scholarly journals Predictive Modeling of Pressure Injury Risk in Patients Admitted to an Intensive Care Unit

2020 ◽  
Vol 29 (4) ◽  
pp. e70-e80
Author(s):  
Mireia Ladios-Martin ◽  
José Fernández-de-Maya ◽  
Francisco-Javier Ballesta-López ◽  
Adrián Belso-Garzas ◽  
Manuel Mas-Asencio ◽  
...  
Keyword(s):  
Machine Learning ◽  
Intensive Care Unit ◽  
Intensive Care ◽  
Injury Risk ◽  
Area Under The Curve ◽  
University Hospital ◽  
Machine Learning Techniques ◽  
Real Environment ◽  
Learning Techniques ◽  
Pressure Injury

Background Pressure injuries are an important problem in hospital care. Detecting the population at risk for pressure injuries is the first step in any preventive strategy. Available tools such as the Norton and Braden scales do not take into account all of the relevant risk factors. Data mining and machine learning techniques have the potential to overcome this limitation. Objectives To build a model to detect pressure injury risk in intensive care unit patients and to put the model into production in a real environment. Methods The sample comprised adult patients admitted to an intensive care unit (N = 6694) at University Hospital of Torrevieja and University Hospital of Vinalopó. A retrospective design was used to train (n = 2508) and test (n = 1769) the model and then a prospective design was used to test the model in a real environment (n = 2417). Data mining was used to extract variables from electronic medical records and a predictive model was built with machine learning techniques. The sensitivity, specificity, area under the curve, and accuracy of the model were evaluated. Results The final model used logistic regression and incorporated 23 variables. The model had sensitivity of 0.90, specificity of 0.74, and area under the curve of 0.89 during the initial test, and thus it outperformed the Norton scale. The model performed well 1 year later in a real environment. Conclusions The model effectively predicts risk of pressure injury. This allows nurses to focus on patients at high risk for pressure injury without increasing workload.

scholarly journals Machine learning predicts early-onset acute organ failure in critically ill patients with sickle cell disease

2019 ◽  
Author(s):  
Akram Mohammed ◽  
Pradeep S. B. Podila ◽  
Robert L. Davis ◽  
Kenneth I. Ataga ◽  
Jane S. Hankins ◽  
...  
Keyword(s):  
Machine Learning ◽  
Intensive Care ◽  
Sickle Cell Disease ◽  
Kidney Disease ◽  
Organ Failure ◽  
Organ Dysfunction ◽  
Sickle Cell ◽  
Machine Learning Techniques ◽  
Cell Disease ◽  
Learning Techniques

AbstractBackgroundSickle cell disease (SCD) is a genetic disorder of the red blood cells, resulting in multiple acute and chronic complications including pain episodes, stroke, and kidney disease. Patients with SCD develop chronic organ dysfunction, which may progress to organ failure during disease exacerbations. Early detection of acute physiological deterioration leading to organ failure is not always attainable. Machine learning techniques that allow for prediction of organ failure may enable earlier identification and treatment, and potentially reduce mortality. We tested the hypothesis that machine learning physiomarkers could predict the development of organ dysfunction in an adult sample of patients with SCD admitted to intensive care units.Methods and FindingsWe studied 63 sequential SCD patients with 163 patient encounters, mean age 33.0±11.0 years, admitted to intensive care units, some of whom (6.7%) had pre-existing cardiovascular or kidney disease. A subset of these patient encounters (37; 23%) met sequential organ failure assessment (SOFA) criteria. The site of organ failure included: central nervous system (32), cardiovascular (11), renal (10), liver (7), respiratory (5) and coagulation (2) systems. Most (81.5%) of the patient encounters who experienced organ failure had single organ failure. The other 126 SCD patient encounters served as controls. A set of signal processing features (such as fast fourier transform, energy, continuous wavelet transform, etc.) derived from heart rate, blood pressure, and respiratory rate were identified to distinguish patients with SCD who developed acute physiological deterioration leading to organ failure, from SCD patients who did not meet the criteria. A random forest model accurately predicted organ failure up to six hours prior to onset, with a five-fold cross-validation accuracy of 94.57% (average sensitivity and specificity of 90.24% and 98.9% respectively).ConclusionsThis study demonstrates the viability of using machine learning to predict acute physiological deterioration heralded organ failure among hospitalized adults with SCD. The discovery of salient physiomarkers through machine learning techniques has the potential to further accelerate the development and implementation of innovative care delivery protocols and strategies for medically vulnerable patients.

