Established Chest Pain Algorithm Not Up to Snuff in U.S. Study

In some U.S. emergency departments (EDs), the European Society of Cardiology (ESC) 0/1-hour troponin algorithm has shown uniquely poor performance ruling out cardiac events in patients presenting with chest pain, according to an observational study.

With death or myocardial infarction (MI) within 30 days excluded based on this high-sensitivity cardiac troponin (hs-cTn) algorithm, people with a known history of coronary artery disease (CAD) nevertheless experienced this outcome in 3.4% of cases, according to Nicklaus Ashburn, MD, MS, of Wake Forest School of Medicine in Winston-Salem, North Carolina, and colleagues.

"This 3.4% missed 30-day death or MI rate is higher than the 1% miss rate most emergency clinicians are generally willing to accept. Thus, our results suggest that the ESC 0/1-hour hs-cTnT algorithm does not safely rule out ACS [acute coronary syndrome] in patients with known CAD," the authors wrote in JAMA Cardiology.

Moreover, the algorithm's negative predictive value (NPV) for 30-day cardiac death or MI was 96.6% among patients with known CAD and 98.9% without known CAD -- both falling under the accepted 99% threshold.

The ESC algorithm "is likely not safe for routine use among U.S. ED patients with chest pain, regardless of CAD status," Ashburn's group suggested based on their study, a subgroup analysis of the STOP-CP study conducted at eight EDs.

The ESC 0/1-hour hs-cTnT algorithm had been designed and validated to rule in or rule out non-ST-segment elevation MI in multiple international cohorts. The question now is whether its poor performance in the present study is a fluke.

Júlia Karády, MD, PhD, a cardiac imager at Massachusetts General Hospital in Boston, and David Morrow, MD, MPH, director of the Cardiac Intensive Care Unit at Brigham and Women's Hospital in Boston, pointed out the algorithm's strong diagnostic performance in the "preponderance of evidence from other studies, including U.S.-based, that demonstrate an NPV greater than 99% for index MI and 30-day events."

"Evaluated mostly in European and Australian cohorts, the ESC 0/1-hour algorithm has demonstrated excellent diagnostic performance to rule out MI among patients presenting to the ED with chest pain," the duo wrote in an accompanying editorial.

"However, fewer studies have evaluated the 0/1-hour strategy in the U.S., where testing patterns in the ED and regulatory requirements for reporting hs-cTn values differ from elsewhere in the world," Karády and Morrow noted.

The editorialists stressed an integrated approach for risk stratification -- including patient history, electrocardiogram, and serial hs-cTn -- as recommended in the 2021 American Heart Association/American College of Cardiology chest pain guidelines.

STOP-CP had been conducted from 2017 to 2018. The present analysis included people age 21 years and older with suspected ACS (but no ST-segment elevation on initial ECG) who had serial troponins ordered.

This 1,430-person cohort averaged 57.6 years old; 54.2% were men, and 57.8% were white.

Participants with 0- and 1-hour hs-cTnT measures were stratified into rule-out, observation, and rule-in zones using the ESC 0/1-hour hs-cTnT algorithm. The study was observational, and the ESC 0/1-hour hs-cTnT algorithm was not used to guide patient care.

A major limitation of the study was that it was conducted mainly at academic sites, limiting its generalizability nationwide, the researchers said. There was also possible selection bias and misclassification bias related to the investigators relying on medical chart review and self-reporting of pre-existing CAD.

Comment