Ventilator Weaning and Terminal Extubation: Withdrawal of Life-Sustaining Therapy in Children. Secondary Analysis of the Death One Hour After Terminal Extubation Study.

OBJECTIVE: Terminal extubation (TE) and terminal weaning (TW) during withdrawal of life-sustaining therapies (WLSTs) have been described and defined in adults. The recent Death One Hour After Terminal Extubation study aimed to validate a model developed to predict whether a child would die within 1 hour after discontinuation of mechanical ventilation for WLST. Although TW has not been described in children, pre-extubation weaning has been known to occur before WLST, though to what extent is unknown. In this preplanned secondary analysis, we aim to describe/define TE and pre-extubation weaning (PW) in children and compare characteristics of patients who had ventilatory support decreased before WLST with those who did not. DESIGN: Secondary analysis of multicenter retrospective cohort study. SETTING: Ten PICUs in the United States between 2009 and 2021. PATIENTS: Nine hundred thirteen patients 0-21 years old who died after WLST. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: 71.4% ( n = 652) had TE without decrease in ventilatory support in the 6 hours prior. TE without decrease in ventilatory support in the 6 hours prior = 71.4% ( n = 652) of our sample. Clinically relevant decrease in ventilatory support before WLST = 11% ( n = 100), and 17.6% ( n = 161) had likely incidental decrease in ventilatory support before WLST. Relevant ventilator parameters decreased were F io2 and/or ventilator set rates. There were no significant differences in any of the other evaluated patient characteristics between groups (weight, body mass index, unit type, primary diagnostic category, presence of coma, time to death after WLST, analgosedative requirements, postextubation respiratory support modality). CONCLUSIONS: Decreasing ventilatory support before WLST with extubation in children does occur. This practice was not associated with significant differences in palliative analgosedation doses or time to death after extubation.

Prehospitalization Trauma and Physiologic Factors Associated with the Presence of Post-traumatic Stress 3 Months After PICU Discharge.

BACKGROUND: Children admitted to the pediatric intensive care unit (PICU) have post-traumatic stress (PTS) rates up to 64%, and up to 28% of them meet criteria for PTS disorder (PTSD). We aim to examine whether a prior trauma history and increased physiologic parameters due to a heightened sympathetic response are associated with later PTS. Our hypothesis was children with history of prehospitalization trauma, higher heart rates, blood pressures, cortisol, and extrinsic catecholamine administration during PICU admission are more likely to have PTS after discharge. METHODS: This is a prospective, observational study of children admitted to the PICU at an urban, quaternary, academic children's hospital. Children aged 8 to 17 years old without developmental delay, severe psychiatric disorder, or traumatic brain injury were included. Children's prehospitalization trauma history was assessed with a semistructured interview. All in-hospital variables were from the electronic medical record. PTS was present if children had 4 of the Diagnostic and Statistical Manual of Mental Disorders IV criteria for PTSD. Student's t- and chi-squared tests were used to compare the presence or absence of prior trauma and all of the PICU-associated variables. RESULTS: Of the 110 children at baseline, 67 had 3-month follow-up. In the latter group, 46% met the criteria for PTS, mean age of 13 years (SD 3), 57% male, a mean PRISM III score of 4.9 (SD 4.3), and intensive care unit length of stay 6.5 days (SD 7.8). There were no statistically significant differences in the demographics of the children with and without PTS. The only variable to show significance was trauma history; children with prehospitalization trauma were more likely to have PTS at 3-month follow-up (P = .02). CONCLUSIONS: Prehospitalization trauma history was associated with the presence of PTS after admission to the PICU. This study suggests future studies should shift to the potential predictive benefit of screening children for trauma history upon PICU admission.

Continuous Prediction of Mortality in the PICU: A Recurrent Neural Network Model in a Single-Center Dataset.

