Michael Vincer

Purpose: Is machine learning (ML) superior to the traditionally used logistic regression (LR) in prediction of neurodevelopmental outcomes in preterm infants? Objectives: To develop and internally validate a ML model to predict neurodevelopmental impairment (NDI) in very preterm infants (0–30⁶ weeks) born between January 2004 and December 2016 in Nova Scotia, Canada. Survivors with neurodevelopmental assessment at 36 months corrected age were included. The study sample was randomly split (80:20) into a development and testing datasets. We compared four methods: LR, elastic net (EN), random forest ensemble (RF) and gradient boosting (XGB), in relation to discrimination (AUC), calibration, and diagnostic properties. Results: Of 811 eligible infants, 663 were included (mean gestational age 28 weeks, mean birth weight 1137 g and 52% male). Of those, 195 (29%) developed NDI and 468 (71%) did not. On internal validation using the testing dataset, all four models provided good discrimination of NDI with comparable AUC. RF was superior to the other three methods with a higher AUC (0.79 vs. 0.74, 0.74, and 0.73 for XGB, EN and LR, respectively), but all models have overlapped CIs. Conclusions: In this population-based cohort of very preterm infants, RF was superior to conventional LR in prediction of NDI at 3 years corrected age. Accurate prediction of preterm infants at risk of NDI enables early referrals for intervention programs and resources allocation toward those who are most likely to benefit.

Objective Accurate and early prediction of neurodevelopmental impairment is a crucial endeavor in caring for very preterm infants (0–30⁶ weeks) born between January 2004 and December 2018 with a neurodevelopmental assessment at 36 months corrected age were included. Three prediction models of neurodevelopmental impairment were compared: 1. A reference model including the gestational age, infant sex, and 2-week and 6-week ultrasound 2. A model including the gestational age, infant sex, and 6-week ultrasound 3. A model including the gestational age, infant sex, and 2-week ultrasound Results Of 786 eligible preterm infants born during the study period, 656/786 survivors were included in the analysis (mean gestational age 27⁵ weeks, mean birth weight 1,133 g, and 55% male infants). At 36 months of corrected age, 30% developed neurodevelopmental impairment. All three models provided comparable discrimination areas under the curve (AUC) of neurodevelopmental impairment at 36 months of corrected age. Both the 6-week and the reference model had similar AUC of 0.68 (95% CI 0.63–0.72) and were not noticeably different from the 2-week model (AUC 0.66 (95% CI 0.61–0.70)). The 6-week model provided the best prediction with the lowest Akaike information criterion (AIC) of 766 for the 6-week-only model, AIC 769 for combined weeks 2 and 6 (reference model), and AIC 784 for the 2-week-only model. Conclusion In this cohort of very preterm infants, a model including 6-week ultrasound only was comparable to a reference model combining 2-week and 6-week ultrasound and showed nearly identical predictive performance of neurodevelopmental impairment at 36 months corrected age across a broad set of metrics; thus, it is redundant to do both the 2-week and 6-week ultrasound. Clinical relevance statement Late ultrasound at 6 weeks of age provided comparable diagnostic and prognostic information to a reference model combining 2-week and 6-week ultrasound and, if anything, was slightly superior to the 2-week ultrasound model, across a broad set of metrics. The 2-week ultrasound can be eliminated with no impact on the prediction of neurodevelopmental impairment at 36 months, promoting prudent resource allocation and stewardship in healthcare. Graphical Abstract

Objective: The aim of this study was to evaluate the neurodevelopmental outcome at 18–24 months in surviving preterm infants with grades I–IV intraventricular hemorrhages (IVHs) compared to those with no IVH. Study Design: We included preterm survivors <29 weeks’ GA admitted to the Canadian Neonatal Network’s NICUs from April 2009 to September 2011 with follow-up data at 18–24 months in a retrospective cohort study. The neonates were grouped based on the severity of the IVH detected on a cranial ultrasound scan and recorded in the database: no IVH; subependymal hemorrhage or IVH without ventricular dilation (grades I–II); IVH with ventricular dilation (grade III); and persistent parenchymal echogenicity/lucency (grade IV). The primary outcomes of neurodevelopmental impairment (NDI), significant neurodevelopmental impairment (sNDI), and the effect modification by other short-term neonatal morbidities were assessed. Using multivariable regression analysis, the adjusted ORs (AOR) and 95% of the CIs were calculated. Results: 2327 infants were included. The odds of NDI were higher in infants with grades III and IV IVHs (AOR 2.58, 95% CI 1.56, 4.28 and AOR 2.61, 95% CI 1.80, 3.80, respectively) compared to those without IVH. Infants with an IVH grade ≤II had similar outcomes for NDI (AOR 1.08, 95% CI 0.86, 1.35) compared to those without an IVH, but the odds of sNDI were higher (AOR 1.58, 95% CI 1.16, 2.17). Conclusions: There were increased odds of sNDI in infants with grades I–II IVHs, and an increased risk of adverse NDI in infants with grades ≥III IVHs is corroborated with the current literature.

Inhaled corticosteroids (ICS) offer targeted treatment for bronchopulmonary dysplasia (BPD) with minimal systemic effects compared to systemic steroids. However, dosing of ICS in the management of infants at high-risk of developing BPD is not well established. The objective of this study was to determine an effective dose of ICS for the treatment of ventilator-dependent infants to facilitate extubation or reduce fractional inspired oxygen concentration. Forty-one infants born at < 32 weeks gestational age (GA) or < 1250 g who were ventilator-dependent at 10–28 days postnatal age were included. A non-randomized dose-ranging trial was performed using aerosolized inhaled beclomethasone with hydrofluoralkane propellant (HFA-BDP). Four dosing groups (200, 400, 600 and 800 μg twice daily for 1 week) with 11, 11, 10 and 9 infants in each group, respectively, were studied. The primary outcome was therapeutic efficacy (successful extubation or reduction in FiO2 of > 75% from baseline) in ≥60% of infants in the group. Oxygen requirements, complications and long-term neurodevelopmental outcomes were also assessed. The median age at enrollment was 22 (10–28) postnatal days. The primary outcome, therapeutic efficacy as defined above, was not achieved in any group. However, there was a significant reduction in post-treatment FiO2 at a dose of 800 μg bid. No obvious trends were seen in long-term neurodevelopmental outcomes. Therapeutic efficacy was not achieved with all studied doses of ICS. A significant reduction in oxygen requirements was noted in ventilator-dependent preterm infants at 10–28 days of age when given 800 μg of HFA-BDP bid. Larger randomized trials of ICS are required to determine efficacy for the management of infants at high-risk for development of BPD. This clinical trial was registered retrospectively on clinicaltrials.gov. The registration number is NCT03503994 .

Skin-to-skin contact (SSC) between mother and infant, commonly referred to as Kangaroo Mother Care (KMC), is recommended as an intervention for procedural pain. Evidence demonstrates its consistent efficacy in reducing pain for a single painful procedure. The purpose of this study is to examine the sustained efficacy of KMC, provided during all routine painful procedures for the duration of Neonatal Intensive Care Unit (NICU) hospitalization, in diminishing behavioral pain response in preterm neonates. The efficacy of KMC alone will be compared to standard care of 24% oral sucrose, as well as the combination of KMC and 24% oral sucrose.