scholarly journals Machine learning-based prediction models for accidental hypothermia patients

2021 ◽  
Vol 9 (1) ◽  
Author(s):  
Yohei Okada ◽  
Tasuku Matsuyama ◽  
Sachiko Morita ◽  
Naoki Ehara ◽  
Nobuhiro Miyamae ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cohort Study ◽  
Random Forest ◽  
Hospital Mortality ◽  
Retrospective Cohort Study ◽  
Retrospective Cohort ◽  
Prediction Models ◽  
Validation Cohort ◽  
Accidental Hypothermia ◽  
Gradient Boosting

Abstract Background Accidental hypothermia is a critical condition with high risks of fatal arrhythmia, multiple organ failure, and mortality; however, there is no established model to predict the mortality. The present study aimed to develop and validate machine learning-based models for predicting in-hospital mortality using easily available data at hospital admission among the patients with accidental hypothermia. Method This study was secondary analysis of multi-center retrospective cohort study (J-point registry) including patients with accidental hypothermia. Adult patients with body temperature 35.0 °C or less at emergency department were included. Prediction models for in-hospital mortality using machine learning (lasso, random forest, and gradient boosting tree) were made in development cohort from six hospitals, and the predictive performance were assessed in validation cohort from other six hospitals. As a reference, we compared the SOFA score and 5A score. Results We included total 532 patients in the development cohort [N = 288, six hospitals, in-hospital mortality: 22.0% (64/288)], and the validation cohort [N = 244, six hospitals, in-hospital mortality 27.0% (66/244)]. The C-statistics [95% CI] of the models in validation cohorts were as follows: lasso 0.784 [0.717–0.851] , random forest 0.794[0.735–0.853], gradient boosting tree 0.780 [0.714–0.847], SOFA 0.787 [0.722–0.851], and 5A score 0.750[0.681–0.820]. The calibration plot showed that these models were well calibrated to observed in-hospital mortality. Decision curve analysis indicated that these models obtained clinical net-benefit. Conclusion This multi-center retrospective cohort study indicated that machine learning-based prediction models could accurately predict in-hospital mortality in validation cohort among the accidental hypothermia patients. These models might be able to support physicians and patient’s decision-making. However, the applicability to clinical settings, and the actual clinical utility is still unclear; thus, further prospective study is warranted to evaluate the clinical usefulness.

scholarly journals Using Machine Learning Algorithms to Predict Hospital Acquired Thrombocytopenia after Operation in the Intensive Care Unit: A Retrospective Cohort Study

Diagnostics ◽  
2021 ◽  
Vol 11 (9) ◽  
pp. 1614
Author(s):  
Yisong Cheng ◽  
Chaoyue Chen ◽  
Jie Yang ◽  
Hao Yang ◽  
Min Fu ◽  
...  
Keyword(s):  
Machine Learning ◽  
Intensive Care Unit ◽  
Cohort Study ◽  
Intensive Care ◽  
Retrospective Cohort Study ◽  
Retrospective Cohort ◽  
Machine Learning Algorithms ◽  
Gradient Boosting ◽  
Test Set ◽  
Hospital Acquired

Hospital acquired thrombocytopenia (HAT) is a common hematological complication after surgery. This research aimed to develop and compare the performance of seven machine learning (ML) algorithms for predicting patients that are at risk of HAT after surgery. We conducted a retrospective cohort study which enrolled adult patients transferred to the intensive care unit (ICU) after surgery in West China Hospital of Sichuan University from January 2016 to December 2018. All subjects were randomly divided into a derivation set (70%) and test set (30%). ten-fold cross-validation was used to estimate the hyperparameters of ML algorithms during the training process in the derivation set. After ML models were developed, the sensitivity, specificity, area under the curve (AUC), and net benefit (decision analysis curve, DCA) were calculated to evaluate the performances of ML models in the test set. A total of 10,369 patients were included and in 1354 (13.1%) HAT occurred. The AUC of all seven ML models exceeded 0.7, the two highest were Gradient Boosting (GB) (0.834, 0.814–0.853, p < 0.001) and Random Forest (RF) (0.828, 0.807–0.848, p < 0.001). There was no difference between GB and RF (0.834 vs. 0.828, p = 0.293); however, these two were better than the remaining five models (p < 0.001). The DCA revealed that all ML models had high net benefits with a threshold probability approximately less than 0.6. In conclusion, we found that ML models constructed by multiple preoperative variables can predict HAT in patients transferred to ICU after surgery, which can improve risk stratification and guide management in clinical practice.

scholarly journals Proton pump inhibitors versus histamine-2 receptor blockers for stress ulcer prophylaxis in patients with sepsis: a retrospective cohort study