OBJECTIVES: Develop, as a proof of concept, a recurrent neural network model using electronic medical records data capable of continuously assessing an individual child's risk of mortality throughout their ICU stay as a proxy measure of severity of illness. DESIGN: Retrospective cohort study. SETTING: PICU in a tertiary care academic children's hospital. PATIENTS/SUBJECTS: Twelve thousand five hundred sixteen episodes (9,070 children) admitted to the PICU between January 2010 and February 2019, partitioned into training (50%), validation (25%), and test (25%) sets. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: On 2,475 test set episodes lasting greater than or equal to 24 hours in the PICU, the area under the receiver operating characteristic curve of the recurrent neural network's 12th hour predictions was 0.94 (CI, 0.93-0.95), higher than those of Pediatric Index of Mortality 2 (0.88; CI, [0.85-0.91]; p < 0.02), Pediatric Risk of Mortality III (12th hr) (0.89; CI, [0.86-0.92]; p < 0.05), and Pediatric Logistic Organ Dysfunction day 1 (0.85; [0.81-0.89]; p < 0.002). The recurrent neural network's discrimination increased with more acquired data and smaller lead time, achieving a 0.99 area under the receiver operating characteristic curve 24 hours prior to discharge. Despite not having diagnostic information, the recurrent neural network performed well across different primary diagnostic categories, generally achieving higher area under the receiver operating characteristic curve for these groups than the other three scores. On 692 test set episodes lasting greater than or equal to 5 days in the PICU, the recurrent neural network area under the receiver operating characteristic curves significantly outperformed their daily Pediatric Logistic Organ Dysfunction counterparts (p < 0.005). CONCLUSIONS: The recurrent neural network model can process hundreds of input variables contained in a patient's electronic medical record and integrate them dynamically as measurements become available. Its high discrimination suggests the recurrent neural network's potential to provide an accurate, continuous, and real-time assessment of a child in the ICU.

Machine Learning to Predict Cardiac Death Within 1 Hour After Terminal Extubation.

OBJECTIVES: Accurate prediction of time to death after withdrawal of life-sustaining therapies may improve counseling for families and help identify candidates for organ donation after cardiac death. The study objectives were to: 1) train a long short-term memory model to predict cardiac death within 1 hour after terminal extubation, 2) calculate the positive predictive value of the model and the number needed to alert among potential organ donors, and 3) examine associations between time to cardiac death and the patient's characteristics and physiologic variables using Cox regression. DESIGN: Retrospective cohort study. SETTING: PICU and cardiothoracic ICU in a tertiary-care academic children's hospital. PATIENTS: Patients 0-21 years old who died after terminal extubation from 2011 to 2018 (n = 237). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The median time to death for the cohort was 0.3 hours after terminal extubation (interquartile range, 0.16-1.6 hr); 70% of patients died within 1 hour. The long short-term memory model had an area under the receiver operating characteristic curve of 0.85 and a positive predictive value of 0.81 at a sensitivity of 94% when predicting death within 1 hour of terminal extubation. About 39% of patients who died within 1 hour met organ procurement and transplantation network criteria for liver and kidney donors. The long short-term memory identified 93% of potential organ donors with a number needed to alert of 1.08, meaning that 13 of 14 prepared operating rooms would have yielded a viable organ. A Cox proportional hazard model identified independent predictors of shorter time to death including low Glasgow Coma Score, high Pao2-to-Fio2 ratio, low-pulse oximetry, and low serum bicarbonate. CONCLUSIONS: Our long short-term memory model accurately predicted whether a child will die within 1 hour of terminal extubation and may improve counseling for families. Our model can identify potential candidates for donation after cardiac death while minimizing unnecessarily prepared operating rooms.

The Impact of Coronavirus Disease 2019 Pandemic on U.S. and Canadian PICUs.