scholarly journals Incidence of pressure injury and nursing care time in intensive care

2020 ◽  
Author(s):  
Yasmin Cardoso Metwaly Mohamed Ali ◽  
Taís Milena Milena Pantaleão Souza ◽  
Paulo Carlos Garcia ◽  
Paula Cristina Nogueira
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Retrospective Study ◽  
Data Analysis ◽  
Nursing Care ◽  
University Hospital ◽  
Central Tendency ◽  
Hospital Data ◽  
Nursing Team ◽  
Pressure Injury

Objectives: To correlate the incidence of pressure injury (PI) with the average time of nursing care in an intensive care unit (ICU). Method: Epidemiological, observational, retrospective study, carried out in the ICU of a university hospital. Data were collected by consulting the PI incidence and the average nursing care time from ICU databases between 2010 and 2014. Measures of central tendency and variability, and Pearson’s correlation coefficient were used for data analysis. Results: The average incidence of PI between 2010 and 2014 was 10.83% (SD = 2.87) and the average time spent in nursing care for patients admitted to the ICU was 15 hours (SD = 0.94). There was no statistically significant correlation between the incidence of PI and the nursing care time (r = -0.17; p = 0.199), however, the results suggested an overload on the nursing team. Conclusion: This study confirms the importance of implementing and reassessing the effectiveness of preventive care protocols for PI, in addition to warning about the work overload of nursing in assisting critically ill patients.

scholarly journals Boosting algorithms for prediction in agriculture: an application of feature importance and feature selection boosting algorithms for prediction crop damage.

2021 ◽  
Author(s):  
Viviane Costa Silva ◽  
Mateus Silva Rocha ◽  
Glaucia Amorim Faria ◽  
Silvio Fernando Alves Xavier Junior ◽  
Tiago Almeida de Oliveira ◽  
...  
Keyword(s):  
Machine Learning ◽  
Area Under The Curve ◽  
Crop Damage ◽  
Machine Learning Techniques ◽  
Gradient Boosting ◽  
Agriculture Sector ◽  
Light Gradient ◽  
Learning Techniques ◽  
Gradient Boosting Machine ◽  
Boosting Algorithms

Abstract The Agriculture sector has created and collected large amounts of data. It can be gathered, stored, and analyzed to assist in decision making generating competitive value, and the use of Machine Learning techniques has been very effective for this market. In this work, a Machine Learning study was carried out using supervised classification models based on boosting to predict disease in a crop, thus identifying the model with the best areas under curve metrics. Light Gradient Boosting Machine, CatBoost Classifier, Extreme Gradient, Gradient Boosting Classifier, Adaboost models were used to qualify the crop as healthy or sick. One can see that the LightGBM algorithm provided a better fit to the data with an area under the curve of 0.76 under the use of BORUTA variable selection.

scholarly journals COVID Mortality Prediction with Machine Learning Methods: A Systematic Review and Critical Appraisal

2021 ◽  
Vol 11 (9) ◽  
pp. 893
Author(s):  
Francesca Bottino ◽  
Emanuela Tagliente ◽  
Luca Pasquini ◽  
Alberto Di Napoli ◽  
Martina Lucignani ◽  
...  
Keyword(s):  
Machine Learning ◽  
Intensive Care Unit ◽  
Clinical Decision Making ◽  
Clinical Decision ◽  
Mortality Prediction ◽  
Machine Learning Techniques ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Learning Techniques ◽  
Ventilation Modes

