Machine Learning for Outcome Prediction in Electroencephalograph (EEG)-Monitored Children in the Intensive Care Unit

2018 ◽  
Vol 33 (8) ◽  
pp. 546-553 ◽  
Author(s):  
Iván Sánchez Fernández ◽  
Arnold J. Sansevere ◽  
Marina Gaínza-Lein ◽  
Kush Kapur ◽  
Tobias Loddenkemper
Keyword(s):  
Machine Learning ◽  
Intensive Care Unit ◽  
Intensive Care ◽  
Hospital Mortality ◽  
Critically Ill ◽  
Explanatory Models ◽  
Critically Ill Children ◽  
Support Vector ◽  
Discriminative Performance ◽  
Number Of Patients

The aim of this study was to evaluate the performance of models predicting in-hospital mortality in critically ill children undergoing continuous electroencephalography (cEEG) in the intensive care unit (ICU). We evaluated the performance of machine learning algorithms for predicting mortality in a database of 414 critically ill children undergoing cEEG in the ICU. The area under the receiver operating characteristic curve (AUC) in the test subset was highest for stepwise selection/elimination models (AUC = 0.82) followed by least absolute shrinkage and selection operator (LASSO) and support vector machine with linear kernel (AUC = 0.79), and random forest (AUC = 0.71). The explanatory models had the poorest discriminative performance (AUC = 0.63 for the model without considering etiology and AUC = 0.45 for the model considering etiology). Using few variables and a relatively small number of patients, machine learning techniques added information to explanatory models for prediction of in-hospital mortality.

scholarly journals Analysis of serum lactate levels to predict in hospital mortality in critically ill children admitted to pediatric intensive care unit

2019 ◽  
Vol 6 (6) ◽  
pp. 2538
Author(s):  
Trilok Rao Srigiri ◽  
Partha Saradhi Manyam ◽  
Uma Mahesh ◽  
Gangadhar Belavadi
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Hospital Mortality ◽  
Blood Lactate ◽  
Critically Ill ◽  
Pediatric Intensive Care ◽  
Critically Ill Children ◽  
Lactate Levels ◽  
Blood Lactate Levels ◽  
High Lactate

Background: The predictive significance of lactate measurement at admission for mortality in critically ill children remains uncertain. Authors  objectives was to study evaluated the predictive value of blood lactate levels at admission and determined the cut-off values for predicting in-hospital mortality in the critically ill pediatric population.Methods: A prospective observational study was done in 100 critically ill admissions to the pediatric intensive care unit (PICU), requiring hemodynamic/respiratory support.  The chi-square test for categorical variables performs the comparison.Results:  Out of 100 patients, 22 (22%) expired. Mortality is highest in 10-16 age (7%). In the non-survivor group, the majority of patients were diagnosed as pneumonia (7.5%). Median lactate levels in non-survivors are 4.5 at admission when compared to 2.0 in survivors (p<0.001). The mortality rates left rate in the high lactate group (73%) is more when compared to intermediate (20%) and low-level groups (7%). Blood lactate was 75% sensitive and 90% specific at the optimal cut-off value of 33.7 mg/dl. The positive likelihood ratio of predicting death is more with a high lactate level (7.5) when compared to intermediate (0.8) and low levels (0.08). Sensitivity and Specificity with elevated lactate levels is the mortality 24 hrs (89%, 92%) than at admission (75%, 90%). The AUROC values with the admission lactate level are 0.86, and after 24 hrs are 0.95.Conclusions: Blood lactate levels at admission predict mortality in critically ill children requiring hemodynamic/respiratory support.

Abstract 15293: Noninvasive Echocardiographic Cardiac Power Output Predicts Mortality in Cardiac Intensive Care Unit Patients

Circulation ◽  
2020 ◽  
Vol 142 (Suppl_3) ◽  
Author(s):  
Barry Burstein ◽  
Vidhu Anand ◽  
Bradley Ternus ◽  
Meir Tabi ◽  
Nandan S Anavekar ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Hospital Mortality ◽  
Critically Ill ◽  
Power Output ◽  
Critically Ill Patients ◽  
Cardiac Intensive Care Unit ◽  
Cardiac Power ◽  
The Mean ◽  
Cardiac Intensive Care