OBJECTIVES: There are limited reports of the impact of the coronavirus disease 2019 pandemic focused on U.S. and Canadian PICUs. This hypothesis-generating report aims to identify the United States and Canadian trends of coronavirus disease 2019 in PICUs. DESIGN AND SETTING: To better understand how the coronavirus disease 2019 pandemic was affecting U.S. and Canadian PICUs, an open voluntary daily data collection process of Canadian and U.S. PICUs was initiated by Virtual Pediatric Systems, LLC (Los Angeles, CA; http://www.myvps.org) in mid-March 2020. Information was made available online to all PICUs wishing to participate. A secondary data collection was performed to follow-up on patients discharged from those PICUs reporting coronavirus disease 2019 positive patients. MEASUREMENTS AND MAIN RESULTS: To date, over 180 PICUs have responded detailing 530 PICU admissions requiring over 3,467 days of PICU care with 30 deaths. The preponderance of cases was in the eastern regions. Twenty-four percent of the patients admitted to the PICUs were over 18 years old. Fourteen percent of admissions were under 2 years old. Nearly 60% of children had comorbidities at admission with the average length of stay increasing by age and by severity of comorbidity. Advanced respiratory support was necessary during 67% of the current days of care, with 69% being conventional mechanical ventilation. CONCLUSIONS: PICUs have been significantly impacted by the pandemic. They have provided care not only for children but also adults. Patients with coronavirus disease 2019 have a high frequency of comorbidities, require longer stays, more ventilatory support than usual PICU admissions. These data suggest several avenues for further exploration.

Development and Validation of an Empiric Tool to Predict Favorable Neurologic Outcomes Among PICU Patients.

OBJECTIVES: To create a novel tool to predict favorable neurologic outcomes during ICU stay among children with critical illness. DESIGN: Logistic regression models using adaptive lasso methodology were used to identify independent factors associated with favorable neurologic outcomes. A mixed effects logistic regression model was used to create the final prediction model including all predictors selected from the lasso model. Model validation was performed using a 10-fold internal cross-validation approach. SETTING: Virtual Pediatric Systems (VPS, LLC, Los Angeles, CA) database. PATIENTS: Patients less than 18 years old admitted to one of the participating ICUs in the Virtual Pediatric Systems database were included (2009-2015). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A total of 160,570 patients from 90 hospitals qualified for inclusion. Of these, 1,675 patients (1.04%) were associated with a decline in Pediatric Cerebral Performance Category scale by at least 2 between ICU admission and ICU discharge (unfavorable neurologic outcome). The independent factors associated with unfavorable neurologic outcome included higher weight at ICU admission, higher Pediatric Index of Morality-2 score at ICU admission, cardiac arrest, stroke, seizures, head/nonhead trauma, use of conventional mechanical ventilation and high-frequency oscillatory ventilation, prolonged hospital length of ICU stay, and prolonged use of mechanical ventilation. The presence of chromosomal anomaly, cardiac surgery, and utilization of nitric oxide were associated with favorable neurologic outcome. The final online prediction tool can be accessed at https://soipredictiontool.shinyapps.io/GNOScore/. Our model predicted 139,688 patients with favorable neurologic outcomes in an internal validation sample when the observed number of patients with favorable neurologic outcomes was among 139,591 patients. The area under the receiver operating curve for the validation model was 0.90. CONCLUSIONS: This proposed prediction tool encompasses 20 risk factors into one probability to predict favorable neurologic outcome during ICU stay among children with critical illness. Future studies should seek external validation and improved discrimination of this prediction tool.

Applying Machine Learning to Pediatric Critical Care Data.

