Study Design
This was a retrospective cohort study based out of two-level 3 NICUs at Foothills Hospital in Calgary, Canada, and the University of Texas Medical Branch (UTMB) in Galveston, United States. The study was approved by the Conjoint Health Research Ethics Board of the University of Calgary, and the UTMB Institutional Review Board. Both NICUs implemented a BPD screening program, including a cardiac care pathway to detect PH by echocardiography at 36 weeks postmenstrual age (PMA) in infants with BPD.
Study Population and sample
The study population included all infants born at < 29 weeks GA between January 2016 and December 2019, admitted to the NICUs at Foothills Medical Center in Calgary, Canada, and the University of Texas Medical Branch (UTMB) in Galveston, United States. Neonatal data of these infants were collected retrospectively; however, follow-up data on these infants were collected prospectively at 18–24 months CA. Infants with significant congenital anomalies, complex congenital heart disease and chromosomal disorders were excluded from the study.
Variables and measurement
The exposure variable of interest was PH. Infants were divided into two groups based on the presence of PH: Group 1: BPD with PH and Group II: BPD with no PH. All premature infants born at < 29 weeks are routinely followed up in the Neonatal follow up clinics (NFC), at both centers, at the ages of four months, eight months, 18–24 months, 36 months, and five years with a multidisciplinary team. Both centers use the Bayley III for neurodevelopmental assessment at follow-up. From April 2020, routine Bayley assessments were stopped due to the COVID-19 pandemic. To enable ongoing developmental assessments of high-risk neonates, the neonatal follow-up clinics had to rely on the Ages and stages questionnaire version 3 (ASQ 3) as a screening tool. This was followed by tailored assessments in the follow-up clinics for infants identified as having concerns on ASQ 3. ASQ is an easy-to-use tool consisting of a set of questions to be answered by parents, which takes about 15 minutes to answer. ASQ has been studied as a screening tool to identify developmental issues in various populations and has shown modest agreement with other formal methods of neurodevelopmental assessment such as Bayley III and WPSSI [17, 18]. One study reported that the sensitivity, specificity, and correlations between measures improved in children with risk factors such as prematurity and increased age at assessment [18]. We used ASQ 3 scores for ND outcomes as a surrogate marker of NDI, especially in the pandemic period when infants did not have Bayley scores due to virtual follow-up.
Data Source
Standardized demographic, perinatal and neonatal data were collected from patients’ medical records and a research coordinator and entered into a computerized neonatal follow-up clinic database upon discharge from the NICU. All surviving premature infants underwent comprehensive developmental assessment by a multidisciplinary team (consisting of a neonatologist or developmental pediatrician, psychologist, occupational therapist, physiotherapist, dietician, speech language pathologist, social worker, nurse, ophthalmologist, audiologist). The Bayley-III was administered at 18–24 months CA by a trained psychologist/psychometrist, audiologist and speech-language pathologist. Members of the multidisciplinary team were not blinded to the infant’s neonatal hospital course details, which may introduce the risk of detection bias.
Definition of BPD and PH
Bronchopulmonary dysplasia was defined as oxygen dependency at 36 weeks’ postmenstrual age (PMA). Infants with any degree of PH were included in the BPD-PH group. PH was diagnosed at 36 weeks’ PMA based on echocardiography features of interventricular septal flattening, right ventricular systolic pressure /systemic systolic blood pressure ratio as measured by tricuspid regurgitation envelope and presence of main pulmonary artery dilation.
Outcome measures
The primary outcome for this study was a composite of mortality or NDI at 18–24 months of CA. Secondary outcomes included individual components (cognitive, language, motor) of the primary outcome, cerebral palsy, visual impairment, and sensorineural or mixed hearing loss.
NDI was defined as a Bayley-III score of < 85 on any one of the components (cognitive, language, motor composite score), and severe NDI defined as a score of < 70 on any one of the components.
Covariates
For infants in both groups, perinatal, birth and neonatal data were extracted from both centers’ neonatal follow-up clinics’ databases. Perinatal factors included maternal hypertension, maternal diabetes, chorioamnionitis, and antenatal corticosteroids (ANCS). Birth data included gestational age, sex, Apgar scores at 5 minutes and small for gestational age (SGA) status. Neonatal morbidities included respiratory distress syndrome (RDS), duration of mechanical ventilation, blood culture or CSF culture-proven sepsis, patent ductus arteriosus (PDA) and its treatment, necrotizing enterocolitis (NEC) Bell’s criteria[19], intraventricular hemorrhage (IVH) grade 3 or higher as defined by Papile classification[20].
Statistical analysis
Maternal and neonatal characteristics and outcomes were compared between the two groups using the Chi-square test or Fisher’s exact test for categorical variables and the t-test or Wilcoxon rank sum test for continuous variables. Logistic regression models were used to adjust for confounding variables. Confounders included gestational age at delivery, ANCS, SGA, PDA requiring treatment, chorioamnionitis and NEC. Bootstrap analysis was done to determine the confidence intervals (CI) of the adjusted odd ratio (aOR). Data were analyzed using SAS version 9.4 (SAS Institute Inc, Cary, NC, USA).
Post hoc Power Calculations
We used a convenience sample of all eligible infants born in the study period for whom we had data for echocardiography screening for PH. With 58 infants in Group I and 308 in Group II, and 60% of infants with BPD without PH having the primary outcome, we had 80% power to detect a difference as small as 20% between the two groups.