What are the risk scores used in Acute Coronary Syndrome (ACS)?

Risk Scores Used in Acute Coronary Syndrome

The primary risk scores used in ACS are GRACE, TIMI, HEART, and PURSUIT, with GRACE demonstrating superior discriminative ability and the HEART score being specifically designed for emergency department chest pain evaluation. 1, 2

Primary Risk Scores

GRACE (Global Registry of Acute Coronary Events) Risk Score

• GRACE is the preferred risk score with superior performance compared to subjective physician assessment, demonstrating a C-statistic of 0.83 for predicting in-hospital death across all ACS presentations 2, 3
• Uses 8 variables: age, Killip class, systolic blood pressure, heart rate, ST-segment deviation, cardiac arrest at presentation, serum creatinine, and positive cardiac biomarkers 1, 2
• Predicts outcomes at multiple time points: in-hospital, 6-month, 1-year, and 3-year mortality or death/MI 1
• Patients with GRACE score >140 are considered high-risk and require aggressive therapy including early invasive strategy 2
• Demonstrates pooled C-statistics of 0.82-0.84 at short and long-term follow-up across validation studies 3

TIMI (Thrombolysis in Myocardial Infarction) Risk Score

• Two versions exist: one for NSTEMI/unstable angina and one for STEMI 1
• NSTEMI/Unstable Angina TIMI uses 7 equally-weighted variables: age ≥65 years, ≥3 CAD risk factors, known coronary stenosis ≥50%, ST-segment deviation ≥0.5mm, ≥2 anginal events in prior 24 hours, aspirin use in prior 7 days, and elevated cardiac biomarkers 1
• STEMI TIMI includes: age categories (65-74 years/≥75 years), Killip class II-IV, systolic blood pressure <100 mmHg, heart rate >100 bpm, anterior ST-elevation or LBBB, diabetes/hypertension/angina, weight <67 kg, and time to treatment >4 hours 1
• Predicts 14-day all-cause death, MI, or urgent revascularization for NSTEMI/unstable angina and 30-day all-cause death for STEMI 1
• Shows lower discriminative ability than GRACE with pooled C-statistics of 0.54-0.77 depending on ACS type and timeframe 3

HEART Score

• Specifically designed for emergency department evaluation of chest pain patients without established ACS diagnosis, representing a more modern approach than historical scores 1, 4
• Consists of 5 components: History (0-2 points), ECG (0-2 points), Age (0-2 points with >65 years = 2 points), Risk factors (0-2 points with ≥3 factors = 2 points), and Troponin (0-2 points with >3× normal limit = 2 points) 4
• Low-risk patients (HEART score 0-3) have <1% risk of 30-day MACE and can be safely discharged, with negative predictive value of 98.2% 4, 5
• Intermediate-risk patients score 4-6 points, while high-risk patients score ≥7 points 4
• Demonstrates superior discrimination compared to TIMI in emergency department populations with C-statistic of 0.753 versus 0.678 for TIMI 5

PURSUIT (Platelet glycoprotein IIb/IIIa in Unstable angina: Receptor Suppression Using Integrilin Therapy) Risk Score

• Mentioned as a historical risk score used in NSTEMI/unstable angina populations 1
• Shows moderate predictive accuracy with C-statistic of 0.630 for death or MI at 1 year 6
• Less commonly used in contemporary practice compared to GRACE, TIMI, and HEART 1

Clinical Application Framework

When to Use Each Score

• Use GRACE at hospital admission for all confirmed ACS patients to predict short and long-term outcomes and guide intensity of therapy 1, 2
• Use HEART score in the emergency department for undifferentiated chest pain patients before ACS is confirmed or excluded 1, 4
• Use TIMI as an alternative when GRACE is not feasible, though it has inferior discriminative ability 3, 7

Important Caveats

• Risk scores are NOT diagnostic tools—they stratify risk only in suspected or confirmed ACS 1
• Historical scores (TIMI, GRACE, PURSUIT) are being replaced by more accurate tools like HEART for emergency department evaluation 1
• There is insufficient evidence that routine use of risk scores translates into reduced cardiovascular events, despite their validated prognostic ability 1, 2
• The HEART score may be less accurate in very early presenters where troponin elevation hasn't yet occurred 4
• Risk scores should be used alongside clinical judgment, serial ECGs, and cardiac biomarkers—not as standalone decision-making tools 1