OBJECTIVES: To explore whether machine learning applied to pediatric critical care data could discover medically pertinent information, we analyzed clinically collected electronic medical record data, after data extraction and preparation, using k-means clustering. DESIGN: Retrospective analysis of electronic medical record ICU data. SETTING: Tertiary Children's Hospital PICU. PATIENTS: Anonymized electronic medical record data from PICU admissions over 10 years. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Data from 11,384 PICU episodes were cleaned, and specific features were generated. A k-means clustering algorithm was applied, and the stability and medical validity of the resulting 10 clusters were determined. The distribution of mortality, length of stay, use of ventilation and pressors, and diagnostic categories among resulting clusters was analyzed. Clusters had significant prognostic information (p < 0.0001). Cluster membership predicted mortality (area under the curve of the receiver operating characteristic = 0.77). Length of stay, the use of inotropes and intubation, and diagnostic categories were nonrandomly distributed among the clusters (p < 0.0001). CONCLUSIONS: A standard machine learning methodology was able to determine significant medically relevant information from PICU electronic medical record data which included prognosis, diagnosis, and therapy in an unsupervised approach. Further development and application of machine learning to critical care data may provide insights into how critical illness happens to children.

Impact of 24/7 In-Hospital Intensivist Coverage on Outcomes in Pediatric Intensive Care. A Multicenter Study.

RATIONALE: The around-the-clock presence of an in-house attending critical care physician (24/7 coverage) is purported to be associated with improved outcomes among high-risk children with critical illness. OBJECTIVES: To evaluate the association of 24/7 in-house coverage with outcomes in children with critical illness. METHODS: Patients younger than 18 years of age in the Virtual Pediatric Systems Database (2009-2014) were included. The main analysis was performed using generalized linear mixed effects multivariable regression models. In addition, multiple sensitivity analyses were performed to test the robustness of our findings. MEASUREMENTS AND MAIN RESULTS: A total of 455,607 patients from 125 hospitals were included (24/7 group: 266,319 patients; no 24/7 group: 189,288 patients). After adjusting for patient and center characteristics, the 24/7 group was associated with lower mortality in the intensive care unit (ICU) (24/7 vs. no 24/7; odds ratio [OR], 0.52; 95% confidence interval [CI], 0.33-0.80; P = 0.002), a lower incidence of cardiac arrest (OR, 0.73; 95% CI, 0.54-0.99; P = 0.04), lower mortality after cardiac arrest (OR, 0.56; 95% CI, 0.340-0.93; P = 0.02), a shorter ICU stay (mean difference, -0.51 d; 95% CI, -0.93 to -0.09), and shorter duration of mechanical ventilation (mean difference, -0.68 d; 95% CI, -1.23 to -0.14). CONCLUSIONS: In this large observational study, we demonstrated that pediatric critical care provided in the ICUs staffed with a 24/7 intensivist presence is associated with improved overall patient survival and survival after cardiac arrest compared with patients treated in ICUs staffed with discretionary attending coverage. However, results from a few sensitivity analyses leave some ambiguity in these results.

Effect of Inhaled Nitric Oxide on Outcomes in Children With Acute Lung Injury: Propensity Matched Analysis From a Linked Database.

