Automated Calculator for the Pediatric Sequential Organ Failure Assessment Score: Development and External Validation in a Single-Center 7-Year Cohort, 2015-2021.

OBJECTIVES: The pediatric Sequential Organ Failure Assessment (pSOFA) score summarizes severity of organ dysfunction and can be used to predict in-hospital mortality. Manual calculation of the pSOFA score is time-consuming and prone to human error. An automated method that is open-source, flexible, and scalable for calculating the pSOFA score directly from electronic health record data is desirable. DESIGN: Single-center, retrospective cohort study. SETTING: Quaternary 40-bed PICU. PATIENTS: All patients admitted to the PICU between 2015 and 2021 with ICU stay of at least 24 hours. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We used 77 records to evaluate the automated score. The automated algorithm had an overall accuracy of 97%. The algorithm calculated the respiratory component of two cases incorrectly. An expert human annotator had an initial accuracy of 75% at the patient level and 95% at the component level. An untrained human annotator with general clinical research experience had an overall accuracy of 16% and component-wise accuracy of 67%. Weighted kappa for agreement between the automated method and the expert annotator's initial score was 0.92 (95% CI, 0.88-0.95), and between the untrained human annotator and the automated score was 0.50 (95% CI, 0.36-0.61). Data from 9146 patients (in-hospital mortality 3.6%) were included to validate externally the discriminability of the automated pSOFA score. The admission-day pSOFA score had an area under the receiver operating characteristic curve of 0.79 (95% CI, 0.77-0.82). CONCLUSIONS: The developed automated algorithm calculates pSOFA score with high accuracy and is more accurate than a trained expert rater and nontrained data abstracter. pSOFA's performance for predicting in-hospital mortality was lower in our cohort than it was for the originally derived score.

Dynamic Prediction of Mortality Using Longitudinally Measured Pediatric Sequential Organ Failure Assessment Scores: A Joint Modeling Approach.

OBJECTIVES: The pediatric Sequential Organ Failure Assessment (pSOFA) score was designed to track illness severity and predict mortality in critically ill children. Most commonly, pSOFA at a point in time is used to assess a static patient condition. However, this approach has a significant drawback because it fails to consider any changes in a patients' condition during their PICU stay and, especially, their response to initial critical care treatment. We aimed to evaluate the performance of longitudinal pSOFA scores for predicting mortality. DESIGN: Single-center, retrospective cohort study. SETTING: Quaternary 40-bed PICU. PATIENTS: All patients admitted to the PICU between 2015 and 2021 with at least 24 hours of ICU stay. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We calculated daily pSOFA scores up to 30 days, or until death or discharge from the PICU, if earlier. We used the joint longitudinal and time-to-event data model for the dynamic prediction of 30-day in-hospital mortality. The dataset, which included 9146 patients with a 30-day in-hospital mortality of 2.6%, was divided randomly into training (75%) and validation (25%) subsets, and subjected to 40 repeated stratified cross-validations. We used dynamic area under the curve (AUC) to evaluate the discriminative performance of the model. Compared with the admission-day pSOFA score, AUC for predicting mortality between days 5 and 30 was improved on average by 6.4% (95% CI, 6.3-6.6%) using longitudinal pSOFA scores from the first 3 days and 9.2% (95% CI, 9.0-9.5%) using scores from the first 5 days. CONCLUSIONS: Compared with admission-day pSOFA score, longitudinal pSOFA scores improved the accuracy of mortality prediction in PICU patients at a single center. The pSOFA score has the potential to be used dynamically for the evaluation of patient conditions.

Changes in global nutrition practices in critically ill children and the impact of emerging evidence: a secondary analysis of the Pediatric International Nutrition Studies (PINS), 2009-2018.

