Risk Stratification in Myocardial Infarction: Critical Role of TIMI, Killip, and GRACE Scores
These three scoring systems are essential tools that predict mortality and guide treatment intensity in MI patients, with GRACE demonstrating superior discriminative ability (C-statistic 0.83) for predicting in-hospital and long-term death across all ACS presentations, while TIMI provides rapid bedside risk assessment, and Killip class—incorporated within GRACE—captures hemodynamic instability that independently predicts outcomes. 1, 2
GRACE Score: The Most Comprehensive Risk Predictor
The GRACE risk model should be calculated at admission for all MI patients to predict in-hospital, 6-month, 1-year, and 3-year mortality or death/MI, as it outperforms subjective physician assessment. 2
Components and Predictive Power
GRACE uses 8 variables that capture both hemodynamic instability and underlying disease severity: older age (OR 1.7 per 10 years), Killip class (OR 2.0 per class), systolic blood pressure (OR 1.4 per 20 mm Hg decrease), ST-segment deviation (OR 2.4), cardiac arrest during presentation (OR 4.3), serum creatinine level (OR 1.2 per 1-mg/dL increase), positive initial cardiac biomarkers (OR 1.6), and heart rate (OR 1.3 per 30-beat/min increase). 1, 2
The model was developed from 11,389 patients and validated in subsequent GRACE and GUSTO IIb cohorts, predicting in-hospital death in patients with STEMI, NSTEMI, or unstable angina with excellent discrimination (C-statistic 0.83). 1
GRACE demonstrates superior long-term prognostic accuracy compared to TIMI, particularly for predicting 3-year mortality in both NSTEMI (p=0.035) and STEMI (p=0.009). 3
Clinical Application Thresholds
Patients with GRACE score >140 are considered high-risk and require aggressive therapy, including early invasive strategy, intensive antiplatelet therapy with glycoprotein IIb/IIIa inhibitors, and close monitoring. 2
The GRACE clinical application tool can be downloaded to a handheld device for bedside use at www.outcomes-umassmed.org/grace. 1
TIMI Risk Score: Rapid Bedside Assessment
The TIMI risk score provides immediate risk stratification using 7 equally-weighted variables available at presentation, with event rates (death, MI, or urgent revascularization within 14 days) ranging from 4.7% for scores 0-1 to 40.9% for scores 6-7. 1, 4
The 7 Components (1 point each)
- Age ≥65 years 4
- ≥3 risk factors for coronary artery disease (hypertension, diabetes, hyperlipidemia, family history, current smoking) 4
- Known coronary stenosis ≥50% on prior catheterization 4
- ST-segment deviation ≥0.5 mm on presenting ECG 4
- ≥2 anginal events in prior 24 hours 1
- Aspirin use in prior 7 days 4
- Elevated cardiac biomarkers (troponin or CK-MB) 4
Risk-Based Treatment Decisions
High TIMI risk scores (≥4) mandate early invasive strategy with coronary angiography and revascularization, intensive antiplatelet therapy with glycoprotein IIb/IIIa inhibitors, and low-molecular-weight heparin over unfractionated heparin. 4
Low TIMI risk scores (0-2, representing 4.7-8.3% event rate) support conservative management with medical therapy and stress testing before discharge, though this should not override clinical judgment when high-risk features are present. 1, 4
The TIMI calculator is available at www.timi.org for verification, though the score can be calculated immediately at bedside without computer assistance. 1, 4
Killip Classification: Hemodynamic Assessment
Killip class is a critical component of the GRACE score and independently predicts mortality, with 1-year mortality rates of 6%, 24%, 42%, and 60% for Killip classes 1-4, respectively. 5
Integration with Other Scores
Applying the TIMI score to patients classified as Killip 1 further stratifies them into low-, medium-, and high-risk subgroups with 1%, 8%, and 19% 1-year mortality rates, significantly improving stratification beyond Killip classification alone. 5
The Killip classification is simpler and less time-consuming compared to the TIMI score, making it valuable for immediate bedside assessment before calculating more comprehensive scores. 5
Within the GRACE model, Killip class carries an odds ratio of 2.0 per class, making it one of the most powerful predictors of adverse outcomes. 1
Comparative Performance and Clinical Decision-Making
Head-to-Head Comparisons
All three risk scores (TIMI, GRACE, and PURSUIT) demonstrate good predictive accuracy for death and MI at 1 year, identifying patients who benefit most from aggressive therapy including early myocardial revascularization. 1
GRACE outperforms TIMI in predicting long-term mortality, with significantly better discrimination for 3-year outcomes in both STEMI and NSTEMI populations. 3
In STEMI specifically, while TIMI and GRACE show similar discriminatory capacity for hospital death (C-statistics both 0.87), TIMI demonstrates better calibration with more appropriate distribution across risk categories. 6
Practical Implementation Strategy
Calculate GRACE score at admission for all confirmed MI patients to guide overall treatment intensity and predict both short-term and long-term outcomes. 2
Use TIMI score for rapid bedside risk assessment, particularly when immediate treatment decisions are needed before complete GRACE calculation. 4
Document Killip class on presentation, as it provides immediate prognostic information and is incorporated into GRACE scoring. 5
High-risk patients identified by any score (GRACE >140, TIMI ≥4, or Killip ≥2) benefit most from myocardial revascularization performed during initial hospital stay. 2, 4, 7
Critical Caveats
Risk scores stratify risk only in suspected or confirmed ACS and should be used alongside clinical judgment, serial ECGs, and cardiac biomarkers—they are not diagnostic tools. 2
The predictive ability of these commonly used risk assessment scores for nonfatal coronary heart disease risk is only moderate, emphasizing the need to focus on mortality and major adverse cardiac events as primary outcomes. 1
Approximately 10-17% of patients with suspected ACS present with normal ECGs yet still experience major adverse cardiac events, highlighting that normal ECG findings do not eliminate risk when other high-risk features are present. 2
Renal impairment represents an additional high-risk feature not fully captured by standard scores, with severe renal dysfunction associated with severely increased short- and long-term mortality risks. 1