OBJECTIVES: To evaluate the effect of inhaled nitric oxide on outcomes in children with acute lung injury. DESIGN: Retrospective study with a secondary data analysis of linked data from two national databases. Propensity score matching was performed to adjust for potential confounding variables between patients who received at least 24 hours of inhaled nitric oxide (inhaled nitric oxide group) and those who did not receive inhaled nitric oxide (no inhaled nitric oxide group). SETTING: Linked data from Virtual Pediatric Systems (LLC) database and Pediatric Health Information System. PATIENTS: Patients less than 18 years old receiving mechanical ventilation for acute lung injury at nine participating hospitals were included (2009-2014). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A total of 20,106 patients from nine hospitals were included. Of these, 859 patients (4.3%) received inhaled nitric oxide for at least 24 hours during their hospital stay. Prior to matching, patients in the inhaled nitric oxide group were younger, with more comorbidities, greater severity of illness scores, higher prevalence of cardiopulmonary resuscitation, and greater resource utilization. Before matching, unadjusted outcomes, including mortality, were worse in the inhaled nitric oxide group (inhaled nitric oxide vs no inhaled nitric oxide; 25.7% vs 7.9%; p < 0.001; standardized mortality ratio, 2.6 [2.3-3.1] vs 1.1 [1.0-1.2]; p < 0.001). Propensity score matching of 521 patient pairs revealed no difference in mortality in the two groups (22.3% vs 20.2%; p = 0.40; standardized mortality ratio, 2.5 [2.1-3.0] vs 2.3 [1.9-2.8]; p = 0.53). However, the other outcomes such as ventilation free days (10.1 vs 13.6 d; p < 0.001), duration of mechanical ventilation (13.8 vs 10.1 d; p < 0.001), duration of ICU and hospital stay (15.5 vs 12.2 d; p < 0.001 and 28.0 vs 24.1 d; p < 0.001), and hospital costs ($150,569 vs $102,823; p < 0.001) were significantly worse in the inhaled nitric oxide group. CONCLUSIONS: This large observational study demonstrated that inhaled nitric oxide administration in children with acute lung injury was not associated with improved mortality. Rather, it was associated with increased hospital utilization and hospital costs.

Association of Freestanding Children's Hospitals With Outcomes in Children With Critical Illness.

OBJECTIVES: Little is known about the relationship between freestanding children's hospitals and outcomes in children with critical illness. The purpose of this study was to evaluate the association of freestanding children's hospitals with outcomes in children with critical illness. DESIGN: Propensity score matching was performed to adjust for potential confounding variables between patients cared for in freestanding or nonfreestanding children's hospitals. We tested the sensitivity of our findings by repeating the primary analyses using inverse probability of treatment weighting method and regression adjustment using the propensity score. SETTING: Retrospective study from an existing national database, Virtual PICU Systems (LLC) database. PATIENTS: Patients less than 18 years old admitted to one of the participating PICUs in the Virtual PICU Systems, LLC database were included (2009-2014). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A total of 538,967 patients from 140 centers were included. Of these, 323,319 patients were treated in 60 freestanding hospitals. In contrast, 215,648 patients were cared for in 80 nonfreestanding hospitals. By propensity matching, 134,656 patients were matched 1:1 in the two groups (67,328 in each group). Prior to matching, patients in the freestanding hospitals were younger, had greater comorbidities, had higher severity of illness scores, had higher incidence of cardiac arrest, had higher resource utilization, and had higher proportion of patients undergoing complex procedures such as cardiac surgery. Before matching, the outcomes including mortality were worse among the patients cared for in the freestanding hospitals (freestanding vs nonfreestanding, 2.5% vs 2.3%; p < 0.001). After matching, the majority of the study outcomes were better in freestanding hospitals (freestanding vs nonfreestanding, mortality: 2.1% vs 2.8%, p < 0.001; standardized mortality ratio: 0.77 [0.73-0.82] vs 0.99 [0.87-0.96], p < 0.001; reintubation: 3.4% vs 3.8%, p < 0.001; good neurologic outcome: 97.7% vs 97.1%, p = 0.001). CONCLUSIONS: In this large observational study, we demonstrated that ICU care provided in freestanding children's hospitals is associated with improved risk-adjusted survival chances compared to nonfreestanding children's hospitals. However, the clinical significance of this change in mortality should be interpreted with caution. It is also possible that the hospital structure may be a surrogate of other factors that may bias the results.

Association of house staff training with mortality in children with critical illness.