BACKGROUND: The timeline of the three Pediatric International Nutrition Studies (PINS) coincided with the publication of 2 major guidelines for the timing of parenteral nutrition (PN) and recommended energy and protein delivery dose. OBJECTIVE: The study's main objective was to describe changes in the nutrition delivery practice recorded in PINS 1 and 2 (conducted in 2009 and 2011, pre-exposure epoch) versus PINS 3 (conducted in 2018,post-exposure epoch), in relation to the published practice guidelines. DESIGN: This study is a secondary analysis of data from a multi-center prospective cohort study. PARTICIPANTS: /setting. Data from 3650 participants, aged 1 month to 18 years, admitted to 100 unique hospitals that participated in three PINS was used for this study. MAIN OUTCOME MEASURES: The time in days from PICU admission to the initiation of PN and enteral nutrition (EN) delivery were the primary outcomes. Prescribed energy and protein goals were the secondary outcomes. STATISTICAL ANALYSES PERFORMED: A frailty model with a random intercept per hospital with stratified baseline hazard function by region for the primary outcomes and a mixed-effects negative binomial regression with random intercept per hospital for the secondary outcomes. RESULTS: The proportion of patients receiving EN (88.3% vs. 80.6%, p-value<0.001) was higher, and those receiving PN (20.6% vs. 28.8%, p-value<0.001) was lower in the PINS3 cohort compared to PINS1-2. In the PINS3 cohort, the odds of initiating PN during the 1st 10 days of PICU admission were lower, compared to the PINS1-2 cohort (HR=0.8, CI=[0.67-0.95], p-value=0.013); and prescribed energy goal was lower compared to the PINS1-2 cohort (IRR=0.918, CI=[0.874-0.965], p=0.001). CONCLUSIONS: The likelihood of initiation of PN delivery significantly decreased in the first ten days post-admission in the PINS3 cohort compared to PINS1-2. Energy goal prescription in mechanically ventilated children significantly decreased in the post-guidelines epoch compared to the pre-guidelines epoch.

Comparing the Quality of Domain-Specific Versus General Language Models for Artificial Intelligence-Generated Differential Diagnoses in PICU Patients.

OBJECTIVES: Generative language models (LMs) are being evaluated in a variety of tasks in healthcare, but pediatric critical care studies are scant. Our objective was to evaluate the utility of generative LMs in the pediatric critical care setting and to determine whether domain-adapted LMs can outperform much larger general-domain LMs in generating a differential diagnosis from the admission notes of PICU patients. DESIGN: Single-center retrospective cohort study. SETTING: Quaternary 40-bed PICU. PATIENTS: Notes from all patients admitted to the PICU between January 2012 and April 2023 were used for model development. One hundred thirty randomly selected admission notes were used for evaluation. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Five experts in critical care used a 5-point Likert scale to independently evaluate the overall quality of differential diagnoses: 1) written by the clinician in the original notes, 2) generated by two general LMs (BioGPT-Large and LLaMa-65B), and 3) generated by two fine-tuned models (fine-tuned BioGPT-Large and fine-tuned LLaMa-7B). Differences among differential diagnoses were compared using mixed methods regression models. We used 1,916,538 notes from 32,454 unique patients for model development and validation. The mean quality scores of the differential diagnoses generated by the clinicians and fine-tuned LLaMa-7B, the best-performing LM, were 3.43 and 2.88, respectively (absolute difference 0.54 units [95% CI, 0.37-0.72], p < 0.001). Fine-tuned LLaMa-7B performed better than LLaMa-65B (absolute difference 0.23 unit [95% CI, 0.06-0.41], p = 0.009) and BioGPT-Large (absolute difference 0.86 unit [95% CI, 0.69-1.0], p < 0.001). The differential diagnosis generated by clinicians and fine-tuned LLaMa-7B were ranked as the highest quality in 144 (55%) and 74 cases (29%), respectively. CONCLUSIONS: A smaller LM fine-tuned using notes of PICU patients outperformed much larger models trained on general-domain data. Currently, LMs remain inferior but may serve as an adjunct to human clinicians in real-world tasks using real-world data.

Validation and Extension of the Association Between Potentially Excess Oxygen Exposure and Death in Mechanically Ventilated Children.

