Prospective risk factor monitoring reduces intracranial hemorrhage rates in preterm infants

Predicting clinical outcomes using artificial intelligence and machine learning in neonatal intensive care units: a systematic review

BACKGROUND

Advances in technology, data availability, and analytics have helped improve quality of care in the neonatal intensive care unit.

OBJECTIVE

To provide an in-depth review of artificial intelligence (AI) and machine learning techniques being utilized to predict neonatal outcomes.

METHODS

The PRISMA protocol was followed that considered articles from established digital repositories. Included articles were categorized based on predictions of: (a) major neonatal morbidities such as sepsis, bronchopulmonary dysplasia, intraventricular hemorrhage, necrotizing enterocolitis, and retinopathy of prematurity; (b) mortality; and (c) length of stay.

RESULTS

A total of 366 studies were considered; 68 studies were eligible for inclusion in the review. The current set of predictor models are primarily built on supervised learning and mostly used regression models built on retrospective data.

CONCLUSION

With the availability of EMR data and data-sharing of NICU outcomes across neonatal research networks, machine learning algorithms have shown breakthrough performance in predicting neonatal disease.

Neuroprotection care bundle implementation is associated with improved long-term neurodevelopmental outcomes in extremely premature infants

OBJECTIVE

To study the impact of an evidence-based neuroprotection care (NPC) bundle on long-term neurodevelopmental impairment (NDI) in infants born extremely premature.

STUDY DESIGN

An NPC bundle targeting predefined risk factors for acute brain injury in extremely preterm infants was implemented. We compared the incidence of composite outcome of death or severe neurodevelopmental impairment (sNDI) at 21 months adjusted age pre and post bundle implementation.

RESULTS

Adjusting for confounding factors, NPC bundle implementation associated with a significant reduction in death or sNDI (aOR, 0.34; 95% CI 0.17-0.68; P = 0.002), mortality (aOR, 0.31; 95% CI (0.12-0.79); P = 0.015), sNDI (aOR, 0.37; 95% CI: 0.12-0.94; P = 0.039), any motor, language, or cognitive composite score <70 (aOR, 0.48; 95% CI: 0.26-0.90; P = 0.021).

CONCLUSION

Implementation of NPC bundle targeting predefined risk factors is associated with a reduction in mortality or sNDI in extremely preterm infants.

Advancements in neonatology through quality improvement

In the past 3 decades, quality improvement methodology has often been employed in medicine to improve patient outcomes. Neonatal medicine has seen an increase in publications using improvement science to ensure the application of potentially better practices to decrease complications and increase survival without major disability. This article reviews quality improvement studies that have impacted neonatal mortality and morbidity, as well as specific disease processes including bronchopulmonary dysplasia, intraventricular hemorrhage, necrotizing enterocolitis, and retinopathy of prematurity. Using improvement science, studies have substantially reduced neonatal mortality and the major complications of preterm birth.

Prevention of severe brain injury in very preterm neonates: A quality improvement initiative

OBJECTIVE

To determine the impact of neuroprotection interventions bundle on the incidence of severe brain injury or early death (intraventricular hemorrhage grade 3/4 or death by 7 days or ventriculomegaly or cystic periventricular leukomalacia on 1-month head ultrasound, primary composite outcome) in very preterm (27^0/7 to ≤ 29^6/7 weeks gestational age) infants.

STUDY DESIGN

Prospective quality improvement initiative, from April 2017-September 2019, with neuroprotection interventions bundle including cerebral NIRS, TcCO₂, and HeRO monitoring-based management algorithm, indomethacin prophylaxis, protocolized bicarbonate and inotropes use, noise reduction, and neutral positioning.

RESULT

There was a decrease in the incidence of the primary composite outcome in the intervention period on unadjusted (N = 11/99, pre-intervention to N = 0/127, intervention period, p < 0.001) and adjusted analysis (adjusted for birthweight and Apgar score <5 at 5 min, aOR = 0.042, 95% CI = 0.003-0.670, p = 0.024).

CONCLUSIONS

Neuroprotection interventions bundle was associated with significant decrease in severe brain injury or early death in very preterm infants.

