New electrocardiographic changes as predictors of acute coronary interventions in the emergency department: a...

Electrocardiograms (EKGs) are integral to the early evaluation of chest pain in the emergency department (ED), and the HEART score is frequently used for risk stratification of acute coronary syndrome (ACS). However, the HEART score typically relies on static EKG assessments, overlooking recent EKG changes that may reveal evolving ischemia. This study investigates the association between serial EKG changes and the need for urgent cardiac interventions, assessing whether integrating serial EKG data can enhance the predictive accuracy of the HEART score.

We conducted a single-center retrospective observational study of ED patients who presented at least twice with chest pain or equivalent symptoms from 2019 to 2023. Eligible patients had serial EKGs and documented HEART scores. New EKG changes were categorized into three groups: (1) new common ischemic patterns (e.g., T-wave inversions, ST-T segment depressions); (2) pseudo-normalization (e.g., resolution of prior abnormalities); and (3) new uncommon ischemic patterns (e.g., QT prolongation, U waves, LBBB, etc.). The primary outcome was the requirement for percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG). Propensity score matching was used to control confounding factors. Inter-rater agreement for EKG classification was assessed using Cohen's kappa.

Among 1,862 patients, 149 (8.0%) underwent PCI/CABG interventions. New EKG changes were observed in nearly 80% of patients who required intervention, compared with less than 30% of those who did not (p < 0.001). The most frequent patterns were T-wave inversions and ST-T depressions. Despite higher HEART scores in the intervention group, 10.7% were misclassified as low risk under traditional scoring.

New EKG changes may be strongly associated with the need for PCI/CABG and provide prognostic value beyond static EKG interpretation. Relying solely on traditional HEART scoring may underestimate risk in a significant subset of patients.