OBJECTIVES: "Cumulative excess oxygen exposure" (CEOE)-previously defined as the mean hourly administered Fio2 above 0.21 when the corresponding hourly Spo2 was 95% or above-was previously shown to be associated with mortality. The objective of this study was to examine the relationship among Fio2, Spo2, and mortality in an independent cohort of mechanically ventilated children. DESIGN: Retrospective cross-sectional study. SETTING: Quaternary-care PICU. PATIENTS: All patients admitted to the PICU between 2012 and 2021 and mechanically ventilated via endotracheal tube for at least 24 hours. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Among 3,354 patients, 260 (8%) died. Higher CEOE quartile was associated with increased mortality (p = 0.001). The highest CEOE quartile had an 87% increased risk of mortality (95% CI, 7-236) compared with the first CEOE quartile. The hazard ratio for extended CEOE exposure, which included mechanical ventilation data from throughout the patients' mechanical ventilation time rather than only from the first 24 hours of mechanical ventilation, was 1.03 (95% CI, 1.02-1.03). CONCLUSIONS: Potentially excess oxygen exposure in patients whose oxygen saturation was at least 95% was associated with increased mortality.

Brain effective connectome based on fMRI and DTI data: Bayesian causal learning and assessment.

Neuroscientific studies aim to find an accurate and reliable brain Effective Connectome (EC). Although current EC discovery methods have contributed to our understanding of brain organization, their performances are severely constrained by the short sample size and poor temporal resolution of fMRI data, and high dimensionality of the brain connectome. By leveraging the DTI data as prior knowledge, we introduce two Bayesian causal discovery frameworks -the Bayesian GOLEM (BGOLEM) and Bayesian FGES (BFGES) methods- that offer significantly more accurate and reliable ECs and address the shortcomings of the existing causal discovery methods in discovering ECs based on only fMRI data. Moreover, to numerically assess the improvement in the accuracy of ECs with our method on empirical data, we introduce the Pseudo False Discovery Rate (PFDR) as a new computational accuracy metric for causal discovery in the brain. Through a series of simulation studies on synthetic and hybrid data (combining DTI from the Human Connectome Project (HCP) subjects and synthetic fMRI), we demonstrate the effectiveness of our proposed methods and the reliability of the introduced metric in discovering ECs. By employing the PFDR metric, we show that our Bayesian methods lead to significantly more accurate results compared to the traditional methods when applied to the Human Connectome Project (HCP) data. Additionally, we measure the reproducibility of discovered ECs using the Rogers-Tanimoto index for test-retest data and show that our Bayesian methods provide significantly more reliable ECs than traditional methods. Overall, our study's numerical and visual results highlight the potential for these frameworks to significantly advance our understanding of brain functionality.

Heart Rate Change as a Potential Digital Biomarker of Brain Death in Critically Ill Children With Acute Catastrophic Brain Injury.

Bedside measurement of heart rate (HR) change (HRC) may provide an objective physiologic marker for when brain death (BD) may have occurred, and BD testing is indicated in children. OBJECTIVES: To determine whether HRC, calculated using numeric HR measurements sampled every 5 seconds, can identify patients with BD among patients with catastrophic brain injury (CBI). DESIGN SETTING AND PARTICIPANTS: Single-center, retrospective study (2008-2020) of critically ill children with acute CBI. Patients with CBI had a neurocritical care consultation, were admitted to an ICU, had acute neurologic injury on presentation or during hospitalization based on clinical and/or imaging findings, and died or survived with Glasgow Coma Scale (GCS) less than 13 at hospital discharge. Patients meeting BD criteria (BD group) were compared with those with cardiopulmonary death (CD group) or those who survived to discharge. MAIN OUTCOMES AND MEASURES: HRC was calculated as the interquartile range of HR divided by median HR using 5-minute windows with 50% overlap for up to 5 days before death or end of recording. HRC was compared among the BD, CD, and survivor groups. RESULTS: Of 96 patients with CBI (69% male, median age 4 years), 28 died (8 BD, 20 CD) and 20 survived (median GCS 9 at discharge). Within 24 hours before death, HRC was lower in BD compared with CD patients or survivors (0.01 vs 0.03 vs 0.04, p = 0.001). In BD patients, HRC decreased at least 1 day before death. HRC discriminated BD from CD patients and survivors with 90% sensitivity, 70% specificity, 44% positive predictive value, 96% negative predictive value (area under the receiver operating characteristic curve 0.88, 95% CI, 0.80-0.93). CONCLUSIONS AND RELEVANCE: HRC is a novel digital biomarker that, with further validation, may be useful as a classifier for BD in the overall course of patients with CBI.

