Clinical Decision Support for Improved Neonatal Care: The Development of a Machine Learning Model for the Prediction of Late-onset Sepsis and Necrotizing Enterocolitis.

OBJECTIVE: To develop an artificial intelligence-based software system for predicting late-onset sepsis (LOS) and necrotizing enterocolitis (NEC) in infants admitted to the neonatal intensive care unit (NICU). STUDY DESIGN: Single-center, retrospective cohort study, conducted in the NICU of the Antwerp University Hospital. Continuous monitoring data of 865 preterm infants born at <32 weeks gestational age, admitted to the NICU in the first week of life, were used to train an XGBoost machine learning (ML) algorithm for LOS and NEC prediction in a cross-validated setup. Afterward, the model's performance was assessed on an independent test set of 148 patients (internal validation). RESULTS: The ML model delivered hourly risk predictions with an overall sensitivity of 69% (142/206) for all LOS/NEC episodes and 81% (67/83) for severe LOS/NEC episodes. The model showed a median time gain of ≤10 hours (IQR, 3.1-21.0 hours), compared with historical clinical diagnosis. On the complete retrospective dataset, the ML model made 721 069 predictions, of which 9805 (1.3%) depicted a LOS/NEC probability of ≥0.15, resulting in a total alarm rate of <1 patient alarm-day per week. The model reached a similar performance on the internal validation set. CONCLUSIONS: Artificial intelligence technology can assist clinicians in the early detection of LOS and NEC in the NICU, which potentially can result in clinical and socioeconomic benefits. Additional studies are required to quantify further the effect of combining artificial and human intelligence on patient outcomes in the NICU.

Expectant Management or Early Ibuprofen for Patent Ductus Arteriosus.

BACKGROUND: Cyclooxygenase inhibitors are commonly used in infants with patent ductus arteriosus (PDA), but the benefit of these drugs is uncertain. METHODS: In this multicenter, noninferiority trial, we randomly assigned infants with echocardiographically confirmed PDA (diameter, >1.5 mm, with left-to-right shunting) who were extremely preterm (<28 weeks' gestational age) to receive either expectant management or early ibuprofen treatment. The composite primary outcome included necrotizing enterocolitis (Bell's stage IIa or higher), moderate to severe bronchopulmonary dysplasia, or death at 36 weeks' postmenstrual age. The noninferiority of expectant management as compared with early ibuprofen treatment was defined as an absolute risk difference with an upper boundary of the one-sided 95% confidence interval of less than 10 percentage points. RESULTS: A total of 273 infants underwent randomization. The median gestational age was 26 weeks, and the median birth weight was 845 g. A primary-outcome event occurred in 63 of 136 infants (46.3%) in the expectant-management group and in 87 of 137 (63.5%) in the early-ibuprofen group (absolute risk difference, -17.2 percentage points; upper boundary of the one-sided 95% confidence interval [CI], -7.4; P<0.001 for noninferiority). Necrotizing enterocolitis occurred in 24 of 136 infants (17.6%) in the expectant-management group and in 21 of 137 (15.3%) in the early-ibuprofen group (absolute risk difference, 2.3 percentage points; two-sided 95% CI, -6.5 to 11.1); bronchopulmonary dysplasia occurred in 39 of 117 infants (33.3%) and in 57 of 112 (50.9%), respectively (absolute risk difference, -17.6 percentage points; two-sided 95% CI, -30.2 to -5.0). Death occurred in 19 of 136 infants (14.0%) and in 25 of 137 (18.2%), respectively (absolute risk difference, -4.3 percentage points; two-sided 95% CI, -13.0 to 4.4). Rates of other adverse outcomes were similar in the two groups. CONCLUSIONS: Expectant management for PDA in extremely premature infants was noninferior to early ibuprofen treatment with respect to necrotizing enterocolitis, bronchopulmonary dysplasia, or death at 36 weeks' postmenstrual age. (Funded by the Netherlands Organization for Health Research and Development and the Belgian Health Care Knowledge Center; BeNeDuctus ClinicalTrials.gov number, NCT02884219; EudraCT number, 2017-001376-28.).

Application and potential of artificial intelligence in neonatal medicine.

Neonatal care is becoming increasingly complex with large amounts of rich, routinely recorded physiological, diagnostic and outcome data. Artificial intelligence (AI) has the potential to harness this vast quantity and range of information and become a powerful tool to support clinical decision making, personalised care, precise prognostics, and enhance patient safety. Current AI approaches in neonatal medicine include tools for disease prediction and risk stratification, neurological diagnostic support and novel image recognition technologies. Key to the integration of AI in neonatal medicine is the understanding of its limitations and a standardised critical appraisal of AI tools. Barriers and challenges to this include the quality of datasets used, performance assessment, and appropriate external validation and clinical impact studies. Improving digital literacy amongst healthcare professionals and cross-disciplinary collaborations are needed to harness the full potential of AI to help take the next significant steps in improving neonatal outcomes for high-risk infants.

Association between maternal cervicovaginal swab positivity for Ureaplasma spp. or other microorganisms and neonatal respiratory outcome and mortality.

OBJECTIVE: We investigated the association between maternal cervicovaginal cultures, its antibiotic treatment, and neonatal outcome. STUDY DESIGN: This retrospective cohort study enrolled 480 neonates born prior to 32 weeks' gestation. They were divided into groups according to maternal cervicovaginal culture results. Multivariate logistic regression analysis was used to predict neonatal outcome based on maternal culture results, adjusted for perinatal risk factors and neonatal morbidities. RESULT: Maternal cervicovaginal Ureaplasma colonization was independently associated with bronchopulmonary dysplasia at 36 weeks (BPD) (OR 8.34; 95% CI 1.21-57.45). In neonates with and without maternal cervicovaginal Ureaplasma colonization BPD occurred in 12.3% and 3.8%, respectively. Maternal colonization with other microorganisms was associated with a higher neonatal mortality (p = 0.002), lower gestational age (p = 0.026), and birth weight (p = 0.036). CONCLUSIONS: This study underscores the role of the maternal cervicovaginal microbiome as a predictor of neonatal outcome. Cervicovaginal Ureaplasma colonization seems not to be an innocent bystander in the multifactorial etiology of BPD.

Machine Learning to Support Hemodynamic Intervention in the Neonatal Intensive Care Unit.

Hemodynamic support in neonatal intensive care is directed at maintaining cardiovascular wellbeing. At present, monitoring of vital signs plays an essential role in augmenting care in a reactive manner. By applying machine learning techniques, a model can be trained to learn patterns in time series data, allowing the detection of adverse outcomes before they become clinically apparent. In this review we provide an overview of the different machine learning techniques that have been used to develop models in hemodynamic care for newborn infants. We focus on their potential benefits, research pitfalls, and challenges related to their implementation in clinical care.

Epilepsy in children with Down syndrome: not so benign as generally accepted.

Down syndrome (DS) is one of the most common causes of mental retardation in children. Many children with DS suffer from neurologic problems, including seizures. Epileptic spasms (ES) are the most frequently reported seizure type. As in the general epilepsy population, ES are rather difficult to control with anti-epileptic drugs. Different treatment regimens have been proposed in the literature, most of them containing vigabatrin or steroids. We present 12 children with DS, who were seen and treated at the Antwerp University Hospital because of seizures. Eight of them presented with ES. Different treatment regimens were used, with varying outcome. This article summarizes our experience with epilepsy in children with DS, describing the different treatment options that were used. We found a poor outcome in these children, compared to most previous reports. Although steroids play an important role in the treatment of ES worldwide, we found a low success rate (8.3 %) of these drugs.