Performance of the Grace ACS Risk Score in Predicting in-Hospital and Six-Month Mortality in Patients with ACS: A Single Center Retrospective Cohort Study

Abstract

Introduction: The incidence of coronary heart disease (CHD) is rising globally. Despite the high burden of CHD in Africa, little is known about appropriate risk stratification systems that can be used to predict outcomes in patients with acute coronary syndromes (ACS) in this region. Methodology: This was a single-center retrospective cohort study conducted at the Aga Khan University hospital. Data was explored descriptively by summarizing categorical variables using frequencies/percentages and continuous variables using medians and interquartile ranges. Risk groups and mortality were compared using Fisher’s exact test or the chi-square test. The area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity were obtained from a binary logistic regression model with mortality as the outcome of interest. The model fit was assessed using Hosmer-Lemeshow statistics. Analysis was performed using R version 4.3.2 (2023–10-31 ucrt) and p-value < 0.05 set as statistical significance. Results: 378 participants were enrolled. NSTEMI was diagnosed in 48.4% of participants and STEMI in 51.6%. ACS in-hospital mortality was 2.11% and 6-months mortality was 12.70%. NSTEMI in-hospital mortality was 0.55% (p = 0.45) and 15.92% (p = 0.028) at 6-months. STEMI in-hospital mortality was 3.72% (p < 0.001) and 18.23% (p < 0.001) at 6-months. The AUROC curve for in-hospital mortality was 0.87 and 0.66 for 6-month mortality, demonstrating acceptable discriminatory capacity. Calibration of the score using the Hosmer-Lemeshow model fit was appropriate (p = 0.999). Conclusion: The GRACE 2 score demonstrated good discriminative power and an acceptable goodness-of-fit for in-hospital outcomes in our setting, but was insufficiently accurate for reliably predicting 6-month post-discharge mortality.