Long-Term Functional Outcomes at 1-Year After Hospital Discharge in Critically Ill Neonates With Congenital Diaphragmatic Hernia.

OBJECTIVES: Congenital diaphragmatic hernia (CDH) is a birth defect associated with long-term morbidity. Our objective was to examine longitudinal change in Functional Status Scale (FSS) after hospital discharge in CDH survivors. DESIGN: Single-center retrospective cohort study. SETTING: Center for comprehensive CDH management at a quaternary, free-standing children's hospital. PATIENTS: Infants with Bochdalek CDH were admitted to the ICU between January 2009 and December 2019 and survived until hospital discharge. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: One hundred forty-two infants (58% male, mean birth weight 3.08 kg, 80% left-sided defects) met inclusion criteria. Relevant clinical data were extracted from the medical record to calculate FSS (primary outcome) at hospital discharge and three subsequent outpatient follow-up time points. The median (interquartile range [IQR]) FSS score at hospital discharge was 8.0 (7.0-9.0); 39 patients (27.5%) had at least moderate impairment (FSS ≥ 9). Median (IQR) FSS at 0- to 6-month ( n = 141), 6- to 12-month ( n = 141), and over 12-month ( n = 140) follow-up visits were 7.0 (7.0-8.0), 7.0 (6.0-8.0), and 6.0 (6.0-7.0), respectively. Twenty-one patients (15%) had at least moderate impairment at over 12-month follow-up; median composite FSS scores in the over 12-month time point decreased by 2.0 points from hospital discharge. Median feeding domain scores improved by 1.0 (1.0-2.0), whereas other domain scores remained without impairment. Multivariable analysis demonstrated right-sided, C- or D-size defects, extracorporeal membrane oxygenation use, cardiopulmonary resuscitation, and chromosomal anomalies were associated with impairment. CONCLUSIONS: The majority of CDH survivors at our center had mild functional status impairment (FSS ≤ 8) at discharge and 1-year follow-up; however, nearly 15% of patients had moderate impairment during this time period. The feeding domain had the highest level of functional impairment. We observed unchanged or improving functional status longitudinally over 1-year follow-up after hospital discharge. Longitudinal outcomes will guide interdisciplinary management strategies in CDH survivors.

Sporadic and Periodic Interictal Discharges in Critically Ill Children: Seizure Associations and Time to Seizure Identification.

PURPOSE: We evaluated interictal discharges (IEDs) as a biomarker for the time to development of electrographic seizures (ES). METHODS: Prospective observational study of 254 critically ill children who underwent continuous electroencephalography (cEEG) monitoring. We excluded neonates and patients with known epilepsy or the sole cEEG indication to characterize events. Interictal discharges included sporadic epileptiform discharges and periodic and rhythmic patterns. Sporadic epileptiform discharges were categorized as low frequency (rare [<1/hour] and occasional [≥1/hour but <1/minute]) and high frequency (frequent, [≥1/minute] and abundant [≥1/10 seconds]). Time variables included time from cEEG start to first IED and time between first IED and ES. RESULTS: Interictal discharges were present in 33% (83/254) of patients. We identified ES in 20% (50/254), and 86% (43/50) had IEDs. High-frequency sporadic epileptiform discharges (odds ratio [OR], 35; 95% confidence interval [CI], 14.5-88; P < 0.0001) and lateralized periodic discharges (OR, 27; 95% CI, 7.3-100; P < 0.0001) were associated with ES. Mildly abnormal EEG background without IEDs or background asymmetry was associated with the absence of seizures (OR, 0.1; 95% CI, 0.04-0.3; P < 0.0001). Time from cEEG start to first IED was 36 minutes (interquartile range, 3-131 minutes), and time between first IED and ES was 9.6 minutes (interquartile range, 0.6-165 minutes). CONCLUSIONS: Interictal discharges are associated with ES and are identified in the first 3 hours of cEEG. High-frequency sporadic epileptiform discharges and periodic patterns have the highest risk of ES. Our findings define a window of high seizure risk after the identification of IEDs in which to allocate resources to improve seizure identification and subsequent treatment.

