Discordance and Performance of the ARC-HBR and PRECISE-DAPT High Bleeding Risk Definitions After Coronary Stenting

BACKGROUND: The aim of the ARC-HBR (Academic Research Consortium for High Bleeding Risk) and PRECISE-DAPT (Predicting Bleeding Complications in Patients Undergoing Stent Implantation and Subsequent Dual Antiplatelet Therapy) score definitions for high bleeding risk is to identify patients who would benefit from shorter or less intensive antiplatelet therapy after coronary stenting. OBJECTIVES: The aim of this study was to assess the performance of the ARC-HBR and PRECISE-DAPT score definitions for high bleeding risk in routine clinical practice. METHODS: Using nationwide registers, all patients in Stockholm, Sweden, who were discharged after coronary stenting with dual antiplatelet therapy (January 1, 2013, to July 1, 2018) were included. Patients were categorized as high bleeding risk according to the 2 risk tools, and risk for bleeding (BARC Bleeding Academic Research Consortium types 3-5 or TIMI major or minor) and ischemic events (myocardial infarction or ischemic stroke) within 1 year after discharge was assessed. RESULTS: Of 7,562 patients, the proportions categorized as high bleeding risk were 27% (2,004 of 7,562) using the ARC-HBR definition and 38% (2,894 of 7,562) using the PRECISE-DAPT score; 22% (1,696 of 7,562) had discordant categorization of high bleeding risk comparing the 2 risk tools. Patients with vs without high bleeding risk according to the ARC-HBR definition had higher risk for BARC type 3 to 5 bleeding (1-year risk 7.1% vs 2.3%; HR: 3.21; 95% CI: 2.47-4.17) and ischemic events (7.8% vs 2.8%; HR: 2.96; 95% CI: 2.31-3.79). Patients with vs without high bleeding risk according to the PRECISE-DAPT score had higher risk for TIMI major or minor bleeding (4.4% vs 2.1%; HR: 2.17; 95% CI: 1.63-2.89) and ischemic events (6.2% vs 2.7%; HR: 2.38; 95% CI: 1.85-3.05). The PRECISE-DAPT score underestimated bleeding risk across almost all score levels (median absolute difference between observed and predicted 1-year risk 1.1%; Q1-Q3: 0.8%-1.4%). CONCLUSIONS: There was substantial discordance in the categorization of high bleeding risk between the ARC-HBR definition and the PRECISE-DAPT score. Both tools identified patients at increased bleeding risk, but those patients also had increased ischemic risk. The PRECISE-DAPT score underestimated bleeding risk. Guideline-recommended high bleeding risk definitions may not be generalizable across patient populations, and refined scoring systems are needed.