More than a year has passed since the report of the first case of coronavirus disease 2019 (COVID), and increasing deaths continue to occur. Minimizing the time required for resource allocation and clinical decision making, such as triage, choice of ventilation modes and admission to the intensive care unit is important. Machine learning techniques are acquiring an increasingly sought-after role in predicting the outcome of COVID patients. Particularly, the use of baseline machine learning techniques is rapidly developing in COVID mortality prediction, since a mortality prediction model could rapidly and effectively help clinical decision-making for COVID patients at imminent risk of death. Recent studies reviewed predictive models for SARS-CoV-2 diagnosis, severity, length of hospital stay, intensive care unit admission or mechanical ventilation modes outcomes; however, systematic reviews focused on prediction of COVID mortality outcome with machine learning methods are lacking in the literature. The present review looked into the studies that implemented machine learning, including deep learning, methods in COVID mortality prediction thus trying to present the existing published literature and to provide possible explanations of the best results that the studies obtained. The study also discussed challenging aspects of current studies, providing suggestions for future developments.

scholarly journals Fluid Overload Phenotypes in Critical Illness—A Machine Learning Approach

2022 ◽  
Vol 11 (2) ◽  
pp. 336
Author(s):  
Anna S. Messmer ◽  
Michel Moser ◽  
Patrick Zuercher ◽  
Joerg C. Schefold ◽  
Martin Müller ◽  
...  
Keyword(s):  
Machine Learning ◽  
Septic Shock ◽  
Logistic Regression ◽  
Intensive Care ◽  
Random Forest ◽  
Fluid Overload ◽  
Area Under The Curve ◽  
Machine Learning Techniques ◽  
Intensive Care Patients ◽  
Multivariable Logistic Regression

Background: The detrimental impact of fluid overload (FO) on intensive care unit (ICU) morbidity and mortality is well known. However, research to identify subgroups of patients particularly prone to fluid overload is scarce. The aim of this cohort study was to derive “FO phenotypes” in the critically ill by using machine learning techniques. Methods: Retrospective single center study including adult intensive care patients with a length of stay of ≥3 days and sufficient data to compute FO. Data was analyzed by multivariable logistic regression, fast and frugal trees (FFT), classification decision trees (DT), and a random forest (RF) model. Results: Out of 1772 included patients, 387 (21.8%) met the FO definition. The random forest model had the highest area under the curve (AUC) (0.84, 95% CI 0.79–0.86), followed by multivariable logistic regression (0.81, 95% CI 0.77–0.86), FFT (0.75, 95% CI 0.69–0.79) and DT (0.73, 95% CI 0.68–0.78) to predict FO. The most important predictors identified in all models were lactate and bicarbonate at admission and postsurgical ICU admission. Sepsis/septic shock was identified as a risk factor in the MV and RF analysis. Conclusion: The FO phenotypes consist of patients admitted after surgery or with sepsis/septic shock with high lactate and low bicarbonate.

scholarly journals Evaluation of PRE-DELIRIC (PREdiction of DELIRium in ICu patients) delirium prediction model for the patients in the intensive care unit

2018 ◽  
Vol 25 (1) ◽  
pp. 14-22 ◽  
Author(s):  
Gabrielė Linkaitė ◽  
Mantas Riauka ◽  
Ignė Bunevičiūtė ◽  
Saulius Vosylius
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Characteristic Curve ◽  
Area Under The Curve ◽  
Confusion Assessment Method ◽  
Assessment Method ◽  
Preventive Therapy ◽  
Apache Ii ◽  
University Hospital ◽  
Risk Patients

