External validation of the GRACE Freedom from Events score

Aim: Acute coronary syndrome (ACS) is a common life-threatening condition but the majority of patients are at low risk of acute adverse events. In 2005, the GRACE Freedom-from-Event score (GFFES) was developed to identify patients with a low risk of adverse in-hospital events. Our aim was to externally validate this score.

Methods: A prospective observational cohort of patients was admitted to a cardiology service with admission diagnoses of chest pain, unstable angina or myocardial infarction (MI). Clinical and investigational data were collected. Defined major adverse cardiac events (MACE) were death, new MI, stroke, acute pulmonary oedema, cardiac arrest or sustained ventricular tachycardia, high degree atrioventricular block, cardiogenic shock, pacemaker or intra-aortic balloon pump insertion, assisted ventilation or new acute renal failure occurring during the index admission. The primary outcome of interest was the predictive performance of the GFFES for MACE, by ROC curve and clinical performance analysis.

Results: 238 patients were studied; median age 67, 56.7% were male. Seventy-eight patients (32.8%) were classified as low risk by the score (GFFES score≥287). There were no MACE in the low risk group. The AUC for predictive performance of the GFFES was 0.74 (95% CI 0.62-0.86). Sensitivity was 100% (95% CI 71.7-100%), specificity 34.7% (95% CI 28.5-41.3%) and negative predictive value 100% (95% CI 94.2-100%).

Conclusion: In this single site prospective validation, GFFES showed good discrimination, sensitivity and negative predictive value. It may be a useful tool for assigning patients to appropriate levels of care based on risk.