Evaluating the performance of the GRACE and TIMI risk scores in acute coronary syndromes: a South African cohort

Khiroya, Mitesh Satish

Evaluating the performance of the GRACE and TIMI risk scores in acute coronary syndromes: a South African cohort

Thesis / Dissertation

2024

Publisher

University of Cape Town

Department

Department of Medicine

Faculty

Faculty of Health Sciences

Abstract

Introduction: The GRACE and TIMI scores are validated risk stratification tools that accurately predict risk of in-hospital, 30-day, and one-year major adverse cardiac events (MACE) in patients with Acute Coronary Syndromes (ACS). The performance of GRACE and TIMI scores in a setting where most ST-elevation myocardial infarction (STEMI) patients receive thrombolytic reperfusion therapy after 6 hours and a considerable proportion of non-ST elevation myocardial infarction (NTSEMI) patients receive delayed angiography and revascularisation after 48 hours, is unknown. Objective: To evaluate the accuracy of GRACE and TIMI risk scores in predicting in-hospital and 30-day mortality in a population characterised by a significant prevalence of delayed ACS presentation, limited access to primary percutaneous coronary intervention (PPCI) and delayed revascularisation. Methods: We conducted a retrospective review of all patients admitted to the coronary care unit (CCU) at Groote Schuur Hospital, Cape Town, with either STEMI or NSTEMI, between January 1 st to December 31st, 2019. For each participant, both GRACE and TIMI risk scores were calculated and recorded electronically. Performance of each score was determined and compared using receiver operating characteristic curve (ROC) analysis. Results: Of 329 participants with ACS, 58.6% presented with STEMI and 41.4% with NSTEMI. Mean age was 61.3 (SD±11.9) years, and 59.6% were male. Mean time from symptom onset to hospital admission was 18.3 (SD ± 37.4) hours, with only 4 participants (2.1%) receiving PPCI. STEMI in-hospital and 30-day mortality was 4.1% and 4.2%, respectively, whereas in-hospital mortality for NSTEMI was 1.5%. In the STEMI cohort, both GRACE and TIMI risk scores were comparable, showed excellent discrimination for in-hospital mortality (AUC=0.927, 95% CI: 0.83- 1.00 versus AUC=0.923, 95% CI: 0.87-0.98; p 0.91), and demonstrated modest accuracy for predicting 30-day mortality (GRACE AUC=0.587, 95% CI: 0.29-0.88; TIMI AUC=0.530, 95% CI: 0.12-0.94; p 0.44). In the NSTEMI cohort, GRACE performed significantly better than TIMI (AUC=0.905, 95% CI: 0.85-0.96 versus AUC=0.278, 95% CI: 0.00-0.68; p 0.001) for predicting in-hospital mortality. Conclusion: Both GRACE and TIMI scores demonstrated high accuracy in predicting in-hospital mortality and their predictive accuracy was modest when predicting 30-day mortality for STEMI patients. In addition, GRACE outperformed the TIMI score in assessing NSTEMI in-hospital mortality. Further research in low-and middle-income countries in SSA is needed to evaluate the potential impact of these scores on treatment strategies and cardiovascular outcomes.