Introduction: A low cardiac power output (CPO), measured invasively, identifies critically ill patients at increased risk of mortality. CPO can also be measured non-invasively with transthoracic echocardiography (TTE), although prognostic data in critically ill patients is not available. Hypothesis: Reduced CPO measured by TTE is associated with increased hospital mortality in cardiac intensive care unit (CICU) patients. Methods: Using a database of CICU patients admitted between 2007 and 2018, we identified patients with TTE within one day (before or after) of CICU admission who had data necessary for calculation of CPO. Multivariable logistic regression determined the relationship between CPO and adjusted hospital mortality. Results: We included 5,585 patients with a mean age of 68.3±14.8 years, including 36.7% females. Admission diagnoses included acute coronary syndrome (ACS) in 57%, heart failure (HF) in 50%, cardiac arrest (CA) in 12%, and cardiogenic shock (CS) in 13%. The mean left ventricular ejection fraction (LVEF) was 47±16%, and the mean CPO was 1.0±0.4 W. CPO was inversely associated with the risk of hospital mortality (Figure A), including among patients with ACS, HF, and CS (Figure B). On multivariable analysis, lower CPO was associated with higher hospital mortality (OR 0.96 per 0.1 W, 95% CI 0.0.93-0.99, p=0.03). Hospital mortality was highest in patients with low CPO coupled with reduced LVEF, increased vasopressor requirements, or higher admission lactate. Hospital mortality was higher among patients with a CPO <0.6 W (adjusted OR 1.57, 95% CI 1.13-2.19, p = 0.007), particularly in the presence of admission lactate level >4 mmol/L (50.9%). Conclusions: Echocardiographic CPO was inversely associated with hospital mortality in CICU patients, particularly among patients with increased lactate and vasopressor requirements. Routine measurement of CPO provides important information beyond LVEF and should be considered in CICU patients.

scholarly journals Use of body composition and phase angle analysis for the assessment of nutritional status and clinical outcomes in critically ill children

2021 ◽  
Author(s):  
Zi-Hong Xiong ◽  
Xue-Mei Zheng ◽  
Guo-Ying Zhang ◽  
Meng-Jun Wu ◽  
Yi Qu
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Nutritional Status ◽  
Clinical Outcomes ◽  
Critically Ill ◽  
Pediatric Intensive Care Unit ◽  
Phase Angle ◽  
Pediatric Intensive Care ◽  
Critically Ill Children ◽  
Duration Time

Abstract BackgroundMalnutrition is highly prevalent in critically ill children in the pediatric intensive care unit .We aimed to investigate the efficiency of bioelectrical impedance analysis (BIA) measurements and phase angle (PhA) analysis for the assessment of nutritional risk and clinical outcomes in critically ill children.MethodsThis single-center observational study included patients admitted to the Pediatric Intensive Care Unit (PICU) of Chengdu Women’s and Children’s Central Hospital. All patients underwent anthropometric measurement in the first 24 h of admission and underwent BIA measurements within 3 days after the admission. The patients were classified into different groups based on body mass index (BMI) for age. Electronic hospital medical records were reviewed to collect clinical data for each patient. All the obtained data were analyzed by the statistics method.ResultsThere were 204 patients enrolled in our study, of which 32.4% were diagnosed with malnutrition. We found that BMI, arm muscle circumference, fat mass, and %body fat were lower in the group with poorer nutritional status (P < 0.05). Evident differences in the score of the Pediatric Risk of Mortality and the duration of mechanical ventilation (MV) among the three groups with different nutritional statuses were observed (P < 0.05). Patients in the severely malnourished group had the longest duration of MV. In the MV groups, there were significant differences (P < 0.05) in albumin level, PhA, and extracellular water/total body water (ECW/TBW ratio). The ECW/TBW ratio and the time for PICU stay had a weak degree of correlation (Pearson correlation coefficient = 0.375). PhA showed a weak degree of correlation with the duration time of medical ventilation (coefficient of correlation = 0.398).ConclusionBIA can be considered an alternative way to assess nutritional status in critically ill children. ECW/TBW ratio and PhA were correlated with PICU stay and duration time of medical ventilation, respectively.

scholarly journals Development and Validation of Unplanned Extubation Prediction Models Using Intensive Care Unit Data: Comparative Machine Learning Study (Preprint)