2021 ◽  
Vol 49 (6) ◽  
pp. 030006052110251
Author(s):  
Minqiang Huang ◽  
Ming Han ◽  
Wei Han ◽  
Lei Kuang
Keyword(s):  
Risk Factors ◽  
Cohort Study ◽  
Hospital Mortality ◽  
Critically Ill ◽  
Proton Pump ◽  
Retrospective Cohort Study ◽  
Retrospective Cohort ◽  
Stress Ulcer ◽  
Increased Risk ◽  
Ulcer Prophylaxis

Objective We aimed to compare the efficacy and risks of proton pump inhibitor (PPI) versus histamine-2 receptor blocker (H2B) use for stress ulcer prophylaxis (SUP) in critically ill patients with sepsis and risk factors for gastrointestinal bleeding (GIB). Methods In this retrospective cohort study, we used the Medical Information Mart for Intensive Care III Clinical Database to identify critically ill adult patients with sepsis who had at least one risk factor for GIB and received either an H2B or PPI for ≥48 hours. Propensity score matching (PSM) was conducted to balance baseline characteristics. The primary outcome was in-hospital mortality. Results After 1:1 PSM, 1056 patients were included in the H2B and PPI groups. The PPI group had higher in-hospital mortality (23.8% vs. 17.5%), GIB (8.9% vs. 1.6%), and pneumonia (49.6% vs. 41.6%) rates than the H2B group. After adjusting for risk factors of GIB and pneumonia, PPI use was associated with a 1.28-times increased risk of in-hospital mortality, 5.89-times increased risk of GIB, and 1.32-times increased risk of pneumonia. Conclusions Among critically ill adult patients with sepsis at risk for GIB, SUP with PPIs was associated with higher in-hospital mortality and higher risk of GIB and pneumonia than H2Bs.

scholarly journals Incidence and Outcomes of Acute Kidney Injury in Patients Admitted to Hospital With COVID-19: A Retrospective Cohort Study

2021 ◽  
Vol 8 ◽  
pp. 205435812110277
Author(s):  
Tyler Pitre ◽  
Angela (Hong Tian) Dong ◽  
Aaron Jones ◽  
Jessica Kapralik ◽  
Sonya Cui ◽  
...  
Keyword(s):  
Mechanical Ventilation ◽  
Acute Kidney Injury ◽  
Cohort Study ◽  
Hospital Mortality ◽  
Disease Severity ◽  
Retrospective Cohort Study ◽  
Odds Ratio ◽  
Retrospective Cohort ◽  
Kidney Injury ◽  
Icu Admission

Background: The incidence of acute kidney injury (AKI) in patients with COVID-19 and its association with mortality and disease severity is understudied in the Canadian population. Objective: To determine the incidence of AKI in a cohort of patients with COVID-19 admitted to medicine and intensive care unit (ICU) wards, its association with in-hospital mortality, and disease severity. Our aim was to stratify these outcomes by out-of-hospital AKI and in-hospital AKI. Design: Retrospective cohort study from a registry of patients with COVID-19. Setting: Three community and 3 academic hospitals. Patients: A total of 815 patients admitted to hospital with COVID-19 between March 4, 2020, and April 23, 2021. Measurements: Stage of AKI, ICU admission, mechanical ventilation, and in-hospital mortality. Methods: We classified AKI by comparing highest to lowest recorded serum creatinine in hospital and staged AKI based on the Kidney Disease: Improving Global Outcomes (KDIGO) system. We calculated the unadjusted and adjusted odds ratio for the stage of AKI and the outcomes of ICU admission, mechanical ventilation, and in-hospital mortality. Results: Of the 815 patients registered, 439 (53.9%) developed AKI, 253 (57.6%) presented with AKI, and 186 (42.4%) developed AKI in-hospital. The odds of ICU admission, mechanical ventilation, and death increased as the AKI stage worsened. Stage 3 AKI that occurred during hospitalization increased the odds of death (odds ratio [OR] = 7.87 [4.35, 14.23]). Stage 3 AKI that occurred prior to hospitalization carried an increased odds of death (OR = 5.28 [2.60, 10.73]). Limitations: Observational study with small sample size limits precision of estimates. Lack of nonhospitalized patients with COVID-19 and hospitalized patients without COVID-19 as controls limits causal inferences. Conclusions: Acute kidney injury, whether it occurs prior to or after hospitalization, is associated with a high risk of poor outcomes in patients with COVID-19. Routine assessment of kidney function in patients with COVID-19 may improve risk stratification. Trial registration: The study was not registered on a publicly accessible registry because it did not involve any health care intervention on human participants.

scholarly journals Machine Learning-based in-hospital Mortality Prediction Models for Patients With Acute Coronary Syndrome

2020 ◽  
Author(s):  
Jun Ke ◽  
Yiwei Chen ◽  
Xiaoping Wang ◽  
Zhiyong Wu ◽  
qiongyao Zhang ◽  
...  
Keyword(s):  
Machine Learning ◽  
Logistic Regression ◽  
Random Forest ◽  
Hospital Mortality ◽  
Operating Characteristic ◽  
Prediction Models ◽  
Characteristic Curve ◽  
Multivariate Logistic Regression Analysis ◽  
Hdl Cholesterol ◽  
Coronary Syndrome

