Predictive power of the grace score in population with diabetes

Area of Science:

• Cardiology
• Clinical Risk Stratification
• Diabetes Mellitus Research

Background:

• Clinical guidelines advocate risk stratification for acute coronary syndrome (ACS) patients upon hospital admission.
• Diabetes mellitus (DM) is a known independent predictor of mortality in ACS patients but is not currently incorporated into the GRACE risk score.

Purpose of the Study:

• To validate the GRACE risk score in a contemporary ACS population.
• To specifically evaluate the GRACE score's performance in diabetic patients.
• To assess the impact of including diabetes mellitus as a variable in the GRACE risk score model.

Main Methods:

• Retrospective cohort study utilizing the ARIAM-SEMICYUC registry.
• Analysis of in-hospital mortality data for ACS patients.
• Evaluation of GRACE score predictive power using ROC curve analysis and calibration assessment.
• Assessment of the effect of DM inclusion via net reclassification improvement.

Main Results:

• The GRACE score demonstrated good predictive power for hospital mortality in the study cohort.
• Moderate calibration was observed for the GRACE score, with no significant variations across ACS types or DM status.
• Incorporating diabetes mellitus as a variable did not enhance the predictive value of the GRACE model.

Conclusions:

• The GRACE risk score exhibits appropriate predictive power and good calibration for ACS patients.
• The GRACE score is clinically applicable to the subgroup of patients with diabetes mellitus.
• Adding diabetes mellitus to the GRACE score did not provide additional predictive benefit.