Accounting for arterial and capillary blood gases for calculation of cerebral blood flow in preterm infants

One of the most feared neurological complications of premature birth is intraventricular hemorrhage, frequently triggered by fluctuations in cerebral blood flow (CBF). Although several techniques for CBF measurement have been developed, they are not part of clinical routine in neonatal intensive care. A promising tool for monitoring of CBF is its numerical assessment using standard clinical parameters such as mean arterial pressure, carbon dioxide partial pressure (pCO₂) and oxygen partial pressure (pO₂). A standard blood gas analysis is performed on arterial blood. In neonates, capillary blood is widely used for analysis of blood gas parameters. The purpose of this study was the assessment of differences between arterial and capillary analysis of blood gases and adjustment of the mathematical model for CBF calculation to capillary values. The statistical analysis of pCO₂ and pO₂ values collected from 254 preterm infants with a gestational age of 23-30 weeks revealed no significant differences between arterial and capillary pCO₂ and significantly lower values for capillary pO₂. The estimated mean differences between arterial and capillary pO₂ of 15.15 mmHg (2.02 kPa) resulted in a significantly higher CBF calculated for capillary pO₂ compared to CBF calculated for arterial pO₂. Two methods for correction of capillary pO₂ were proposed and compared, one based on the mean difference and another one based on a regression model.

CONCLUSION

Capillary blood gas analysis with correction for pO₂ as proposed in the present work is an acceptable alternative to arterial sampling for the assessment of CBF.

WHAT IS KNOWN

• Arterial blood analysis is the gold standard in clinical practice. However, capillary blood is widely used for estimating blood gas parameters. • There is no significant difference between the arterial and capillary pCO₂ values, but the capillary pO₂ differs significantly from the arterial one.

WHAT IS NEW

• The lower capillary pO₂ values yield significantly higher values of calculated CBF compared to CBF computed from arterial pO₂ measurements. • Two correction methods for the adjustment of capillary pO₂ to arterial pO₂ that made the difference in the calculated CBF insignificant have been proposed.

Impact of a “Brain Protection Bundle” in Reducing Severe Intraventricular Hemorrhage in Preterm Infants <30 Weeks GA: A Retrospective Single Centre Study

Background: despite advances in perinatal care, periventricular/intraventricular hemorrhage (IVH) continues to remain high in neonatal intensive care units (NICUs) worldwide. Studies have demonstrated the benefits of implementing interventions during the antenatal period, stabilization after birth (golden hour management) and postnatally in the first 72 h to reduce the incidence of IVH. Objective: to compare the incidence of severe intraventricular hemorrhage (IVH ≥ Grade III) before and after implementation of a “brain protection bundle” in preterm infants <30 weeks GA. Study design: a pre- and post-implementation retrospective cohort study to compare the incidence of severe IVH following execution of a “brain protection bundle for the first 72 h from 2015 to 2018. Demographics, management practices at birth and in the NICU, cranial ultrasound results and short-term morbidities were compared. Results: a total of 189 and 215 infants were included in the pre- and post-implementation phase, respectively. No difference in the incidence of severe IVH (6.9% vs. 9.8%, p = 0.37) was observed on the first cranial scan performed after 72 h of age. Conclusion: the implementation of a “brain protection bundle” was not effective in reducing the incidence of severe IVH within the first 72 h of life in our centre.

Extremely low gestational age infants: Developing a multidisciplinary care bundle

BACKGROUND

Clinical experience in managing extremely low gestational age infants, particularly those born <24 weeks' gestation, is limited in Canada. Our goal was to develop a bedside care bundle for infants born <26 weeks' gestation, with special considerations for infants of <24 weeks, to harmonize and improve quality of care.

METHODS

We created a multidisciplinary working group with experience in caring for preterm infants, searched the literature from 2000 to 2019 to identify best practices for the care of extremely preterm infants and consulted colleagues across Canada and internationally. Iterative improvements were made following the Plan-Do-Study-Act methodology.

RESULTS

A care bundle, created in October 2015, was divided into three time periods: initial resuscitation/stabilization, the first 72 hours and days 4 to 7, with each period subdivided in 8 to 12 care themes. Revisions and practice changes were implemented to improve skin integrity, admission temperature, timing of initiation of feeds, reliability of transcutaneous CO2 monitoring and ventilation. Of 127 infants <26 weeks admitted between implementation and end of 2019, 78 survived to discharge (61%).

