Dataset Information

Validation of a Retrospective Computing Model for Mortality Risk in the Intensive Care Unit.

ABSTRACT:

Objective

To compare the predictive performance of Epic Systems Corporation's proprietary intensive care unit (ICU) mortality risk model (IMRM) with that of the Acute Physiology and Chronic Health Evaluation (APACHE) IV score.

Methods

This is a retrospective cohort study of patients treated from January 1, 2008, through January 1, 2018. This single-center study was performed at Mayo Clinic (Rochester, MN), a tertiary care teaching and referral center. The primary outcome was death in the ICU. Discrimination of each risk model for hospital mortality was assessed by comparing area under the receiver operating characteristic curve (AUROC).

Results

The cohort mostly comprised older patients (median age, 64 years) and men (56.7%). The mortality rate of the cohort was 3.5% (2251 of 63,775 patients). The AUROC for mortality prediction was 89.7% (95% CI, 89.5% to 89.9%) for the IMRM, which was significantly greater than the AUROC of 88.2% (95% CI, 87.9% to 88.4%) for APACHE IV (P<.001).

Conclusion

The IMRM was superior to the commonly used APACHE IV score and may be easily integrated into electronic health records at any hospital using Epic software.

SUBMITTER: Tan EM

PROVIDER: S-EPMC7560567 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Validation of a Retrospective Computing Model for Mortality Risk in the Intensive Care Unit.

Tan Eugene M EM Kashyap Rahul R Olson Ian C IC O'Horo John C JC

Mayo Clinic proceedings. Innovations, quality & outcomes 20201006 5

<h4>Objective</h4>To compare the predictive performance of Epic Systems Corporation's proprietary intensive care unit (ICU) mortality risk model (IMRM) with that of the Acute Physiology and Chronic Health Evaluation (APACHE) IV score.<h4>Methods</h4>This is a retrospective cohort study of patients treated from January 1, 2008, through January 1, 2018. This single-center study was performed at Mayo Clinic (Rochester, MN), a tertiary care teaching and referral center. The primary outcome was death ...[more]

PMID: 33083706

Similar Datasets

Project description:Critically ill intensive care unit (ICU) patients commonly develop severe muscle wasting and impaired muscle function, leading to delayed recovery, with subsequent increased morbidity and financial costs, and decrease quality of life of survivors. Acute Quadriplegic Myopathy (AQM) is one of the most common neuromuscular disorders associated with ICU-acquired muscle weakness. Although there are no available treatments for the ICU-acquired muscle weakness, it has been demonstrated that early mobilization can improve its prognosis and functional outcomes. This study aims at improving our understanding of the effects of passive mechanical loading on skeletal muscle structure and function by using a unique experimental rat ICU model allowing analyses of the temporal sequence of changes in mechanically ventilated and pharmacologically paralyzed animals at durations varying from 6 h to 14 days. Results show that passive mechanical loading alleviated the muscle wasting and the loss of force-generation associated with the ICU intervention, resulting in a doubling of the functional capacity of the loaded vs. unloaded muscles after a 2-week ICU intervention. We demonstrated that the improved maintenance of muscle structure and function is likely a consequence of a reduced oxidative stress, and a reduced loss of the molecular motor protein myosin. A complex temporal gene expression pattern, delineated by microarray analysis, was observed with loading-induced changes in transcript levels of sarcomeric proteins, muscle developmental processes, stress response, ECM/cell adhesion proteins and metabolism. Thus, the results from this study show that passive mechanical loading alleviates the severe negative consequences on muscle structure and function associated with mechanical silencing in ICU patients, strongly supporting early and intense physical therapy in immobilized ICU patients. This study aims to unravel the effects of passive mechanical loading on skeletal muscle structure and function in an experimental rat ICU model at duration varying between 6h and 14 days. A total of 23 experimental female Sprague-Dawley rats were included in this study. The experimental rats were anaesthetized, treated with the neuromuscular blocking agent (NMBA) M-NM-1-cobrotoxin, mechanically ventilated and monitored for durations varying from 6h to 4 days (n=13), from 5 to 8 days (n=4), and from 9 to 14 days (n=6). The left leg of the animal was activated for 6 hours at the shortest duration and 12 hours per day at durations 12 hours and longer throughout the experiment, using a mechanical lever arm that produced a continuous passive maximal ankle joint flexions-extensions at a speed of 13.3 cycles per minute. Muscle biopsies were obtained from gastrocnemius muscle (proximal part) immediately after euthanasia, were quickly frozen in liquid propane cooled by liquid nitrogen, and stored at -80M-BM-0C. RNA was extracted.

Project description:BackgroundHeart failure (HF) is a common disease and a major public health problem. HF mortality prediction is critical for developing individualized prevention and treatment plans. However, due to their lack of interpretability, most HF mortality prediction models have not yet reached clinical practice.ObjectiveWe aimed to develop an interpretable model to predict the mortality risk for patients with HF in intensive care units (ICUs) and used the SHapley Additive exPlanation (SHAP) method to explain the extreme gradient boosting (XGBoost) model and explore prognostic factors for HF.MethodsIn this retrospective cohort study, we achieved model development and performance comparison on the eICU Collaborative Research Database (eICU-CRD). We extracted data during the first 24 hours of each ICU admission, and the data set was randomly divided, with 70% used for model training and 30% used for model validation. The prediction performance of the XGBoost model was compared with three other machine learning models by the area under the curve. We used the SHAP method to explain the XGBoost model.ResultsA total of 2798 eligible patients with HF were included in the final cohort for this study. The observed in-hospital mortality of patients with HF was 9.97%. Comparatively, the XGBoost model had the highest predictive performance among four models with an area under the curve (AUC) of 0.824 (95% CI 0.7766-0.8708), whereas support vector machine had the poorest generalization ability (AUC=0.701, 95% CI 0.6433-0.7582). The decision curve showed that the net benefit of the XGBoost model surpassed those of other machine learning models at 10%~28% threshold probabilities. The SHAP method reveals the top 20 predictors of HF according to the importance ranking, and the average of the blood urea nitrogen was recognized as the most important predictor variable.ConclusionsThe interpretable predictive model helps physicians more accurately predict the mortality risk in ICU patients with HF, and therefore, provides better treatment plans and optimal resource allocation for their patients. In addition, the interpretable framework can increase the transparency of the model and facilitate understanding the reliability of the predictive model for the physicians.

Dataset Information

Validation of a Retrospective Computing Model for Mortality Risk in the Intensive Care Unit.

Objective

Methods

Results

Conclusion

Publications

Validation of a Retrospective Computing Model for Mortality Risk in the Intensive Care Unit.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets