The large environmental impact (EI) of healthcare is of growing concern, especially given the increasing negative health effects of environmental deterioration.1,2 Paediatric intensive care units (PICU) are major contributors to this EI, partly due to the use of consumables and electricity.3–5 Environmental hotspots in clinical pathways (CP) must be identified to guide practical and effective interventions to lower the EI.5,6 Life cycle assessment (LCA) is currently the golden standard for EI assessment. However, LCA is time-consuming and highly complex. Thus, execution of LCAs on a large scale to analyse full CPs is not feasible in healthcare settings. Furthermore, the required data are often not available. Financial costs may be used as a proxy (spend-based LCA), but their representativeness is questionable.6,7 Furthermore, the CPs of patients in the PICU differ highly and cannot be represented by a single standard CP. Therefore, this research aimed to develop a process-based framework for environmental hotspot identification that requires less time and expertise compared to LCA and accounts for differences between patients, using a case study of six PICU post-cardiac surgery patients. Modules were designed as building blocks to represent medical events in the CP with the flexibility to deal with interpatient variation. Associated consumable and electricity use were allocated to each module based on medical protocols and discussions with clinical staff. From this iterative process, a set of allocation rules was established. The material composition of each product was analysed and recorded in a database. Per module, the median frequency of occurrence was calculated in the patient cohort data. The carbon emissions (kg CO2-equivalent) associated with each module per occurrence were calculated based on impact factors from an open-source database.8 The total of each module was defined as the EI per module occurrence (in kg material and in kg CO2-equivalent) multiplied by the median module frequency. This research was a first step towards an accurate and flexible approach to environmental hotspot identification within CPs related to consumable weight and bedside electricity consumption. The modules and the data analysis algorithm can be reused and expanded for other CPs, saving time and effort in future analyses. The allocation rules ensure standardised allocation of consumables and electricity across the current modules and when new modules are added. Challenges in this approach lie in the availability of product information from manufacturers and reliable, open-source impact factors, especially for pharmaceuticals.

Study objectives: Conventional sleep scoring is based on the scoring criteria of the American Association of Sleep Medicine (AASM) but may not be suited to describe sleep in critically ill children admitted to the Pediatric Intensive Care Unit (PICU). In this study, an anomaly detection model using Gaussian Models trained on sleep stages in data from non-critically ill children is developed to assess if polysomnography(PSG)-derived electroencephalography (EEG) data from critically ill children can be categorized into sleep stages based on these AASM scoring criteria.
Methods: A retrospective study at Erasmus MC Sophia Children’s Hospital, using PSG recordings obtained in non-critically ill children between 2017 and 2021 and in critically ill children between 2020 and 2022.
Gaussian Models were individually trained for each sleep stage using data from non-critically ill children. Anomaly detection was carried out by computing the Mahalanobis Distances and assigning datapoints to specific sleep stages or categorizing them as anomalous. Errors were quantified by calculating the ratio of anomalous epochs to the total number of epochs. The trained Gaussian Models were applied to distinct sleep stages in the data from non-critically ill children. Subsequently, the models were applied to data from critically ill children to determine the categorization of their epochs. This was also analyzed over time and involved comparisons related to medication, mechanical ventilation, and the severity of illness assessed by the PELOD-2 score.
Results: In non-critically ill children the models obtained validation errors aligning with the margin error of the training set. The models could not fully differentiate the distinct sleep stages. In critically ill children, the majority of epochs were classified into multiple sleep stages. High error rates were evident for sleep stages N1, R, and N. Some patients exhibited elevated error rates specifically for sleep stage N1. REM sleep was reduced, consistent with findings from previous studies. In contrast, N3 sleep did not show a reduction. When compared to the sleep stage labels assigned by neurophysiologists, the model classified epochs into multiple sleep stages, while neurophysiologists frequently used the label N. A higher PELOD-2 score did not consistently correlate with an increased occurrence of anomalous classifications in the epochs of these patients to those with lower PELOD-2 scores.
Discussion: Overlap of sleep stages was observed in non-critically ill children. Epochs from critically ill children were classified into multiple sleep stages without clear associations in time or severity of illness. Building upon the established anomaly detection framework is recommended by employing more advanced anomaly detection methods using an informative feature selection. This study marks an initial step, indicating that applying the AASM.