CONCLUSION

It will be important to determine, with ongoing auditing and further evaluation, whether our care bundle led to improvements of short- and long-term outcomes in this population. Our experience may be useful to others caring for extremely low gestational age infants.

Precision Medicine in Neonates: A Tailored Approach to Neonatal Brain Injury

Despite advances in neonatal care to prevent neonatal brain injury and neurodevelopmental impairment, predicting long-term outcome in neonates at risk for brain injury remains difficult. Early prognosis is currently based on cranial ultrasound (CUS), MRI, EEG, NIRS, and/or general movements assessed at specific ages, and predicting outcome in an individual (precision medicine) is not yet possible. New algorithms based on large databases and machine learning applied to clinical, neuromonitoring, and neuroimaging data and genetic analysis and assays measuring multiple biomarkers (omics) can fulfill the needs of modern neonatology. A synergy of all these techniques and the use of automatic quantitative analysis might give clinicians the possibility to provide patient-targeted decision-making for individualized diagnosis, therapy, and outcome prediction. This review will first focus on common neonatal neurological diseases, associated risk factors, and most common treatments. After that, we will discuss how precision medicine and machine learning (ML) approaches could change the future of prediction and prognosis in this field.

Evaluating the Effect of a Neonatal Care Bundle for the Prevention of Intraventricular Hemorrhage in Preterm Infants

Germinal matrix intraventricular hemorrhage (IVH) remains a severe and common complication in preterm infants. A neonatal care bundle (NCB) was implemented as an in-house guideline at a tertiary neonatal intensive care unit to reduce the incidence of IVH in preterm infants. The NCB was applied either to preterm infants <1250 g birth weight or <30 weeks gestational age or both, and standardized patient positioning, nursing care, and medical procedures within the first week of life. A retrospective cohort study was performed to investigate the effect of the NCB and other known risk factors on the occurrence and severity of IVH. Data from 229 preterm infants were analyzed. The rate of IVH was 26.2% before and 27.1% after implementing the NCB. The NCB was associated neither with reducing the overall rate of IVH (odds ratio (OR) 1.02; 95% confidence interval (CI) 0.57–1.84; p = 0.94) nor with severe IVH (OR 1.0; 95% CI 0.67–1.55; p = 0.92). After adjustment for group differences and other influencing factors, amnion infection syndrome and early intubation were associated with an increased risk for IVH. An NCB focusing on patient positioning, nursing care, and medical interventions had no impact on IVH in preterm infants. Known risk factors for IVH were confirmed.

Machine learning models for identifying preterm infants at risk of cerebral hemorrhage

Intracerebral hemorrhage in preterm infants is a major cause of brain damage and cerebral palsy. The pathogenesis of cerebral hemorrhage is multifactorial. Among the risk factors are impaired cerebral autoregulation, infections, and coagulation disorders. Machine learning methods allow the identification of combinations of clinical factors to best differentiate preterm infants with intra-cerebral bleeding and the development of models for patients at risk of cerebral hemorrhage. In the current study, a Random Forest approach is applied to develop such models for extremely and very preterm infants (23-30 weeks gestation) based on data collected from a cohort of 229 individuals. The constructed models exhibit good prediction accuracy and might be used in clinical practice to reduce the risk of cerebral bleeding in prematurity.

Umbilical Cord Milking: A Review

This is a review of umbilical cord milking, a controversial technique where the umbilical cord is squeezed several times before it is clamped an cut. While not physiological or natural for newborns, the question lies as to whether it is useful in certain circumstances, namely the depressed newborn. Here we review the literature and discuss why it could be considered as an alternative for the current practice of delayed cord clamping.

Singapore Neonatal Resuscitation Guidelines 2016

We present the revised Neonatal Resuscitation Guidelines for Singapore. The 2015 International Liaison Committee on Resuscitation Neonatal Task Force's consensus on science and treatment recommendations (2015), and guidelines from the American Heart Association and European Resuscitation Council were debated and discussed. The final recommendations of the National Resuscitation Council, Singapore, were derived after the task force had carefully reviewed the current available evidence in the literature and addressed their relevance to local clinical practice.