamin A in our ELBW
population. All infants in our center are started on caffeine on
admission to NICU immediately after birth. The number of deaths,
presence or absence of BPD based on these 4 definitions, and the
composite outcome of death or BPD were compared between Epochs 1 and 2.
Maternal, Infant and Postnatal Characteristics : Maternal,
neonatal, and postnatal characteristics known to influence the
development of BPD were compared between Epochs 1 and 2, including
antenatal steroids, mode of delivery, any pregnancy and labor
complications, and any postnatal conditions that may have contributed to
neonatal illness severity and potential confounders that could influence
outcome of BPD [15-18].
BPD Definitions: The presence or absence of BPD was defined by:
1) O2 use at 36 weeks PMA (VON Definition) [11], 2)
NIH consensus severity-based definition [12], 3) the Canadian
Neonatal Network criteria of O2 use at 40 weeks PMA
[13], and 4) Jensen-NRN criteria (henceforth referred to as NRN
criteria) identifying mode of ventilation irrespective of
O2 use at 36 weeks PMA [14].
Clinical Factors Impacting BPD Severity : Further analyses of
pre- and postnatal factors influencing BPD severity were performed
comparing those neonates who developed moderate/severe BPD using the NIH
severity-based definition in Epoch 1 versus 2. Given the more severe
spectrum of clinical disease that these infants possessed, and the
knowledge that they would likely have a greater utilization of medical
resources and a guarded long-term prognosis, we sought to understand the
differences in their hospital course between Epochs 1 and 2.
Statistics: Univariate analyses were performed to examine
differences in demographic and clinical variables between epochs and
between neonates who did and did not develop BPD. Continuous variables
were compared between groups using students two-tailed t-test and
categorical variables were compare using chi-square tests. All
statistical testing was two-sided with alpha=0.05. A logistic regression
model was performed for the composite outcome of BPD (based on VON
definition and NIH moderate to severe definition) and death, adjusting
for each of the potential confounders as predictors, including
chorioamnionitis, mode of delivery, multiple births, sex, inborn/outborn
status.