AIM: To evaluate the association of house staff training with mortality in children with critical illness. METHODS: Patients <18 years of age in the Virtual PICU Systems (VPS, LLC) Database (2009-2013) were included. The study population was divided in two study groups: hospitals with residency programme only and hospitals with both residency and fellowship programme. Control group constituted hospitals with no residency or fellowship programme. The primary study outcome was mortality before intensive care unit (ICU) discharge. Multivariable logistic regression models were fitted to evaluate association of training programmes with ICU mortality. RESULTS: A total of 336 335 patients from 108 centres were included. Case-mix of patients among the hospitals with training programmes was complex; patients cared for in the hospitals with training programmes had greater severity of illness, had higher resource utilisation and had higher overall admission risk of death compared to patients cared for in the control hospitals. Despite caring for more complex and sicker patients, the hospitals with training programmes were associated with lower odds of ICU mortality. CONCLUSION: Our study establishes that ICU care provided in hospitals with training programmes is associated with improved adjusted survival rates among the Virtual PICU database hospitals in the United States.

Evidence-Based Pediatric Outcome Predictors to Guide the Allocation of Critical Care Resources in a Mass Casualty Event.

OBJECTIVE: ICU resources may be overwhelmed by a mass casualty event, triggering a conversion to Crisis Standards of Care in which critical care support is diverted away from patients least likely to benefit, with the goal of improving population survival. We aimed to devise a Crisis Standards of Care triage allocation scheme specifically for children. DESIGN: A triage scheme is proposed in which patients would be divided into those requiring mechanical ventilation at PICU presentation and those not, and then each group would be evaluated for probability of death and for predicted duration of resource consumption, specifically, duration of PICU length of stay and mechanical ventilation. Children will be excluded from PICU admission if their mortality or resource utilization is predicted to exceed predetermined levels ("high risk"), or if they have a low likelihood of requiring ICU support ("low risk"). Children entered into the Virtual PICU Performance Systems database were employed to develop prediction equations to assign children to the exclusion categories using logistic and linear regression. Machine Learning provided an alternative strategy to develop a triage scheme independent from this process. SETTING: One hundred ten American PICUs SUBJECTS: : One hundred fifty thousand records from the Virtual PICU database. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The prediction equations for probability of death had an area under the receiver operating characteristic curve more than 0.87. The prediction equation for belonging to the low-risk category had lower discrimination. R for the prediction equations for PICU length of stay and days of mechanical ventilation ranged from 0.10 to 0.18. Machine learning recommended initially dividing children into those mechanically ventilated versus those not and had strong predictive power for mortality, thus independently verifying the triage sequence and broadly verifying the algorithm. CONCLUSION: An evidence-based predictive tool for children is presented to guide resource allocation during Crisis Standards of Care, potentially improving population outcomes by selecting patients likely to benefit from short-duration ICU interventions.

Association of center volume with outcomes in critically ill children with acute asthma.

BACKGROUND: Little is known about the relation between center volume and outcomes in children requiring intensive care unit (ICU) admission for acute asthma. OBJECTIVE: To evaluate the association of center volume with the odds of receiving positive pressure ventilation and length of ICU stay. METHODS: Patients 2 to 18 years of age with the primary diagnosis of asthma were included (2009-2012). Center volume was defined as the average number of mechanical ventilator cases per year for any diagnoses during the study period. In multivariable analysis, the odds of receiving positive pressure ventilation (invasive and noninvasive ventilation) and ICU length of stay were evaluated as a function of center volume. RESULTS: Fifteen thousand eighty-three patients from 103 pediatric ICUs with the primary diagnosis of acute asthma met the inclusion criteria. Seven hundred fifty-two patients (5%) received conventional mechanical ventilation and 964 patients (6%) received noninvasive ventilation. In multivariable analysis, center volume was not associated with the odds of receiving any form of positive pressure ventilation in children with acute asthma, with the exception of high- to medium-volume centers. However, ICU length of stay varied with center volume and was noted to be longer in low-volume centers compared with medium- and high-volume centers. CONCLUSION: In children with acute asthma, this study establishes a relation between center volume and ICU length of stay. However, this study fails to show any significant relation between center volume and the odds of receiving positive pressure ventilation; further analyses are needed to evaluate this relation in more detail.

Narrowing the gap: expected versus deployment performance.

