Abstract 16730: Computerized Prediction of Avoidable Serum Potassium Testing in Critically Ill Cardiac Patients

Circulation ◽  
2020 ◽  
Vol 142 (Suppl_3) ◽  
Author(s):  
Francesca Sperotto ◽  
Bhaven B Patel ◽  
Mathieu Molina ◽  
Satoshi Kimura ◽  
Marlon Delgado ◽  
...  
Keyword(s):  
Machine Learning ◽  
Critically Ill ◽  
Serum Potassium ◽  
Critically Ill Patients ◽  
Urine Output ◽  
Pediatric Patients ◽  
Cardiac Patients ◽  
Machine Learning Techniques ◽  
Median Percentage ◽  
K Value

Introduction: Electrolytes are frequently monitored among critically ill patients, especially among cardiac patients due to higher risk of arrhythmias. However, a percentage of these blood draws may potentially be saved. To date, no studies have investigated strategies to safely reduce the density of electrolyte monitoring in critically ill patients. Hypothesis: We hypothesized that machine learning models can identify potentially avoidable blood draws for serum potassium (K) among pediatric patients following cardiac surgery. Methods: We retrospectively reviewed data of all patients admitted to the CICU at Boston Children’s Hospital during 2010-2018, having a length of stay ≥4 days and ≥2 recorded serum K measurements. We collected variables related to K homeostasis, including serum chemistry, hourly K intake, diuretics, and urine output. Using established machine learning techniques (Random Forest classifiers and hyperparameters) we created models predicting whether a patient’s K would be normal or abnormal based on the most recent K level, medications administered, urine output, and markers of renal function. We developed multiple models based on different age categories and temporal proximity of the most recent K measurement. We assessed the predictive performance of the models using an independent test set. Results: Of the 7,269 admissions (6,196 patients) included, 95,674 serum K was measured on average of 1 (IQR 0-1) time per day. 96% of patients received at least one dose of IV diuretic and 83% received a form of K supplementation. Our models predicted a normal K value with a median positive predictive value of 0.90. A median percentage of 2.1% measurements (mean 2.5%, IQR 1.3%-3.7%) were incorrectly predicted as normal when they were abnormal. A median percentage of 0.0% (IQR 0.0%-0.4%) were incorrectly predicted as normal while being critically low or high. A median of 27.2% (IQR 7.8%-32.4%) of samples were correctly predicted to be normal and could have been potentially avoided. Conclusions: Machine-learning methods can be used to accurately predict avoidable blood tests for serum K in critically ill pediatric patients. A median of 27.2% of samples could have been saved, with decreased costs and risk of infection or anemia.

Pediatric Diabetic Ketoacidosis and Hypoglycemia

2020 ◽  
Author(s):  
Rachel J Williams ◽  
Samantha L. Wood
Keyword(s):  
Emergency Department ◽  
Key Words ◽  
Diabetic Ketoacidosis ◽  
Critically Ill ◽  
Cerebral Edema ◽  
Critically Ill Patients ◽  
Pediatric Patients ◽  
Serum Glucose ◽  
Neurologic Function ◽  
Threat To Life

Abnormalities of serum glucose in pediatric patients are commonly encountered in the emergency department and represent an acute threat to life and neurologic function. Rapidly identifying and aggressively treating hyperglycemia with diabetic ketoacidosis and hypoglycemia are critical to ensure the best possible outcome. This review will guide the emergency provider in the identification, resuscitation, workup, and disposition of these critically ill patients. This review contains 6 figures, 13 tables, and 50 reviews. Key Words: Cerebral edema, diabetic ketoacidosis, hyperglycemia, hypoglycemia

Association Between Urine Output and Mortality in Critically Ill Patients

2021 ◽  
Vol Publish Ahead of Print ◽  
Author(s):  
Aaron J. Heffernan ◽  
Stephanie Judge ◽  
Stephen M. Petrie ◽  
Rakshitha Godahewa ◽  
Christoph Bergmeir ◽  
...  
Keyword(s):  
Critically Ill ◽  
Critically Ill Patients ◽  
Urine Output

scholarly journals Serum potassium levels and outcomes in critically ill patients in the medical intensive care unit

2018 ◽  
Vol 46 (3) ◽  
pp. 1254-1262 ◽  
Author(s):  
Surat Tongyoo ◽  
Tanuwong Viarasilpa ◽  
Chairat Permpikul
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Critically Ill ◽  
Serum Potassium ◽  
Critically Ill Patients ◽  
Medical Intensive Care Unit ◽  
Apache Ii ◽  
Apache Ii Score ◽  
Icu Mortality ◽  
Medical Intensive Care

Objective To compare the outcomes of patients with and without a mean serum potassium (K+) level within the recommended range (3.5–4.5 mEq/L). Methods This prospective cohort study involved patients admitted to the medical intensive care unit (ICU) of Siriraj Hospital from May 2012 to February 2013. The patients’ baseline characteristics, Acute Physiology and Chronic Health Evaluation II (APACHE II) score, serum K+ level, and hospital outcomes were recorded. Patients with a mean K+ level of 3.5 to 4.5 mEq/L and with all individual K+ values of 3.0 to 5.0 mEq/L were allocated to the normal K+ group. The remaining patients were allocated to the abnormal K+ group. Results In total, 160 patients were included. Their mean age was 59.3±18.3 years, and their mean APACHE II score was 21.8±14.0. The normal K+ group comprised 74 (46.3%) patients. The abnormal K+ group had a significantly higher mean APACHE II score, proportion of coronary artery disease, and rate of vasopressor treatment. An abnormal serum K+ level was associated with significantly higher ICU mortality and incidence of ventricular fibrillation. Conclusion Critically ill patients with abnormal K+ levels had a higher incidence of ventricular arrhythmia and ICU mortality than patients with normal K+ levels.

