PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
September 13, 2016International Journal of Environmental Research and Public HealthOpen Access

Utilizing Chinese Admission Records for MACE Prediction of Acute Coronary Syndrome

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Does a machine learning model utilizing unstructured admission records improve MACE prediction compared to GRACE and TIMI scores in patients with ACS?

Population

2,930 patients with acute coronary syndrome (ACS) from a Chinese hospital

Comparison

Machine learning MACE prediction models… vs GRACE and TIMI risk score tools

Design

Cohort

Key result

A machine learning model utilizing unstructured admission records achieved an AUC of 72% for MACE prediction in ACS patients, outperforming GRACE and TIMI risk scores.

Authors

DHDanqing HuZHZhengxing HuangTCTak-Ming Chan

Discussion

Loading...

Member takes

Overview

May aid early ACS risk stratification from admission notes; extends ML evidence beyond GRACE/TIMI but leaves open prospective validation.

Study Design

Type

Observational (n=2,930)

Multicenter

No

Structured PICO

Does a machine learning model utilizing unstructured admission records improve MACE prediction compared to GRACE and TIMI scores in patients with ACS?

P
Population
2,930 patients with acute coronary syndrome (ACS) from a Chinese hospital
I
Intervention
Machine learning MACE prediction models utilizing unstructured admission records (hybrid approach with rule-based NLP and Conditional Random Fields)
C
Comparator
GRACE and TIMI risk score tools
O
Outcome
Major adverse cardiovascular event (MACE) prediction performance (Area Under the Curve [AUC])composite

Main Result

Effect estimate: AUC 72%

Machine learning models utilizing unstructured admission records can effectively predict MACE in ACS patients early in their hospitalization, outperforming traditional GRACE and TIMI risk scores.

Cite This Study

Hu et al. (2016) conducted an observational in Acute coronary syndrome (n=2,930). Machine learning models using admission records vs. GRACE and TIMI risk scores was evaluated on MACE prediction (AUC 72%). A machine learning model utilizing unstructured admission records achieved an AUC of 72% for MACE prediction in ACS patients, outperforming GRACE and TIMI risk scores.

synapsesocial.com/papers/6a08f464a2bc65e38873a81fhttps://doi.org/10.3390/ijerph13090912
View Full Paper
Ask AI
Bookmark
Share