Background: Newborn resuscitation is a time-pressured activity requiring teams of healthcare professionals to coordinate invasive procedures in a specific sequence of steps in a highly confined physical space. Audiovisual recordings of resuscitations provide both more accuracy than written records and ideal substrate upon which to debrief.
Objective: Implement and evaluate a HIPAA-compliant system for recording resuscitations.
Methods: We worked with our hospital’s Risk Management and Compliance Departments to place cameras with built-in microphones in a subset of our delivery rooms to record neonatal resuscitations. Selected recordings are reviewed monthly during formal confidential debriefings attended by the members of our multidisciplinary resuscitation teams.
Results: In its first five years, the system has captured > 4300 deliveries, 17.6% of which involved CPAP or higher levels of resuscitation. Each such resuscitation has been scored for technical and behavioral performance. Debriefings have resulted in performance improvements including but not limited to timely removal of wet linen, reduction of unnecessary suctioning, improvement in bedside ergonomics, establishment of limitations
on attempts at intubation, clarification of chain of command issues, and implementation of guidelines designed to activate and optimize the number of healthcare professionals at deliveries.
Conclusion: Although challenges exist to recording neonatal resuscitations, all can be overcome with a methodical approach that is mindful of the legal, cultural and psychological issues inherent in this clinical activity. Once established, such a program provides an extremely valuable quality improvement tool that reduces error, enhances training, and serves as an objective and powerful adjunct to debriefings.

Introduction
A significant proportion of term neonates will develop physiological jaundice, some requiring overhead phototherapy treatment. One approach to decrease need for phototherapy and mother-infant separation is use of a ‘Biliblanket’.
We hypothesised that using ‘Biliblanket’ in infants approaching the treatment line for phototherapy would decrease phototherapy use.

Methods
A single-centre randomised control trial was carried out in a level III maternity hospital. Infants ≥35 weeks gestation and >24 hours of age with suspected physiological jaundice were eligible if they had a serum bilirubin within 35µmol/L of the treatment line for their age. Infants admitted to the neonatal unit or with a positive pre-randomisation direct Coombs test were excluded. Sample size calculation required 52 infants per group to identify a reduction in phototherapy from 35% to 24% (80% power and alpha 0.05).
Infants were randomised to two groups: intervention arm (treatment with Biliblanket) or control arm (routine care). Primary outcome was admission for phototherapy.
The trial was registered - ISRCTN29045982.

Results
One hundred and ten infants were recruited with 6 excluded due to breach of entry criteria or unavailability of data. 104 were included in an intention-to-treat analysis, 58 in intervention arm and 46 in control arm. Baseline characteristics were similar.
Eight (14%) infants in the intervention group and 7 (15%) infants in the control group required overhead phototherapy, with no statistically significant difference (p=0.99).

Conclusion
Our results suggest that the use of a ‘Biliblanket’ to treat physiological jaundice approaching the treatment line does not decrease need for overhead phototherapy.

Background:
Hypoxic ischaemic encephalopathy(HIE) is one of the leading causes of morbidity and mortality in infants worldwide. We have previously reported a machine learning algorithm which may support clinical decision making in neonatal hypoxic ischaemic encephalopathy.

Objective:
We aimed to validate a HIE prediction algorithm to identify infants at risk of HIE immediately after birth using readily available clinical data.

Methods:
A retrospective review of electronic health record data of all term deliveries between January 2017-December 2021. Infants>36 weeks’ gestation with the following clinical variables available were included; Apgar score at one and five minutes, postnatal pH, base deficit, and lactate values taken within one hour of birth. Previously trained open-source logistic regression(LR) and random forest(RF) prediction algorithms (https://www.infantcentre.ie/predictionapp.html) were used to calculate a probability index(PI) for each infant for the occurrence of HIE.

Results:
Over the study period, 1093 infants had a complete dataset available: 76 with HIE and 1017 non-HIE. 46(60.5%) infants received therapeutic hypothermia (7[19%] mild, 29[100%] moderate and 10[100%] severe). Median(IQR) probability index in the HIE group was 0.70(0.52-0.86) vs 0.05(0.016-0.16) (p<0.001) in the non-HIE group. AUROC for prediction of HIE = 0.925(0.893-0.958), (p<0.001). Using a PI cut off to optimise sensitivity of 0.30; 947 of the 1093(86.6%) infants were correctly classified by the algorithm (sensitivity=86%, specificity=87%). This cut off identified 78.4% of mild HIE, 93% of moderate HIE and 100% of severe HIE cases.

Conclusion:
In a large unseen dataset, an open-source algorithm could identify infants at risk of HIE in the immediate postnatal period.

Background:
Our group previously demonstrated that a patent ductus arteriosus severity score (PDAsc) showed promise in reducing the rate of chronic lung disease (CLD) when used to identify and treat high-risk infants with a haemodynamically significant(hs) PDA.

Objective:
To present the impact of early haemodynamic screening and treatment using the PDAsc in a cohort of preterm infants.

Methods:
Retrospective cohort study, prospectively gathered data. A cohort of infants <30 weeks from July 2020 onwards who received a targeted hsPDA treatment approach based on the PDAsc (Epoch 2) were compared with a historical high-risk cohort who underwent screening using the PDAsc but were not offered treatment (Epoch 1). Echocardiogram was performed between 24-48 hours. Treatment was with ibuprofen/paracetamol if deemed high-risk (PDAsc ≥5.0). Repeat treatment was administered if PDA patency was documented after initial treatment. The composite outcome of CLD/mortality and its individual components were assessed.

Results:
Eighty-five infants from Epoch 2 were compared to 104 infants from Epoch 1. Demographics and outcomes in Table 1. Infants in Epoch 1 had a higher birthweight and rate of males and a lower use of antenatal steroids. No difference in PDA diameter or PDAsc between groups. Infants in Epoch 2 had a lower number of ventilation days, CLD/Death, CLD alone. No difference in other outcomes including mortality (all p-values and odds ratios adjusted for birthweight, male sex, antenatal steroids).

Conclusion:
Precise identification of high-risk infants is key to the success of early-targeted PDA therapy and may have an effect on reducing respiratory morbidity.

Introduction
With improved quality and availability of MR imaging, cranial ultrasound (cUS) is used less frequently to evaluate hypoxic ischemic injury. We hypothesize that cUS has a role in prognosis and timing of injury.
Aims
To assess whether cUS can help predict timing of injury in HIE. To assess the correlation between neuroimaging scoring systems (cUS and MR) to predict neurodevelopment outcome.
Methods
This was a retrospective study of infants with HIE admitted to a single tertiary NICU between 2010 -2021. Validated scoring systems were used to evaluate neuroimaging performed as part of routine clinical care.[1][2].
Results
One hundred and five infants with cUS suitable for scoring were included, of which 94 had MR imaging suitable for scoring. Seventy-four (70%) had moderate encephalopathy and 31 (30%) had severe encephalopathy. Fifty-seven (40%) infants had evidence for an acute sentinel event.
There was a strong correlation between cUS injury within 12 hours (n=28) of birth and first pH after birth. Infants with a higher pH were more likely to have white matter injury (r=0.39, p=0.039), suggesting a subacute/chronic insult.
The strength of the relationship between cUS and MR total injury scores differed depending on timing of cUS (Table 1).
Conclusion
This study supports the use of early cUS to time injury in HIE. White matter injury seen on early cUS supports a sub-acute prenatal injury as the cause of HIE. cUS performed from 48 hours after birth correlates well with MR injury in HIE and may be useful adjunct to prognosis.