scholarly journals Mortality prediction in critically ill patients using machine learning score

2020 ◽  
Vol 788 ◽  
pp. 012029
Author(s):  
F Dzaharudin ◽  
A M Ralib ◽  
U K Jamaludin ◽  
M B M Nor ◽  
A Tumian ◽  
...  
Keyword(s):  
Machine Learning ◽  
Critically Ill ◽  
Critically Ill Patients ◽  
Mortality Prediction ◽  
Learning Score

scholarly journals Effectiveness and Safety of Potassium Replacement in Critically Ill Patients: A Retrospective Cohort Study

2019 ◽  
Vol 39 (1) ◽  
pp. e13-e18
Author(s):  
Drayton A. Hammond ◽  
Jarrod King ◽  
Niranjan Kathe ◽  
Kristina Erbach ◽  
Jelena Stojakovic ◽  
...  
Keyword(s):  
Intensive Care ◽  
Critically Ill ◽  
Serum Potassium ◽  
Critically Ill Patients ◽  
Odds Ratio ◽  
Target Range ◽  
Target Attainment ◽  
Serum Potassium Concentration ◽  
Rule Of Thumb ◽  
Potassium Replacement

Background Rules of thumb for potassium replacement are used in intensive care units despite minimal empirical validation. Objective To evaluate the effectiveness and safety of rule-of-thumb potassium replacement in critically ill patients with mild and moderate hypokalemia. Methods A retrospective, observational study was done of patients with mild (potassium, 3-3.9 mEq/L) and moderate (potassium, 2-2.9 mEq/L) hypokalemia admitted to a medical intensive care unit who received potassium replacement. Expected and actual frequencies of replacement that achieved target potassium concentrations (≥ 4 mEq/L) were compared by using a χ2 test. Logistic regression analysis was used to assess whether rule-of-thumb administration affected the probability of target attainment within 24 hours of replacement. Results Serum potassium concentrations were checked within 24 hours after potassium replacement on 354 of 577 days (61.4%) when replacement was provided. Concentrations were within target range in 82 instances (23.2%). Of 62 episodes of replacement expected to achieve the target according to the rule-of-thumb estimation, 22 did (35%). Rule-of-thumb administration was associated with greater likelihood of target attainment (odds ratio, 2.12; 95% CI, 1.18-3.85; P = .01). This difference in likelihood remained significant after adjustment for covariates (odds ratio, 2.18; 95% CI, 1.04-4.56; P = .04). Conclusion In critically ill patients given potassium replacement without regard to a formal protocol, the target serum potassium concentration was achieved more often than expected according to the rule-of-thumb estimation but less than one-third of the time.

scholarly journals Using machine learning tools to predict outcomes for emergency department intensive care unit patients

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Qiangrong Zhai ◽  
Zi Lin ◽  
Hongxia Ge ◽  
Yang Liang ◽  
Nan Li ◽  
...  
Keyword(s):  
Machine Learning ◽  
Intensive Care Unit ◽  
Emergency Department ◽  
Intensive Care ◽  
Critically Ill ◽  
Critically Ill Patients ◽  
Predictive Accuracy ◽  
Medical Center ◽  
Learning Tools ◽  
Critical Patients

AbstractThe number of critically ill patients has increased globally along with the rise in emergency visits. Mortality prediction for critical patients is vital for emergency care, which affects the distribution of emergency resources. Traditional scoring systems are designed for all emergency patients using a classic mathematical method, but risk factors in critically ill patients have complex interactions, so traditional scoring cannot as readily apply to them. As an accurate model for predicting the mortality of emergency department critically ill patients is lacking, this study’s objective was to develop a scoring system using machine learning optimized for the unique case of critical patients in emergency departments. We conducted a retrospective cohort study in a tertiary medical center in Beijing, China. Patients over 16 years old were included if they were alive when they entered the emergency department intensive care unit system from February 2015 and December 2015. Mortality up to 7 days after admission into the emergency department was considered as the primary outcome, and 1624 cases were included to derive the models. Prospective factors included previous diseases, physiologic parameters, and laboratory results. Several machine learning tools were built for 7-day mortality using these factors, for which their predictive accuracy (sensitivity and specificity) was evaluated by area under the curve (AUC). The AUCs were 0.794, 0.840, 0.849 and 0.822 respectively, for the SVM, GBDT, XGBoost and logistic regression model. In comparison with the SAPS 3 model (AUC = 0.826), the discriminatory capability of the newer machine learning methods, XGBoost in particular, is demonstrated to be more reliable for predicting outcomes for emergency department intensive care unit patients.

scholarly journals Incidence, timing and outcome of AKI in critically ill patients varies with the definition used and the addition of urine output criteria

2017 ◽  
Vol 18 (1) ◽  
Author(s):  
J. Koeze ◽  
F. Keus ◽  
W. Dieperink ◽  
I. C. C. van der Horst ◽  
J. G. Zijlstra ◽  
...  
Keyword(s):  
Critically Ill ◽  
Critically Ill Patients ◽  
Urine Output
Sign in / Sign up

Export Citation Format

Share Document