Introduction: Critically ill children admitted to the Paediatric Intensive Care Unit (PICU) have a high risk of disruption of their normal sleep rhythm, which is associated with disturbances in physiology and negative effects on psychological and cognitive functioning. There is a need for real-time, automatic sleep monitoring to minimise disruptions in sleep patterns. The main objective of this thesis was to develop a machine learning model that can classify sleep based on vital signs in critically ill children. In addition, methods were investigated to optimise the decision criteria in multi-class problems.
Methods: Three machine learning algorithms, logistic regression, random forest and extreme gradient boosting (XGBoost), were developed based on both the combination of features extracted from electrocardiogram (ECG) signal and pulse transit time (PTT) and on ECG features alone. To gain insight into the number of sleep stages that could be distinguished, the models were developed for two-class, three-class, four-class and five-class staging. The models were developed, trained and evaluated on a diagnostic dataset (n = 90) containing polysomnography (PSG) measurements of non-critically ill children. During model development, the decision criteria for the different classes were jointly optimised. To evaluate whether the models were generalisable to the PICU population, external validation was performed on a set of 8 of PICU patients.
Results: For each number of sleep stages, the three models performed similarly. However, there was an increase in performance with a decrease in the number of sleep stages with balanced accuracies varying between 0.70 and 0.72 in two-class staging and between 0.41 and 0.42 in five-class staging. External validation on the PICU dataset showed a markedly worse performance for all three models with balanced accuracies varying between 0.56 and 0.62 in two-class staging and between 0.22 and 0.23 in five-class staging.
Conclusion: Machine learning models for sleep classification in children based on vital signs have been developed and show promising results. Nonetheless, the developed models are not generalisable to the PICU population. Further research is recommended with a focus on improving the models such that they can be applied in the PICU. Combining information extracted from vital signs with EEG signal and developing the models directly on PICU data should be considered to improve the performance and thereby contribute to personalised care and minimising sleep disturbances.

Introduction Traumatic brain injury (TBI) is a leading cause of childhood morbidity, disability and mortality worldwide. In the treatment of TBI, neuromonitoring is essential to prevent secondary neurological damage. However, the use of neuromonitoring and specifically therapeutic targets is currently limited evidence-based.
Objective The primary objective is to implement suitable neuromonitoring parameters into a novel bedside dashboard for paediatric TBI patients. In addition, basic insights into neuromonitoring parameters between different outcome groups are provided.
Methods Intracranial pressure (ICP), cumulative ICP (ICP_cum), cerebral perfusion pressure (CPP), pressure reactivity index (PRx), wavelet transform PRx, optimal CPP (CPP_opt), partial pressure of arterial oxygen (P_aO₂), partial pressure of arterial carbon dioxide (P_aCO₂), end tidal CO2 (etCO₂), temperature and sodium levels were implemented into a dashboard. The parameters (except wPRx) were retrospectively analysed in TBI patients admitted to the paediatric intensive care unit (PICU) of the Erasmus MC- Sophia. Patients were divided into a good (GO) or bad (BO) outcome group. Per outcome group, the mean and 95 % confidence interval or the median and interquartile range of parameters were calculated.
Results A total of 35 patients (24 GO, 11 BO) were included. Significant differences between GO and BO were observed in median PRx (-0.06 [-0.13 – 0.01] vs. 0.17 [-0.05 – 0.39]), median percentage outside of the CPP_opt (curve) range (15 % [4 – 26] vs. 30 % [26 – 34]), ICP_cum distributions (12.0 mmHg [8.0 – 16.0] vs. 12.0 mmHg [8.5 – 15.5]) and mean sodium levels (146.4 mmol/L (144.5 – 148.3) vs. 151.2 mmol/L (148.6 – 153.8)).
Conclusion In a retrospective analysis, PRx, CPP_opt (curve) and ICP_cum showed potential to improve prognostication for paediatric TBI and to determine therapeutic targets. Based on our results, we recommend implementing MAP, ICP, ICP_cum, CPP alongside CPP_opt (curve), and sodium levels in a future neuromonitoring dashboard.