Related Experiment Video
Updated: Feb 14, 2026
Author Spotlight: Developing Innovative Therapeutic Strategies for Hemorrhagic Shock Research
Published on: March 22, 2024
Development and evaluation of a prediction model for adult ICU hemorrhage using only continuous cardiorespiratory
Andrew Barros1, Brynne Sullivan2, Matthew T Clark3
1Department of Medicine, University of Virginia, 1215 Lee Street, Charlottesville, VA 22903, United States of America.
Insights
A new model accurately predicts hemorrhage in intensive care unit (ICU) patients using cardiorespiratory data. This tool shows promise for earlier bleeding detection and improved patient outcomes, though performance varies across demographics.
Area of Science:
- Critical Care Medicine
- Biomedical Engineering
- Health Informatics
Background:
- Patients in intensive care units (ICUs) face a high risk of bleeding due to severe illness.
- Early identification of hemorrhage is crucial for improving patient outcomes in critical care settings.
- Existing methods for hemorrhage detection may lack timeliness or generalizability.
Purpose of the Study:
- To develop and evaluate a predictive model for hemorrhage in ICU patients using cardiorespiratory data.
- To assess the generalizability of the hemorrhage detection model across diverse ICU cohorts.
- To compare the performance of the developed hemorrhage model against the shock index.
Main Methods:
- Collected cardiorespiratory monitoring and packed red blood cell administration data from four ICU cohorts (one development, three evaluation).
- Defined hemorrhage as three or more transfusions within 24 hours.
- Trained a penalized logistic regression model to predict hemorrhage within 8 hours and externally validated its performance.
Main Results:
- The model achieved a cross-validated AUC of 0.706 in the development cohort and 0.712 across evaluation cohorts.
- The model demonstrated good calibration (slope 1.041) and predicted increased hemorrhage risk hours before clinical recognition.
- Performance was lower in older patients (>75), at specific institutions (Pitt), and for Black patients and females, though the hemorrhage score outperformed the shock index.
Conclusions:
- A risk model utilizing continuous cardiorespiratory data can predict hemorrhage in ICU patients with clinically relevant accuracy.
- The model demonstrates generalizability across different ICU settings, monitoring devices, and electronic health record systems.
- While generally effective, the model's performance disparities across patient subgroups warrant further investigation and potential refinement.
Abstract:
Objective.Develop and evaluate whether a model trained to detect the physiological signature of hemorrhage in intensive care unit (ICU) patients generalizes to other cohorts.Approach.We collected cardiorespiratory monitoring data and packed red blood cell administration data from consecutive adult admissions in one development and three evaluation ICU cohorts. We defined hemorrhage as three or more transfusions within 24 h. We trained a penalized logistic regression model to predict hemorrhage within 8 h and externally evaluated the predictions.Main results.The evaluation ICU cohorts comprised more than 6M q15 min observations. The cross-validated area under the receiver operating characteristic (AUC) in the development cohort was 0.706 (141 event admissions, 95% confidence intervals (CIs): 0.656-0.757) and 0.712 (968 event admissions, 95% CI: 0.693-0.726) for 17 591 medical and surgical ICU patients in the combination of 3 evaluation cohorts. The calibration slope of the hemorrhage model was close to unity (1.041, 95%CI: 0.956-1.127). Predicted risk increased significantly in the 8 h preceding clinical recognition of bleeding. There was no evidence of model performance drifting over time. There was evidence for lower performance for patients over 75 (20% lower than patients 18-44), among patients at University of Pittsburgh (Pitt) (24% lower than MIMIC III), Black patients (11% lower than White patients), and females (12% lower than males). The shock index also had reduced performance at Pitt, for female patients, and for patients over 75, though not for Black patients. The hemorrhage score had a higher net benefit than the shock index.Significance.Patients in ICUs have an increased risk for bleeding due to their chronic and acute illness, and earlier bleeding identification leads to better outcomes. A risk model for hemorrhage based only on continuous cardiorespiratory data has clinically relevant predictive performance that generalizes across three cohorts with different monitoring devices and electronic health record systems.
Related Concept Videos
02:27Predicting Molecular Geometry
01:15Model Approaches for Pharmacokinetic Data: Physiological Models
01:14Model Approaches for Pharmacokinetic Data: Compartment Models
Two primary types of compartment models are recognized: mammillary and catenary. The more...
01:29Self-Evaluation Maintenance Model
01:03Prediction Intervals
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.
01:06Model Approaches for Pharmacokinetic Data: Distributed Parameter Models
The distributed parameter models are specifically designed to account for variations and differences in some drug classes. This model is particularly useful for assessing regional concentrations of anticancer or...