Impact of implementing an evidence-based definition of enteral nutrition intolerance on nutrition delivery: A prospective, cross-sectional cohort study.

BACKGROUND: Enteral nutrition (EN) interruptions because of EN intolerance impede nutrient delivery. We aimed to examine whether revising the EN intolerance definition of an algorithm would decrease EN interruptions and improve nutrient delivery in critically ill children. METHODS: We performed a cross-sectional cohort study including patients who were admitted to our intensive care unit (ICU) for >24 h and received EN. The EN intolerance definition in our nutrition algorithm was modified to include two symptoms of EN intolerance. We compared time to 60% EN adequacy (EN delivered/EN prescribed x 100) and EN interruptions before and after this intervention. RESULTS: We included 150 eligible patients, 78 and 72 patients in the preimplementation and postimplementation cohorts, respectively. There were no significant differences in demographics and clinical characteristics. The preimplementation and postimplementation cohorts achieved 60% EN adequacy 4 (2-5) days and 3 (2-5) days after ICU admission, respectively (P = 0.59). The preimplementation cohort had a median of 1 (1-2) interruption per patient and the postimplementation cohort 2 (1-3; P = 0.08). The frequency of interruptions because of EN intolerance within the first 8 days of ICU admission was 17 in the preimplementation and 10 in the postimplementation cohorts. CONCLUSION: Modifying the EN intolerance definition of a nutrition algorithm did not change the time to 60% EN adequacy or total number of EN interruptions in critically ill children. EN intolerance and interruptions continue to limit nutrient delivery. Research on the best definition for EN intolerance and its effect on nutrition outcomes is needed.

Pharmacokinetic Modeling of Optimized Midazolam and Pentobarbital Dosing Used in Treatment Protocols of Refractory Status Epilepticus.

OBJECTIVES: To model bolus dosing, infusion rate, and weaning rate on theoretical serum concentration of midazolam and pentobarbital used in the treatment of refractory status epilepticus (RSE). DESIGN: One- and two-compartment in silico pharmacokinetic models of midazolam and pentobarbital. SETTING: Not applicable. SUBJECTS: Not applicable. INTERVENTIONS: We compared the model variables used in midazolam and pentobarbital protocols for standard RSE. MEASUREMENTS AND MAIN RESULTS: Standard RSE treatment protocols result in steady-state serum concentrations that are 6.2-9.0-fold higher for the one-compartment model and 2.3-4.7-fold higher for the two-compartment model. In the model, not including bolus doses delays the achievement of serum steady-state concentration by 0.5 and 2.7 hours for midazolam and pentobarbital, respectively. Abrupt discontinuation of these medications reduces modeled medication exposure by 1.1 and 6.4 hours, respectively. CONCLUSIONS: Our in silico pharmacokinetic modeling of standard midazolam and pentobarbital dosing protocols for RSE suggests potential variables to optimize in future clinical studies.

Factors Associated With Mechanical Ventilation Duration in Pediatric Burn Patients in a Regional Burn Center in the United States.

