A Supervised Approach for Patient-Specific ICU Mortality Prediction Using Feature Modeling

2019 ◽  
pp. 287-295
Author(s):  
Gokul S. Krishnan ◽  
S. Sowmya Kamath
Keyword(s):  
Mortality Prediction ◽  
Feature Modeling ◽  
Patient Specific ◽  
Icu Mortality

scholarly journals High mortality rate of obstetric critically ill women in Rwanda and its predictability

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Alcade Rudakemwa ◽  
Amyl Lucille Cassidy ◽  
Théogène Twagirumugabe
Keyword(s):  
Prediction Models ◽  
Average Length ◽  
Mortality Prediction ◽  
Routine Practice ◽  
Icu Mortality ◽  
Resource Limited ◽  
Obstetric Haemorrhage ◽  
Tertiary Hospitals ◽  
Mortality Prediction Models ◽  
Qsofa Score

Abstract Background Reasons for admission to intensive care units (ICUs) for obstetric patients vary from one setting to another. Outcomes from ICU and prediction models are not well explored in Rwanda owing to lack of appropriate scores. This study aimed to assess reasons for admission and accuracy of prediction models for mortality of obstetric patients admitted to ICUs of two public tertiary hospitals in Rwanda. Methods We prospectively collected data from all obstetric patients admitted to the ICUs of the two public tertiary hospitals in Rwanda from March 2017 to February 2018 to identify reasons for admission, demographic and clinical characteristics, outcome including death and its predictability by both the Modified Early Obstetric Warning Score (MEOWS) and quick Sequential Organ Failure Assessment (qSOFA). We analysed the accuracy of mortality prediction models by MEOWS or qSOFA by using logistic regression adjusting for factors associated with mortality. Area under the Receiver Operating characteristic (AUROC) curves is used to show the predicting capacity for each individual tool. Results Obstetric patients (n = 94) represented 12.8 % of all 747 ICU admissions which is 1.8 % of all 4.999 admitted women for pregnancy or labor. Sepsis (n = 30; 31.9 %) and obstetric haemorrhage (n = 24; 25.5 %) were the two commonest reasons for ICU admission. Overall ICU mortality for obstetric patients was 54.3 % (n = 51) with average length of stay of 6.6 ± 7.525 days. MEOWS score was an independent predictor of mortality (adjusted (a)OR 1.25; 95 % CI 1.07–1.46) and so was qSOFA score (aOR 2.81; 95 % CI 1.25–6.30) with an adjusted AUROC of 0.773 (95 % CI 0.67–0.88) and 0.764 (95 % CI 0.65–0.87), indicating fair accuracy for ICU mortality prediction in these settings of both MEOWS and qSOFA scores. Conclusions Sepsis and obstetric haemorrhage were the commonest reasons for obstetric admissions to ICU in Rwanda. MEOWS and qSOFA scores could accurately predict ICU mortality of obstetric patients in resource-limited settings, but larger studies are needed before a recommendation for their use in routine practice in similar settings.

scholarly journals High mortality rate of obstetric critically ill patients in Rwanda and its predictability

2020 ◽  
Author(s):  
Alcade Rudakemwa ◽  
Amy Lucille Cassidy ◽  
Theogene Twagirumugabe
Keyword(s):  
Prediction Models ◽  
Average Length ◽  
Roc Curves ◽  
Mortality Prediction ◽  
Icu Mortality ◽  
Average Length Of Stay ◽  
Referral Hospitals ◽  
Tertiary Hospitals ◽  
Mortality Prediction Models ◽  
Qsofa Score

Abstract Background Reasons for admission at the intensive care units (ICU) for obstetric patients vary from a setting to another. Outcomes from ICU and its prediction models are not well explored in Rwanda because of lack of appropriate scores. This study intended to assess profile and accuracy of predictive models for obstetric patients admitted in ICU in the two public tertiary hospitals in Rwanda.Methods We prospectively collected data from all obstetric patients admitted in the ICU of public referral hospitals in Rwanda from March 2017 to February 2018 to identify reasons for admissions and factors for prognosis. We analysed the accuracy of mortality prediction models including the quick Sequential Organ Failure Assessment (qSOFA) and Modified Early Obstetric Warning Score (MEOWS) by using the Logistic Regression and adjusted Receiver Operating characteristic (ROC) curves. Results Obstetric patients represented 12.8% of all ICU admissions and 1.8% of all deliveries. Sepsis (31.9%) and haemorrhage (25.5%) were the two commonest reasons for ICU admission in our study participants. The overall ICU mortality for our obstetric patients was 54.3% while the average length of stay was 6.6 days. MEOWS score was an independent predictor to mortality (adjusted OR=1.25[1.07-1.46]; p=0.005) and so was the qSOFA score (adjusted OR=2.81[1.25-6.30]; p=0.012). The adjusted Area Under the ROC (AUROC) for MEOWS was 0.773[0.666-0.880] and that of the qSOFA was 0.764[0.654-0.873] signing fair accuracies for ICU mortality prediction in these settings for both models.Conclusion Sepsis is the commonest reason for admissions to ICU for obstetric patients in Rwanda. Simple models comprising MEOWS and qSOFA could accurately predict the mortality for those patients but further larger studies are needed before generalization.