2020 ◽  
Author(s):  
Sujeong Hur ◽  
Ji Young Min ◽  
Junsang Yoo ◽  
Kyunga Kim ◽  
Chi Ryang Chung ◽  
...  
Keyword(s):  
Machine Learning ◽  
Intensive Care Unit ◽  
Patient Safety ◽  
Intensive Care ◽  
Predictive Value ◽  
Prediction Models ◽  
Support Vector ◽  
Unplanned Extubation ◽  
Electronic Health Record Data ◽  
Icu Patients

BACKGROUND Patient safety in the intensive care unit (ICU) is one of the most critical issues, and unplanned extubation (UE) is considered as the most adverse event for patient safety. Prevention and early detection of such an event is an essential but difficult component of quality care. OBJECTIVE This study aimed to develop and validate prediction models for UE in ICU patients using machine learning. METHODS This study was conducted an academic tertiary hospital in Seoul. The hospital had approximately 2,000 inpatient beds and 120 intensive care unit (ICU) beds. The number of patients, on daily basis, was approximately 9,000 for the out-patient. The number of annual ICU admission was approximately 10,000. We conducted a retrospective study between January 1, 2010 and December 31, 2018. A total of 6,914 extubation cases were included. We developed an unplanned extubation prediction model using machine learning algorithms, which included random forest (RF), logistic regression (LR), artificial neural network (ANN), and support vector machine (SVM). For evaluating the model’s performance, we used area under the receiver operator characteristic curve (AUROC). Sensitivity, specificity, positive predictive value negative predictive value, and F1-score were also determined for each model. For performance evaluation, we also used calibration curve, the Brier score, and the Hosmer-Lemeshow goodness-of-fit statistic. RESULTS Among the 6,914 extubation cases, 248 underwent UE. In the UE group, there were more males than females, higher use of physical restraints, and fewer surgeries. The incidence of UE was more likely to occur during the night shift compared to the planned extubation group. The rate of reintubation within 24 hours and hospital mortality was higher in the UE group. The UE prediction algorithm was developed, and the AUROC for RF was 0.787, for LR was 0.762, for ANN was 0.762, and for SVM was 0.740. CONCLUSIONS We successfully developed and validated machine learning-based prediction models to predict UE in ICU patients using electronic health record data. The best AUROC was 0.787, which was obtained using RF. CLINICALTRIAL N/A

DOES CHLORHEXIDINE WIPE HELP IN REDUCING HOSPITAL ACQUIRED INFECTIONS (HAIS) AMONG CRITICALLY ILL CHILDREN ADMITTED TO A PAEDIATRIC INTENSIVE CARE UNIT?

2021 ◽  
pp. 35-37
Author(s):  
Madhan Kumar ◽  
Jolly Chandran ◽  
Pragathesh Pragathesh ◽  
Ebor Jacob Gnananayagam ◽  
Hema Paul ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Critically Ill ◽  
Paediatric Intensive Care Unit ◽  
Paediatric Intensive Care ◽  
Critically Ill Children ◽  
Intervention Period ◽  
Hospital Acquired Infections ◽  
Hospital Acquired ◽  
Historical Controls

OBJECTIVE: To determine the effect of chlorhexidine wipes in reducing the incidence of hospital acquired infections (HAIs) among critically ill children admitted in Paediatric Intensive Care Unit (PICU). METHODS: An interventional study, wherein enrolled children were wiped with chlorhexidine after routine bath. The incidence of HAIs were noted and compared with data from historical controls of previous year during the same period (pre-intervention). RESULTS: One hundred and ninety nine children in the intervention period were compared with 271 children from pre-intervention period. The numbers of ventilator-days were 777 and 696 respectively for the intervention period and pre-intervention periods. Incidence of ventilator associated pneumonia (VAP) reduced from 12.9/1000 ventilator-days in the pre-intervention period to 6.4/1000 ventilator-days in the intervention period (p=0.1). VAP prevalence was 3.3% in the pre-intervention period as compared to 2.5% in the intervention period (p=0.6). The incidence of CLABSI was 3.6/1000 catheter-days (catheter days: 1377) with prevalence of 2.5% in the intervention period, whereas among the historic controls of the previous year it was 4.2/1000 days (catheter days 1432) with a prevalence of 2.2% (p= 0.8). No untoward effect was reported. CONCLUSION: The use of chlorhexidine wipes in ICU was feasible but did not signicantly decrease HAIs.

