Study protocol for Rule-Out Acute Myocardial Infarction Using Artificial Intelligence Electrocardiogram 2 (ROMIAE-2) in patients with chest pain: a multicenter randomized noninferiority trial

Q: What is the clinical evidence from this study?

Study design: RCT. Population: Acute chest pain or symptoms suggestive of acute myocardial infarction (n=4670). Intervention: AI-ECG-guided strategy (AiTiAMI Ver 2.00.00 combined with HEART score) vs. Standard care (physician ECG interpretation, hs-cTn algorithms, and HEART score without AI-ECG). Primary outcome: 30-day major adverse cardiovascular events (MACE: composite of all-cause mortality, myocardial infarction, stroke, urgent target-vessel revascularization, and definite or probable stent thrombosis).

Taeyoung Kong; Tae Nyoung Chung; Min Sung Lee; Tae Gun Shin; Kyuseok Kim; on behalf of ROMIAE-2 study group

doi:10.15441/ceem.26.305

October 22, 2026Clinical and Experimental Emergency MedicineOpen Access

Study protocol for Rule-Out Acute Myocardial Infarction Using Artificial Intelligence Electrocardiogram 2 (ROMIAE-2) in patients with chest pain: a multicenter randomized noninferiority trial

View Full Paper

Ask AI

Bookmark

Share

Key result

Trial tests AI-ECG plus HEART score noninferiority for 30-day MACE in ED chest pain.

n=4,670

Why the study?

Although artificial intelligence-based ECG interpretation systems have shown strong diagnostic accuracy for acute myocardial infarction, no multicenter randomized controlled trial had evaluated their clinical impact when integrated into routine emergency care.

Does a real-time AI-ECG-guided strategy (AiTiAMI Ver 2.00.00 combined with the HEART score) safely preserve 30-day MACE outcomes (noninferiority) while improving operational efficiency compared to standard care in adults presenting to the emergency department with undifferentiated acute chest pain?

Key Points

To assess whether real-time artificial intelligence-based electrocardiogram interpretation can be safely integrated into emergency department care to rule out acute myocardial infarction in patients with chest pain.
Prospective, multicenter, randomized, open-label trial with blinded endpoint adjudication conducted across 12 academic tertiary emergency departments in South Korea (N=4,670; ClinicalTrials.gov NCT07027891).
Primary endpoint evaluation is designed to assess 30-day major adverse cardiovascular events (MACE) under a noninferiority margin of 3.0%, powered at 80% with a one-sided significance level of 0.025.
Secondary endpoints will examine operational and diagnostic outcomes, including missed acute myocardial infarction at index visit, emergency department length of stay, early discharge, 7-day revisit, and direct medical costs.

Study Design

Type

RCT (n=4,670)

Blinding

Open-label with blinded endpoint adjudication (PROBE)

Randomization

1:1

Multicenter

Yes

Structured PICO

Does a real-time AI-ECG-guided strategy (AiTiAMI Ver 2.00.00 combined with the HEART score) safely preserve 30-day MACE outcomes (noninferiority) while improving operational efficiency compared to standard care in adults presenting to the emergency department with undifferentiated acute chest pain?

P

Population

Planned enrollment of 4,670 adults (aged ≥19 years) presenting to 12 South Korean emergency departments with acute chest pain or AMI-equivalent symptoms undergoing hs-cTn evaluation, excluding definitive STEMI, followed for 30 days.

I

Intervention

Real-time AI-ECG-guided strategy using a deep learning algorithm (AiTiAMI Ver 2.00.00) to analyze the initial 12-lead digital ECG (10 s, 500 Hz) and generate a continuous AMI probability score (0.0-100.0), combined with the HEART score to classify patients into four integrated risk categories (Low to Very High) for clinical decision-making, alongside standard emergency department care.

C

Comparator

Standard emergency department care without access to AI-ECG results, including physician interpretation of the ECG, 0/1-h or 0/2-h high-sensitivity cardiac troponin (hs-cTn) algorithms at physician discretion, and HEART score calculation (both arms permitted all standard diagnostic and therapeutic interventions, but prohibited from using any other AI-based clinical decision-support tools).

O

Outcome

30-day major adverse cardiovascular events (MACE), defined as a composite of all-cause mortality, myocardial infarction, stroke, urgent target-vessel revascularization, and definite or probable stent thrombosis, assessed using a noninferiority framework with a 3.0% absolute risk difference margin and adjudicated by two emergency medicine specialists blinded to group allocation and AI-ECG results.composite

This study protocol outlines the first multicenter randomized controlled trial designed to evaluate whether integrating real-time AI-ECG interpretation with the HEART score is noninferior to standard care for 30-day MACE while improving operational efficiency in emergency department patients with acute chest pain.

Limitations

Study protocol only; no clinical results or outcome data are available yet
Open-label design may introduce performance bias as treating physicians are aware of group allocation
Conducted exclusively at 12 academic tertiary emergency departments in South Korea, which may limit generalizability to other healthcare settings
Complete blinding is not feasible due to the nature of the intervention (open-label PROBE design)
User and patient satisfaction questionnaires have not been formally validated
No formal cost-effectiveness analysis is planned

Implication

Planned noninferiority trial of AI-ECG plus HEART score in 4670 ED patients; leaves open whether it safely matches standard care for 30-day MACE.

Abstract

Objective Acute myocardial infarction (AMI) requires rapid identification in the emergency department (ED). Although artificial intelligence (AI)-based electrocardiogram (ECG) interpretation systems have shown strong diagnostic accuracy, no multicenter randomized controlled trial (RCT) has evaluated their clinical impact. This protocol describes a RCT for assessing whether real-time AI-ECG can be safely integrated into routine emergency care for undifferentiated chest pain.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Kong et al. (2026) conducted an RCT in Acute chest pain or symptoms suggestive of acute myocardial infarction (n=4,670). AI-ECG-guided strategy (AiTiAMI Ver 2.00.00 combined with HEART score) vs. Standard care (physician ECG interpretation, hs-cTn algorithms, and HEART score without AI-ECG) was evaluated on 30-day major adverse cardiovascular events (MACE: composite of all-cause mortality, myocardial infarction, stroke, urgent target-vessel revascularization, and definite or probable stent thrombosis). AI-ECG-guided care (AiTiAMI Ver 2.00.00 plus HEART score) will be evaluated in 4,670 ED chest pain patients for noninferiority in 30-day MACE (3.0% margin) versus standard care in this trial protocol.

synapsesocial.com/papers/6adc966ca704d151585a5e60 https://doi.org/10.15441/ceem.26.305

View Full Paper

Ask AI

Bookmark

Share

Study protocol for Rule-Out Acute Myocardial Infarction Using Artificial Intelligence Electrocardiogram 2 (ROMIAE-2) in patients with chest pain: a multicenter randomized noninferiority trial

Key Points

Study Design

Structured PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Study protocol for Rule-Out Acute Myocardial Infarction Using Artificial Intelligence Electrocardiogram 2 (ROMIAE-2) in patients with chest pain: a multicenter randomized noninferiority trial

Key Points

Study Design

Structured PICO

Limitations

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading