GRACE ACS Risk Score for Mortality Prediction

About

The GRACE risk score was developed from the Global Registry of Acute Coronary Events, a multinational observational study that enrolled over 100,000 patients across 30 countries in Europe, North America, South America, Australia, and Asia. The score was first published in 2003 by Granger et al. in the European Heart Journal, with the primary objective of creating a practical, clinically applicable tool for bedside risk stratification across the entire spectrum of ACS. Unlike some earlier risk scores derived from clinical trial populations with strict inclusion and exclusion criteria, the GRACE score was developed from a real-world registry that reflects the broad diversity of ACS patients encountered in everyday clinical practice, including those with comorbid conditions, advanced age, and complications. The score incorporates eight key variables readily available at initial assessment: age, heart rate, systolic blood pressure, serum creatinine, cardiac arrest at admission, ST-segment deviation, elevated cardiac enzymes, and Killip class. Each variable contributes weighted points based on logistic regression coefficients from the original derivation cohort. The total score ranges from 1 to 372, with higher scores indicating greater mortality risk. The GRACE score has been extensively validated across multiple populations and healthcare settings, demonstrating excellent discrimination (c-statistics consistently >0.80) and calibration in diverse ACS populations. Its evidence level is Grade A, supported by the original derivation in a >100,000 patient registry and multiple large-scale validation studies across geographic regions. The ESC guidelines for both STEMI and NSTE-ACS give the GRACE score a Class I recommendation for initial risk stratification, and it is now mandated for use in many national ACS quality improvement programs and registries worldwide.

Formula

GRACE Score = Age Points + Heart Rate Points + Systolic BP Points + Creatinine Points + Cardiac Arrest (39) + ST Deviation (28) + Elevated Enzymes (14) + Killip Class Points

The GRACE score is calculated by summing weighted points for eight variables. Age is categorized by decade (<30: 0, 30-39: 8, 40-49: 25, 50-59: 41, 60-69: 58, 70-79: 75, 80-89: 91, ≥90: 100). Heart rate points increase with tachycardia (1 point for 50-69 bpm, up to 38 for ≥200 bpm). Systolic BP points are inversely related (0 for ≥200 mmHg, up to 58 for <80 mmHg). Creatinine points increase with renal impairment (3 for 0.4-0.79 mg/dL, up to 28 for ≥4 mg/dL). Cardiac arrest at presentation adds 39 points. ST-segment deviation adds 28 points. Elevated cardiac enzymes add 14 points. Killip class is scored as Class I: 0, Class II: 20, Class III: 39, Class IV: 59 points. Total score ranges from 1 to 372. Patients are stratified into four risk categories: low (≤108, in-hospital mortality <2%), intermediate (109-140, 3-8%), high (141-168, 9-20%), and very high (>168, >20%). Predicted 6-month mortality follows similar patterns. The score maintains excellent discrimination across the spectrum of ACS presentations.

Score Interpretation

The GRACE score is the most extensively validated risk score for ACS mortality prediction and carries a Class I recommendation from the European Society of Cardiology for initial risk stratification in both STEMI and NSTE-ACS. Unlike the TIMI scores which were derived from clinical trial populations, GRACE was developed from a real-world registry of over 100,000 patients, making it more generalizable to routine clinical practice. Its primary clinical impact is in guiding the urgency of invasive management. High-risk patients (GRACE >140) derive the greatest benefit from early revascularization, while low-risk patients may be managed more conservatively without compromising outcomes. The GRACE score also enables objective communication of prognosis to patients and families, facilitates appropriate triage to intensive care versus ward-based care, and serves as a standardized tool for quality benchmarking across institutions. Its utility has been demonstrated across diverse healthcare systems worldwide, with consistent performance regardless of geographic region, ethnicity, or healthcare delivery model.

Reference Ranges