OBJECTIVES: Among burned children who arrive at a burn center and require invasive mechanical ventilation (IMV), some may have prolonged IMV needs. This has implications for patient-centered outcomes as well as triage and resource allocation decisions. Our objective was to identify factors associated with the duration of mechanical ventilation in pediatric patients with acute burn injury in this setting. DESIGN: Single-center, retrospective cohort study. SETTING: Registry data from a regional, pediatric burn center in the United States. PATIENTS: Children less than or equal to 18 years old admitted with acute burn injury who received IMV between January 2005 and December 2020. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Ventilator days were defined as any full or partial day having received IMV via an endotracheal tube or tracheostomy, not inclusive of time spent ventilated for procedures. Of 5,766 admissions for acute burn care, 4.3% ( n = 249) required IMV with a median duration of 10 days. A multivariable model for freedom from mechanical ventilation showed that the presence of inhalational injury (subhazard ratio [sHR], 0.62; 95% CI, 0.46-0.85) and burns to the head and neck region (sHR, 0.94; 95% CI, 0.90-0.98) were associated with increased risk of remaining mechanically ventilated at any time point. Older (sHR, 1.03; 95% CI, 1.01-1.04) and male children (sHR, 1.39; 95% CI, 1.05-1.84) were more likely to discontinue mechanical ventilation. A majority of children (94.8%) survived to hospital discharge. CONCLUSIONS: The presence of inhalational injury and burns to the head and neck region were associated with a longer duration of mechanical ventilation. Older age and male gender were associated with a shorter duration of mechanical ventilation. These factors should help clinicians better estimate a burned child's expected trajectory and resource-intensive needs upon arrival to a burn center.

Optic Nerve Sheath Viscoelastic Properties: Re-Examination of Biomechanical Behavior and Clinical Implications.

BACKGROUND: Meta-analyses show a variable relationship between optic nerve sheath diameter (ONSD) and the presence of raised intracranial pressure (ICP). Because optic nerve sheath (ONS) tissue can be deformed, it is possible that ONSD reflects not only the current ICP but also prior deforming biomechanical exposures. In this post hoc analysis of two published data sets, we characterize ONS Young's modulus (E, mechanical stress per unit of strain) and calculate threshold pressure for plastic deformation. METHODS: The authors of two previously published articles contributed primary data for these unique post hoc analyses. Human cadaveric ex vivo measurements of ONSD (n = 10) and luminal distending pressure (range 5 to 65 mm Hg) were used to calculate E and the threshold pressure for plastic deformation. Clinical in vivo measurements of ONSD and ICP during endotracheal tube suction from patients with traumatic brain injury (n = 15) were used to validate the ex vivo cadaveric findings. RESULTS: Ex vivo ONS estimate of E was 140 ± 1.3 mm Hg (mean ± standard error), with evidence of plastic deformation occurring with distending pressure at 45 mm Hg. Similar E (71 ± 10 mm Hg) was estimated in vivo with an average ICP of 34 ± 2 mm Hg. CONCLUSIONS: Ex vivo, ONS plastic deformation occurs at levels of pressure commonly seen in patients with raised ICP, leading to distortion of the ICP-ONSD relationship. This evidence of plastic deformation may illustrate why meta-analyses fail to identify a single threshold in ONSD associated with the presence of raised ICP. Future studies characterizing time-dependent viscous characteristics of the ONS will help determine the time course of ONS tissue biomechanical behavior.

Artificial Neural Network Algorithms to Predict Resting Energy Expenditure in Critically Ill Children.

INTRODUCTION: Accurate assessment of resting energy expenditure (REE) can guide optimal nutritional prescription in critically ill children. Indirect calorimetry (IC) is the gold standard for REE measurement, but its use is limited. Alternatively, REE estimates by predictive equations/formulae are often inaccurate. Recently, predicting REE with artificial neural networks (ANN) was found to be accurate in healthy children. We aimed to investigate the role of ANN in predicting REE in critically ill children and to compare the accuracy with common equations/formulae. STUDY METHODS: We enrolled 257 critically ill children. Nutritional status/vital signs/biochemical values were recorded. We used IC to measure REE. Commonly employed equations/formulae and the VCO2-based Mehta equation were estimated. ANN analysis to predict REE was conducted, employing the TWIST system. RESULTS: ANN considered demographic/anthropometric data to model REE. The predictive model was good (accuracy 75.6%; R2 = 0.71) but not better than Talbot tables for weight. After adding vital signs/biochemical values, the model became superior to all equations/formulae (accuracy 82.3%, R2 = 0.80) and comparable to the Mehta equation. Including IC-measured VCO2 increased the accuracy to 89.6%, superior to the Mehta equation. CONCLUSIONS: We described the accuracy of REE prediction using models that include demographic/anthropometric/clinical/metabolic variables. ANN may represent a reliable option for REE estimation, overcoming the inaccuracies of traditional predictive equations/formulae.