scholarly journals Assessment of delirium in children admitted into the Intensive Care Unit: psCAM-ICU Tool

2020 ◽  
Vol 38 (2) ◽  
pp. 140-148
Author(s):  
Ángela María Henao Castaño ◽  
Edwar Yamith Pinzon Casas
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Critically Ill ◽  
Pediatric Intensive Care ◽  
Confusion Assessment Method ◽  
Assessment Method ◽  
Primary Diagnosis ◽  
Critically Ill Children ◽  
Associated Risk Factors ◽  
Prolonged Hospital

Background: Delirium has been identified as a risk factor for the mortality of critically ill patients, generating great social and economic impacts, since patients require more days of mechanical ventilation and a prolonged hospital stay in the intensive care unit (ICU), thus increasing medical costs. Objective: To describe the prevalence and characteristics of delirium episodes in a sample of 6-month to 5-year-old children who are critically ill. Methods: Cohort study at a Pediatric Intensive Care Unit (PICU) in Bogotá (Colombia). Participants were assessed by the Preschool Confusion Assessment Method for the ICU (psCAM-ICU) within the first twenty-four hours of hospitalization. Results: One quarter of the participants (25.8%) presented some type of delirium. Among them, two sub-types of delirium were observed: 62.5% of the cases were hypoactive and 37.5% hyperactive. Moreover, from them, six were male (75%) and 2 female (25%). Primary diagnosis was respiratory tract infection in 62.55% of the patients, while respiratory failure was diagnosed in the remaining 37.5%. Conclusions: The implementation of delirium monitoring tools in critically ill children provides a better understanding of the clinical manifestation of this phenomenon and associated risk factors in order to contribute to the design of efficient intervention strategies.

scholarly journals 2 Systemic glucocorticoids and bradycardia in critically ill children: a retrospective study

2020 ◽  
Vol 25 (Supplement_2) ◽  
pp. e1-e1
Author(s):  
Camille Maltais-Bilodeau ◽  
Maryse Frenette ◽  
Geneviève Morissette ◽  
Dennis Bailey ◽  
Karine Cloutier ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Heart Rate ◽  
Intensive Care ◽  
Retrospective Study ◽  
Critically Ill ◽  
Medical Staff ◽  
Pediatric Population ◽  
Critically Ill Children ◽  
Rate Variation ◽  
The Impact

Abstract Background Glucocorticoids are widely used in the pediatric population. They are associated with numerous side effects including repercussions on the cardiovascular system. The impact on heart rate is not well known, but bradycardia has been reported, mostly with high doses. Objectives We described the occurrence of bradycardias and the variation of heart rate in critically ill children receiving glucocorticoids. Design/Methods We conducted a retrospective study including 1 month old to 18 year old children admitted to the Pediatric Intensive Care Unit between 2014 and 2017, who received a glucocorticoid dose equivalent to 1 to 15 mg/kg/day of prednisone. We collected data on exposition to glucocorticoids, heart rate before, during and after the exposition, and interventions from the medical staff in response to bradycardia. The primary outcome was the occurrence of bradycardia and the secondary outcomes were the magnitude of heart rate variation and the clinical management of bradycardias. Results We included 92 admissions (85 patients). The median dose of glucocorticoid used was 2.80 mg/kg/day of prednisone (2.08—3.80). We found 70 cases (76%) with at least one bradycardia. Before treatment, all patients had a mean heart rate higher than the 5th percentile for age. During exposition to glucocorticoids, 8 patients (10%, n = 83) had a median heart rate ≤ 5th percentile. We noted 46 cases of bradycardia (50%) that led to an intervention from the medical staff, but no patient had a major event associated to bradycardia. We found a significant association between bradycardia and age (estimate -0.136, 95% CI -0.207—-0.065, p &lt; 0.001), glucocorticoid dose (estimate 4.820, 95% CI 2.048—7.592, p &lt; 0.001) and intravenous administration (estimate 8.709, 95% CI 1.893—15.524, p = 0.012). Conclusion In our study, most children hospitalized at the intensive care unit receiving standard doses of glucocorticoid experienced bradycardia. The majority of episodes led to an intervention from the medical staff. Presence of bradycardia was associated with younger age, higher dose and IV administration of glucocorticoids.

Sign in / Sign up

Export Citation Format

Share Document