An empirical study of using radiology reports and images to improve ICU-mortality prediction

2021 ◽  
Author(s):  
Mingquan Lin ◽  
Song Wang ◽  
Ying Ding ◽  
Lihui Zhao ◽  
Fei Wang ◽  
...  
Keyword(s):  
Empirical Study ◽  
Mortality Prediction ◽  
Icu Mortality ◽  
Radiology Reports

scholarly journals Heart Failure Diagnosis, Readmission, and Mortality Prediction Using Machine Learning and Artificial Intelligence Models

2020 ◽  
Vol 7 (4) ◽  
pp. 212-219 ◽  
Author(s):  
Aixia Guo ◽  
Michael Pasque ◽  
Francis Loh ◽  
Douglas L. Mann ◽  
Philip R. O. Payne
Keyword(s):  
Machine Learning ◽  
Heart Failure ◽  
Outcome Prediction ◽  
Predictive Accuracy ◽  
Imbalanced Data ◽  
Mortality Prediction ◽  
Machine Learning Techniques ◽  
Patient Specific ◽  
Learning Models ◽  
Machine Learning Models

Abstract Purpose of Review One in five people will develop heart failure (HF), and 50% of HF patients die in 5 years. The HF diagnosis, readmission, and mortality prediction are essential to develop personalized prevention and treatment plans. This review summarizes recent findings and approaches of machine learning models for HF diagnostic and outcome prediction using electronic health record (EHR) data. Recent Findings A set of machine learning models have been developed for HF diagnostic and outcome prediction using diverse variables derived from EHR data, including demographic, medical note, laboratory, and image data, and achieved expert-comparable prediction results. Summary Machine learning models can facilitate the identification of HF patients, as well as accurate patient-specific assessment of their risk for readmission and mortality. Additionally, novel machine learning techniques for integration of diverse data and improvement of model predictive accuracy in imbalanced data sets are critical for further development of these promising modeling methodologies.

scholarly journals Thyroid Hormones Predict ICU Mortality After Cardiopulmonary Bypass in Congenital Heart Disease Patients Under 3 Months Old

2020 ◽  
Author(s):  
Di Yu ◽  
Liang Zou ◽  
Yueshuang Cun ◽  
Yaping Li ◽  
Qingfeng Wang ◽  
...  
Keyword(s):  
Congenital Heart Disease ◽  
Heart Disease ◽  
Cardiopulmonary Bypass ◽  
Thyroid Hormones ◽  
Congenital Heart ◽  
Heart Surgery ◽  
Mortality Prediction ◽  
Binary Logistic Regression Analysis ◽  
Icu Mortality ◽  
Significant Difference

Abstract Backgroup: To study the effectiveness of thyroid hormones in predicting intensive care unit (ICU) mortality after cardiopulmonary bypass (CPB) in infants with congenital heart disease (CHD). Methods: We retrospective observational analyzed data from 133 patients under 3 months old who underwent cardiac surgery with CPB from June 2017 to November 2019. ICU mortality prediction was assessed by multivariate binary logistic regression analysis and area under the curve (AUC) analysis. Results: Non-survivors were younger (17.46±17.10 days vs. 38.63±26.87 days, P=0.006), with a higher proportion of neonates (9/13 vs. 41/120, P=0.017) and a higher proportion of individuals with Risk Adjustment in Congenital Heart Surgery-1 (RACHS-1) score ≥4 (8/13 vs. 31/120, P=0.020). No significant difference was found in CPB and aortic cross-clamping (ACC) time. The levels of free triiodothyronine (FT3) (3.91±0.99 pmol/L vs. 5.11±1.55 pmol/L, P=0.007) and total triiodothyronine (TT3) (1.55±0.35 nmol/L vs. 1.90±0.57 nmol/L, P=0.032) were higher in survivors compared with non-survivors. In the ICU mortality prediction assessment, only FT3 was an independent mortality predictor and showed a good AUC (0.856 ± 0.040). Conclusion: FT3 was a powerful and the only independent predictor of ICU mortality in CHD infants under 3 months old after CPB.

Stress ulcer prophylaxis and treatment drugs in critical illness

2016 ◽  
Author(s):  
Waleed Alhazzani ◽  
Deborah J. Cook
Keyword(s):  
Length Of Stay ◽  
Gastrointestinal Bleeding ◽  
Stress Ulcer ◽  
Randomized Clinical Trials ◽  
Epidemiological Data ◽  
Stress Ulcer Prophylaxis ◽  
Patient Specific ◽  
Icu Mortality ◽  
Moderate Quality ◽  
Ulcer Prophylaxis

Many changes have occurred over the last three decades in the field of stress ulcer gastrointestinal bleeding and its prevention. The topic is controversial, fuelled by disparate data, studies at risk of bias, and the impression that the problem is not as serious as it once was. Indeed, compared with over four decades ago when mucosal ulceration of the stomach causing serious bleeding was first described, a relatively small proportion of critically-ill patients now develop clinically important bleeding. Acid suppression is commonly prescribed for stress ulcer prophylaxis (SUP), targeting subgroups of patients at high risk in the intensive care unit (ICU), rather than universal prevention. The randomized clinical trials to date suggest a significant reduction in CIB with use of histamine-2-receptor antagonists (H2RAs) compared with no SUP, with no impact on pneumonia, ICU mortality, or length of stay. However, these trials are of moderate quality. More recent RCTs suggest proton pump inhibitors compared with H2RAs may significantly reduce the risk of CIB without influencing the risk of pneumonia, ICU mortality, or length of stay. These trials are also of moderate quality. Today, the decision whether to use SUP, and which agent to use, is complex. Clinical considerations include local epidemiological data (for centres documenting these outcomes), and patient-specific risks of gastrointestinal bleeding and infection, indexed to case mix.

Sign in / Sign up

Export Citation Format

Share Document