OBJECTIVES: Successful model development requires both an accurate a priori understanding of future performance and high performance on deployment. Optimistic estimations of model performance that are unrealized in real-world clinical settings can contribute to nonuse of predictive models. This study used 2 tasks, predicting ICU mortality and Bi-Level Positive Airway Pressure failure, to quantify: (1) how well internal test performances derived from different methods of partitioning data into development and test sets estimate future deployment performance of Recurrent Neural Network models and (2) the effects of including older data in the training set on models' performance. MATERIALS AND METHODS: The cohort consisted of patients admitted between 2010 and 2020 to the Pediatric Intensive Care Unit of a large quaternary children's hospital. 2010-2018 data were partitioned into different development and test sets to measure internal test performance. Deployable models were trained on 2010-2018 data and assessed on 2019-2020 data, which was conceptualized to represent a real-world deployment scenario. Optimism, defined as the overestimation of the deployed performance by internal test performance, was measured. Performances of deployable models were also compared with each other to quantify the effect of including older data during training. RESULTS, DISCUSSION, AND CONCLUSION: Longitudinal partitioning methods, where models are tested on newer data than the development set, yielded the least optimism. Including older years in the training dataset did not degrade deployable model performance. Using all available data for model development fully leveraged longitudinal partitioning by measuring year-to-year performance.

Development of a deep learning model that predicts Bi-level positive airway pressure failure.

Delaying intubation for patients failing Bi-Level Positive Airway Pressure (BIPAP) may be associated with harm. The objective of this study was to develop a deep learning model capable of aiding clinical decision making by predicting Bi-Level Positive Airway Pressure (BIPAP) failure. This was a retrospective cohort study in a tertiary pediatric intensive care unit (PICU) between 2010 and 2020. Three machine learning models were developed to predict BIPAP failure: two logistic regression models and one deep learning model, a recurrent neural network with a Long Short-Term Memory (LSTM-RNN) architecture. Model performance was evaluated in a holdout test set. 175 (27.7%) of 630 total BIPAP sessions were BIPAP failures. Patients in the BIPAP failure group were on BIPAP for a median of 32.8 (9.2-91.3) hours prior to intubation. Late BIPAP failure (intubation after using BIPAP > 24 h) patients had fewer 28-day Ventilator Free Days (13.40 [0.68-20.96]), longer ICU length of stay and more post-extubation BIPAP days compared to those who were intubated ≤ 24 h from BIPAP initiation. An AUROC above 0.5 indicates that a model has extracted new information, potentially valuable to the clinical team, about BIPAP failure. Within 6 h of BIPAP initiation, the LSTM-RNN model predicted which patients were likely to fail BIPAP with an AUROC of 0.81 (0.80, 0.82), superior to all other models. Within 6 h of BIPAP initiation, the LSTM-RNN model would identify nearly 80% of BIPAP failures with a 50% false alarm rate, equal to an NNA of 2. In conclusion, a deep learning method using readily available data from the electronic health record can identify which patients on BIPAP are likely to fail with good discrimination, oftentimes days before they are intubated in usual practice.

Are all ICUs the same?

The ability to compare intensive care units (ICUs) and determine whether they provide the same level of care with regard to efficacy, efficiency, and quality is a cornerstone of understanding critical care and improving the quality of that care. Without collecting high-quality data, adjusted for severity of illness and analyzed in a comparative fashion, it would not be possible to describe best practices objectively, to identify which ICUs are doing a good job or to learn from those units that are. This review article discusses how and why ICUs are compared. Particular attention is focused on the severity of illness scores, standardized mortality, and comparative reporting. A data collecting network, Virtual Pediatric Systems, limited liability corporation (VPS, LLC), designed for the purposes of determining where differences in critical care can be identified and the value that this adds in improving quality is discussed. Finally, results from this large data sharing collaborative describing the practice of pediatric critical care are included for the purpose of pediatric intensive care units practice benchmarks.