Related Experiment Video
Updated: Jan 9, 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
7.5K
Machine learning for risk stratification in the emergency department (MARS-ED): a randomized controlled trial.
Paul M E L van Dam1, William P T M van Doorn2, Lotte Sevenich3
1Department of Internal Medicine, Division of General Internal Medicine, Section Acute Medicine, Maastricht University Medical Center +, Maastricht, Netherlands. paul.van.dam@mumc.nl.
Nature Communications
|December 1, 2025
Summary
A new machine learning tool, RISK INDEX, accurately predicts 31-day mortality using lab values. However, its prognostic accuracy did not translate into improved clinical decisions or outcomes in the emergency department.
Area of Science:
- Emergency medicine
- Clinical informatics
- Health services research
Background:
- Emergency department (ED) crowding requires efficient risk stratification.
- Existing tools (NEWS, APACHE II, SOFA) have limitations in generalizability and data requirements.
- Machine learning offers potential for improved predictive accuracy using routine data.
Purpose of the Study:
- To develop and evaluate the RISK INDEX, a machine learning tool predicting 31-day mortality using laboratory values, age, and sex.
- To assess the clinical impact and prognostic accuracy of the RISK INDEX compared to standard care and traditional scoring systems.
- To determine if enhanced prognostic accuracy translates into improved clinical decision-making and patient outcomes in the ED.
Main Methods:
- An investigator-initiated, open-label, randomized non-inferiority trial (MARS-ED) was conducted at a university medical center ED.
- Adult patients (≥18 years) with ≥4 laboratory tests were randomized (1:1) to standard care or standard care plus access to the RISK INDEX.
- Primary outcomes included prognostic accuracy for 31-day mortality and clinical impact, assessed via area under the receiver operating characteristic curve (AUROC) and treatment plan changes.
Main Results:
- The RISK INDEX demonstrated superior prognostic accuracy for 31-day mortality compared to clinical intuition and traditional scores (AUROC 0.84 vs. 0.65-0.76).
- RISK INDEX predictions diverged from clinician expectations in approximately half of cases, particularly among less experienced physicians.
- Despite high accuracy, the RISK INDEX did not significantly alter treatment plans (0.16% changes) or clinical outcomes, with perceived low added value by clinicians.
Conclusions:
- Prognostic accuracy alone is insufficient to ensure clinical impact in the emergency department setting.
- User-centered design, actionability, and trust are crucial for the successful integration of machine learning tools into clinical workflows.
- Future development should focus on creating clinically actionable insights rather than solely on predictive performance.
Related Concept Videos
01:13Randomized Experiments
8.8K
The randomization process involves assigning study participants randomly to experimental or control groups based on their probability of being equally assigned. Randomization is meant to eliminate selection bias and balance known and unknown confounding factors so that the control group is similar to the treatment group as much as possible. A computer program and a random number generator can be used to assign participants to groups in a way that minimizes bias.
Simple randomization
Simple...
Simple randomization
Simple...
8.8K
01:12Hazard Ratio
549
The hazard ratio (HR) is a widely used measure in clinical trials to compare the risk of events, such as death or disease recurrence, between two groups over time. It reflects the ratio of hazard rates—the instantaneous risk of the event occurring—between a treatment group and a control group. This measure provides valuable insights into the relative effectiveness of a treatment by assessing how the risk of an event differs between the two groups.
For example, in a clinical trial...
For example, in a clinical trial...
549
01:20Comparing the Survival Analysis of Two or More Groups
538
Survival analysis is a cornerstone of medical research, used to evaluate the time until an event of interest occurs, such as death, disease recurrence, or recovery. Unlike standard statistical methods, survival analysis is particularly adept at handling censored data—instances where the event has not occurred for some participants by the end of the study or remains unobserved. To address these unique challenges, specialized techniques like the Kaplan-Meier estimator, log-rank test, and...
538
01:25Statistical Methods for Analyzing Epidemiological Data
875
Epidemiological data primarily involves information on specific populations' occurrence, distribution, and determinants of health and diseases. This data is crucial for understanding disease patterns and impacts, aiding public health decision-making and disease prevention strategies. The analysis of epidemiological data employs various statistical methods to interpret health-related data effectively. Here are some commonly used methods:
875
01:23Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
382
Biopharmaceutical studies constitute a vital field aiming to enhance drug delivery methods and refine therapeutic approaches, drawing upon diverse interdisciplinary knowledge. In research methodologies, the choice between controlled and non-controlled studies significantly influences the study's reliability and accuracy.
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...
382