Introduction. Delirium not only compromises patient care, but is also associated with poorer outcomes: increased duration of mechanical ventilation, higher mortality, and greater long-term cognitive dysfunction. The PRE-DELIRIC model is a tool used to calculate the risk of the development of delirium. The classification of the patients into groups by risk allows efficient initiation of preventive measures. The goal of this study was to validate the PRE-DELIRIC model using the CAM-ICU (The Confusion Assessment Method for the Intensive Care Unit) method for the diagnosis of delirium. Materials and methods. Patients admitted to the University Hospital of Vilnius during February 2015 were enrolled. Every day, data were collected for APACHE-II and PRE-DELIRIC scores. Out of 167 patients, 38 (23%) were included and screened using the CAM-ICU method within 24 hours of admission to the ICU. We defined patients as having delirium when they had at least one positive CAM-ICU screening or haloperidol administration due to sedation. To validate the PRE-DELIRIC model, we calculated the area under receiver operating characteristic curve. Results. The mean age of the patients was 69.2 ± 17.2 years, 19 (50%) were male, APACHE-II mean score 18.0 ± 7.4 points. Delirium was diagnosed in 22 (58%) of 38 patients. Data used for validation of the PRE-DELIRIC model resulted in an area under the curve of 0.713 (p < 0.05, 95% CI 0.539–0.887); sensitivity and specificity for the patients with 20% risk were, accordingly, 77.3% and 50%; 40% risk – 45.5% and 81.3%, 60% – 36.4%, and 87.5%. Conclusions. The PRE-DELIRIC model predicted delirium in the patients within 24 hours of admission to the ICU. Preventive therapy could be efficiently targeted at high-risk patients if both of the methods are to be implemented.

scholarly journals Optimizing Outcome Prediction in Diffuse Large B-Cell Lymphoma by Use of Machine Learning and Nationwide Lymphoma Registries: A Nordic Lymphoma Group Study

2018 ◽  
pp. 1-13 ◽  
Author(s):  
Jorne L. Biccler ◽  
Sandra Eloranta ◽  
Peter de Nully Brown ◽  
Henrik Frederiksen ◽  
Mats Jerkeman ◽  
...  
Keyword(s):  
Machine Learning ◽  
Prognostic Model ◽  
Area Under The Curve ◽  
B Cell Lymphoma ◽  
Machine Learning Techniques ◽  
Prognostic Models ◽  
Brier Score ◽  
Large B Cell Lymphoma ◽  
Learning Techniques ◽  
Swedish Cohort

Purpose Prognostic models for diffuse large B-cell lymphoma (DLBCL), such as the International Prognostic Index (IPI) are widely used in clinical practice. The models are typically developed with simplicity in mind and thus do not exploit the full potential of detailed clinical data. This study investigated whether nationwide lymphoma registries containing clinical data and machine learning techniques could prove to be useful for building modern prognostic tools. Patients and Methods This study was based on nationwide lymphoma registries from Denmark and Sweden, which include large amounts of clinicopathologic data. Using the Danish DLBCL cohort, a stacking approach was used to build a new prognostic model that leverages the strengths of different survival models. To compare the performance of the stacking approach with established prognostic models, cross-validation was used to estimate the concordance index (C-index), time-varying area under the curve, and integrated Brier score. Finally, the generalizability was tested by applying the new model to the Swedish cohort. Results In total, 2,759 and 2,414 patients were included from the Danish and Swedish cohorts, respectively. In the Danish cohort, the stacking approach led to the lowest integrated Brier score, indicating that the survival curves obtained from the stacking model fitted the observed survival the best. The C-index and time-varying area under the curve indicated that the stacked model (C-index: Denmark [DK], 0.756; Sweden [SE], 0.744) had good discriminative capabilities compared with the other considered prognostic models (IPI: DK, 0.662; SE, 0.661; and National Comprehensive Cancer Network–IPI: DK, 0.681; SE, 0.681). Furthermore, these results were reproducible in the independent Swedish cohort. Conclusion A new prognostic model based on machine learning techniques was developed and was shown to significantly outperform established prognostic indices for DLBCL. The model is available at https://lymphomapredictor.org .

Sign in / Sign up

Export Citation Format

Share Document