A Machine Learning Classifier Improves Mortality Prediction Compared With Pediatric Logistic Organ Dysfunction-2 Score: Model Development and Validation.

OBJECTIVES: To determine whether machine learning algorithms can better predict PICU mortality than the Pediatric Logistic Organ Dysfunction-2 score. DESIGN: Retrospective study. SETTING: Quaternary care medical-surgical PICU. PATIENTS: All patients admitted to the PICU from 2013 to 2019. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We investigated the performance of various machine learning algorithms using the same variables used to calculate the Pediatric Logistic Organ Dysfunction-2 score to predict PICU mortality. We used 10,194 patient records from 2013 to 2017 for training and 4,043 patient records from 2018 to 2019 as a holdout validation cohort. Mortality rate was 3.0% in the training cohort and 3.4% in the validation cohort. The best performing algorithm was a random forest model (area under the receiver operating characteristic curve, 0.867 [95% CI, 0.863-0.895]; area under the precision-recall curve, 0.327 [95% CI, 0.246-0.414]; F1, 0.396 [95% CI, 0.321-0.468]) and significantly outperformed the Pediatric Logistic Organ Dysfunction-2 score (area under the receiver operating characteristic curve, 0.761 [95% CI, 0.713-0.810]; area under the precision-recall curve (0.239 [95% CI, 0.165-0.316]; F1, 0.284 [95% CI, 0.209-0.360]), although this difference was reduced after retraining the Pediatric Logistic Organ Dysfunction-2 logistic regression model at the study institution. The random forest model also showed better calibration than the Pediatric Logistic Organ Dysfunction-2 score, and calibration of the random forest model remained superior to the retrained Pediatric Logistic Organ Dysfunction-2 model. CONCLUSIONS: A machine learning model achieved better performance than a logistic regression-based score for predicting ICU mortality. Better estimation of mortality risk can improve our ability to adjust for severity of illness in future studies, although external validation is required before this method can be widely deployed.

Implementation of an Analgesia-Sedation Protocol Is Associated With Reduction in Midazolam Usage in the PICU.

OBJECTIVES: Examine the association of a revised analgesia-sedation protocol with midazolam usage in the PICU. DESIGN: A single-center nonrandomized before-after study. SETTING: PICU at a quaternary pediatric hospital (Boston Children's Hospital, Boston, MA). PATIENTS: Children admitted to the PICU who were mechanically ventilated for greater than 24 hours. The preimplementation cohort included 190 eligible patients admitted between July 29, 2017, and February 28, 2018, and the postimplementation cohort included 144 patients admitted between July 29, 2019, and February 28, 2020. INTERVENTIONS: Implementation of a revised analgesia-sedation protocol. MEASUREMENTS AND MAIN RESULTS: Our primary outcome, total dose of IV midazolam administered in mechanically ventilated patients up to day 14 of ventilation, decreased by 72% (95% CI [61-80%]; p < 0.001) in the postimplementation cohort. Dexmedetomidine usage increased 230% (95% CI [145-344%]) in the postimplementation cohort. Opioid usage, our balancing metric, was not significantly different between the two cohorts. There were no significant differences in ventilator-free days, PICU length of stay, rate of unplanned extubations, failed extubations, cardiorespiratory arrest events, and 24-hour readmissions to the PICU. CONCLUSIONS: We successfully implemented an analgesia-sedation protocol that primarily uses dexmedetomidine and intermittent opioids, and it was associated with significant decrease in overall midazolam usage in mechanically ventilated patients in the PICU. The intervention was not associated with changes in opioid usage or prevalence of adverse events.

