Related Experiment Video
Updated: Jun 2, 2026
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Development of a Predictive Model for Identifying High-Risk Older Adults for Geriatric Emergency Department Screening
Karen A Hauser1, Jeremy Swartzberg2, David Schlessinger3
1Department of Hospital Medicine, Kaiser Permanente San Francisco Medical Center, San Francisco, California, USA.
Summary
A new Geriatric Emergency Department (GED) screening score (SS) effectively identifies older adults at risk of hospital use and mortality. This automated tool shows improved accuracy over existing methods, aiding better patient care in emergency settings.
Area of Science:
- Geriatric Emergency Medicine
- Health Services Research
- Clinical Informatics
Background:
- Geriatric emergency department (GED) programs struggle with accurate risk stratification for older adults.
- Existing tools like the Identification of Seniors At Risk (ISAR) score have limitations in predictive accuracy and sustainable implementation.
- There is a need for improved screening tools to identify high-risk older adults in the emergency department (ED).
Purpose of the Study:
- To develop, evaluate, and validate a novel screening score (GED-SS) for older adults in emergency departments.
- To create an automated tool deployable at ED triage to predict risk of subsequent acute care utilization and short-term mortality.
- To compare the performance of the GED-SS model against the existing ISAR score.
Main Methods:
- Developed the GED-SS score model using a multicenter cohort of ED patients aged ≥70 years (January 2018-December 2019).
- Defined the composite outcome as ≥3 days of acute care (ED, observation, inpatient) or death within 90 days.
- Prospectively validated the GED-SS model in 1313 ED patients and compared it with nurse-performed ISAR score screenings.
Main Results:
- The GED-SS score model demonstrated superior discrimination with an AUC of 0.73 compared to 0.66 for the ISAR score model.
- At a 43% sensitivity threshold, GED-SS showed higher specificity (86% vs 78%) and positive predictive value (59% vs 47%) than ISAR.
- The GED-SS model flagged fewer patients (23.3% vs 29.1%) while maintaining better predictive performance.
Conclusions:
- A validated GED-SS score model effectively identifies older adult ED patients at high risk for short-term mortality and acute care utilization.
- The GED-SS model, utilizing structured data for automated calculation, offers improved performance over the ISAR score.
- This automated screening tool can enhance risk stratification within integrated healthcare systems.