Abstract BackgroundThe purpose of this study is to identify the risk factors of in-hospital mortality in patients with acute coronary syndrome (ACS) and to evaluate the performance of traditional regression and machine learning prediction models.MethodsThe data of ACS patients who entered the emergency department of Fujian Provincial Hospital from January 1, 2017 to March 31, 2020 for chest pain were retrospectively collected. The study used univariate and multivariate logistic regression analysis to identify risk factors for in-hospital mortality of ACS patients. The traditional regression and machine learning algorithms were used to develop predictive models, and the sensitivity, specificity, and receiver operating characteristic curve were used to evaluate the performance of each model.ResultsA total of 7810 ACS patients were included in the study, and the in-hospital mortality rate was 1.75%. Multivariate logistic regression analysis found that age and levels of D-dimer, cardiac troponin I, N-terminal pro-B-type natriuretic peptide (NT-proBNP), lactate dehydrogenase (LDH), high-density lipoprotein (HDL) cholesterol, and calcium channel blockers were independent predictors of in-hospital mortality. The study found that the area under the receiver operating characteristic curve of the models developed by logistic regression, gradient boosting decision tree (GBDT), random forest, and support vector machine (SVM) for predicting the risk of in-hospital mortality were 0.963, 0.960, 0.963, and 0.959, respectively. Feature importance evaluation found that NT-proBNP, LDH, and HDL cholesterol were top three variables that contribute the most to the prediction performance of the GBDT model and random forest model.ConclusionsThe predictive model developed using logistic regression, GBDT, random forest, and SVM algorithms can be used to predict the risk of in-hospital death of ACS patients. Based on our findings, we recommend that clinicians focus on monitoring the changes of NT-proBNP, LDH, and HDL cholesterol, as this may improve the clinical outcomes of ACS patients.

scholarly journals Risk factors for mortality in hospitalized patients with COVID-19 from three hospitals in Peru: a retrospective cohort study

F1000Research ◽  
2021 ◽  
Vol 10 ◽  
pp. 224
Author(s):  
Cristian Díaz-Vélez ◽  
Diego Urrunaga-Pastor ◽  
Anthony Romero-Cerdán ◽  
Eric Ricardo Peña-Sánchez ◽  
Jorge Luis Fernández Mogollon ◽  
...  
Keyword(s):  
Risk Factors ◽  
Cohort Study ◽  
Hospital Mortality ◽  
Oxygen Saturation ◽  
Retrospective Cohort Study ◽  
Retrospective Cohort ◽  
Cox Regression ◽  
Vulnerable Groups ◽  
Vital Functions ◽  
The Mean

Background: Peru was one of the countries with the highest COVID-19 mortality worldwide during the first stage of the pandemic. It is then relevant to evaluate the risk factors for mortality in patients hospitalized for COVID-19 in three hospitals in Peru in 2020, from March to May, 2020.  Methods: We carried out a retrospective cohort study. The population consisted of patients from three Peruvian hospitals hospitalized for a diagnosis of COVID-19 during the March-May 2020 period. Independent sociodemographic variables, medical history, symptoms, vital functions, laboratory parameters and medical treatment were evaluated. In-hospital mortality was assessed as the outcome. We performed Cox regression models (crude and adjusted) to evaluate risk factors for in-hospital mortality. Hazard ratios (HR) with their respective 95% confidence intervals (95% CI) were calculated.  Results: We analyzed 493 hospitalized adults; 72.8% (n=359) were male and the mean age was 63.3 ± 14.4 years. COVID-19 symptoms appeared on average 7.9 ± 4.0 days before admission to the hospital, and the mean oxygen saturation on admission was 82.6 ± 13.8. While 67.6% (n=333) required intensive care unit admission, only 3.3% (n=16) were admitted to this unit, and 60.2% (n=297) of the sample died. In the adjusted regression analysis, it was found that being 60 years old or older (HR=1.57; 95% CI: 1.14-2.15), having two or more comorbidities (HR=1.53; 95% CI: 1.10-2.14), oxygen saturation between 85-80% (HR=2.52; 95% CI: 1.58-4.02), less than 80% (HR=4.59; 95% CI: 3.01-7.00), and being in the middle (HR=1.65; 95% CI: 1.15-2.39) and higher tertile (HR=2.18; 95% CI: 1.51-3.15) of the neutrophil-to-lymphocyte ratio, increased the risk of mortality.  Conclusions: The risk factors found agree with what has been described in the literature and allow the identification of vulnerable groups in whom monitoring and early identification of symptoms should be prioritized in order to reduce mortality.

Sign in / Sign up

Export Citation Format

Share Document