EEG features of brain injury during extracorporeal membrane oxygenation in children.

OBJECTIVE: To examine EEG features of major pathophysiology in children undergoing extracorporeal membrane oxygenation (ECMO). METHODS: This was a single-center, retrospective study of 201 pediatric patients on ECMO, using the first 24 hours of continuous EEG (cEEG) monitoring, collating background activity and electrographic seizures (ES) with imaging, ECMO type, and outcome. RESULTS: Severely abnormal cEEG background occurred in 12% (25/201), and was associated with death (sensitivity 0.23, specificity 0.97). ES occurred in 16% (33/201) within 3.2 (0.6-20.3) hours (median [interquartile range]) of cEEG commencement, and higher ES burden was associated with death. ES was always associated with ipsilateral injury (p = 0.006), but occurred in only one-third of cases with abnormal imaging. In 28 patients with isolated hemisphere lesion, type of arterial ECMO cannulation was associated with side of injury: right carotid cannulation was associated with right hemisphere lesions, and ascending aorta cannulation with left hemisphere lesions (odds ratio, 0.29 [95% confidence interval, 0.08-0.98], p = 0.03). CONCLUSIONS: After starting ECMO, cEEG background activity has the potential to inform prognosis. Type of arterial (carotid vs aortic) ECMO correlates with side of focal cerebral injury, which in ≈33% is associated with presence of ES. We hypothesize that the differential distribution reflects abnormal flow dynamics or embolic injury.

Meaning of Intracranial Pressure-to-Blood Pressure Fisher-Transformed Pearson Correlation-Derived Optimal Cerebral Perfusion Pressure: Testing Empiric Utility in a Mechanistic Model.

OBJECTIVES: Time-averaged intracranial pressure-to-blood pressure Fisher-transformed Pearson correlation (PRx) is used to assess cerebral autoregulation and derive optimal cerebral perfusion pressure. Empirically, impaired cerebral autoregulation is considered present when PRx is positive; greater difference between time series median cerebral perfusion pressure and optimal cerebral perfusion pressure (ΔCPP) is associated with worse outcomes. Our aims are to better understand: 1) the potential strategies for targeting optimal cerebral perfusion pressure; 2) the relationship between cerebral autoregulation and PRx; and 3) the determinants of greater ΔCPP. DESIGN: Mechanistic simulation using a lumped compartmental model of blood pressure, intracranial pressure, cerebral autoregulation, cerebral blood volume, PaCO2, and cerebral blood flow. SETTING: University critical care integrative modeling and precision physiology research group. SUBJECTS: None, in silico studies. INTERVENTIONS: Simulations in blood pressure, intracranial pressure, PaCO2, and impairment of cerebral autoregulation, with examination of "output" cerebral perfusion pressure versus PRx-plots, optimal cerebral perfusion pressure, and ΔCPP. MEASUREMENTS AND MAIN RESULTS: In regard to targeting optimal cerebral perfusion pressure, a shift in mean blood pressure or mean intracranial pressure with no change in mean blood pressure, with intact cerebral autoregulation, impacts optimal cerebral perfusion pressure. Second, a positive PRx occurs even with intact cerebral autoregulation. In relation to ΔCPP, for a given input blood pressure profile, with constant intracranial pressure, altering the degree of impairment in cerebral autoregulation or the level of PaCO2 maintains differences to within ±5 mm Hg. Change in intracranial pressure due to either an intermittently prolonged pattern of raised intracranial pressure or terminal escalation shows ΔCPP greater than 10 mm Hg and less than -10 mm Hg, respectively. CONCLUSIONS: These mechanistic simulations provide insight into the empiric basis of optimal cerebral perfusion pressure and the significance of PRx and ΔCPP. PRx and optimal cerebral perfusion pressure deviations do not directly reflect changes in cerebral autoregulation but are, in general, related to the presence of complex states involving well-described clinical progressions with raised intracranial pressure.