Performance of EDACS, INTERCHEST, SCARE, and HEART Scores in Predicting Major Adverse Cardiac Events: A Prospective Multicenter Cohort Study
摘要
Early diagnosis of acute coronary syndrome is critical for timely intervention. While risk scores like HEART are well-validated, they require biomarkers and ECG, potentially delaying triage decisions. History-based scores (EDACS, INTERCHEST, SCARE) enable immediate assessment, but their comparative ED performance remains unclear. This study evaluated their diagnostic accuracy versus HEART for predicting 1-month major adverse cardiac events (MACE) in ED chest pain patients.
MethodsThis prospective multicenter observational study enrolled adults with chest pain presenting to EDs between January and August 2025. Clinical data were collected at presentation to calculate EDACS, INTERCHEST, SCARE, and HEART scores without influencing clinical management. The primary outcome was 1-month MACE (acute coronary syndrome, revascularization, or death). Predictive accuracy was assessed using ROC analysis and DeLong pairwise comparisons. Independent MACE predictors were identified through multivariable logistic regression.
ResultsAmong 335 patients (mean age 54.5 ± 15.2 years; 58.5% male), MACE occurred in 51 (15.2%). AUC values were 0.864 (HEART), 0.829 (EDACS), 0.801 (INTERCHEST), and 0.785 (SCARE). EDACS and INTERCHEST performed comparably to HEART, while SCARE showed lower performance. EDACS demonstrated highest sensitivity (88.2%) and NPV (96.8%); INTERCHEST had highest specificity (85.9%). Independent predictors included age, diaphoresis, exertional pain, radiation patterns, clinician concern, and patient-perceived cardiac etiology.
ConclusionEDACS and INTERCHEST scores achieved diagnostic performance comparable to HEART score, while SCARE score demonstrated lower accuracy. History-based risk scores enable immediate risk stratification without laboratory testing, potentially informing triage decisions regarding low-risk patients and high-risk patients requiring closer observation while optimizing resource allocation.