- Utilizing electronic health records of over 48,000 patients, the model uses multinomial logistic regression to develop ICU Admission and Outcomes Prediction Risk Score Model.
- The tool identifies individuals at higher risk for ICU admission, enabling proactive health interventions to improve patient outcomes.
- Supported by the National Institute on Aging, this pioneering study marks a significant advancement in critical care but requires further validation.
Overview
Regenstrief Institute and Indiana University School of Medicine researchers have developed the Prediction Risk Score, a population-level prediction model for intensive care unit (ICU) survivorship and mortality in older adults. This tool, derived from electronic health records, aims to identify patients at risk of being admitted to the ICU and their potential outcomes. It’s essential because ICU survivors often face long-term cognitive and physical impairments.
Prediction Risk Score Model Development
Researchers conducted a population-based cohort analysis to determine the ICU admission and mortality rates within two years after the primary care visit date. The study involved 48,127 patients who were 50 years old and above. They used data from electronic health records (EHR) to identify variables that could predict ICU survivorship. The study employed multinomial logistic regression and cross-validation techniques to identify and validate predictive variables for distinguishing between patient outcomes. Key risk factors identified included age, chronic illnesses, and specific laboratory abnormalities.
Enhancing patient care
The Prediction Risk Score is designed for use by health systems and researchers. It helps identify adults at higher risk of ICU admission and their likelihood of survival. This can lead to proactive healthcare measures and the development of new care models to improve patient outcomes.
The study, supported by the National Institute on Aging of the National Institutes of Health, is a crucial step towards improving care for critically ill patients and is published in the peer-reviewed journal Health Science Reports. The developed risk scores are a practical tool for identifying patients at increased ICU admission and mortality risk. However, further validation in diverse patient populations is necessary.
Reference
Khan, Sikandar H., Anthony J. Perkins, Mikita Fuchita, Emma Holler, Damaris Ortiz, Malaz Boustani, Babar A. Khan, and Sujuan Gao. 2023. “Development of a Population-Level Prediction Model for Intensive Care Unit (ICU) Survivorship and Mortality in Older Adults: A Population-Based Cohort Study.” Health Science Reports 6 (10): e1634. https://doi.org/10.1002/hsr2.1634
You must be logged in to post a comment.