Predictive accuracy of CHA2DS2-VASc and HAS-BLED scores in patients without atrial fibrillation undergoing percutaneous coronary intervention and discharged on dual antiplatelet therapy

Background The CHA2DS2-VASc and HAS-BLED are well-validated stroke risk prediction scores for atrial fibrillation (AF), but their role in risk stratification of major adverse cardiac events (MACEs) and major bleeding for non-AF patients undergoing percutaneous coronary intervention (PCI) is unknown. Methods Consecutive patients without AF undergoing PCI at two Italian centers were analyzed with different measures of discrimination, net reclassification improvement and net prognostic benefit. MACE included all-cause death, myocardial infarction, destabilizing symptoms leading to hospitalization, and nonfatal stroke. Major bleeding was defined according to the TIMI classification. Results Of 1437 subjects undergoing PCI, 1330 (mean age 63.6 ± 10.9 years, 75.7% male) fulfilled the inclusion criteria. During 2.7 ± 1.2 years of follow-up representing 3539 patient-years at risk, 187 patients had a MACE (5.3%/year) and 48 had a major bleeding (1.4%/year). The cumulative incidences of MACE were significantly stratified by both high CHA2DS2-VASc (P = 0.020) or HAS-BLED (P < 0.001) scores, whereas major bleeding episodes were not. The CHA2DS2-VASc and the HAS-BLED scores had similar C-statistics for MACE (0.56 vs 0.60; P = 0.52) and major bleeding (0.63 vs 0.60; P = 0.63). Compared with CHA2DS2-VASc, the HAS-BLED score more accurately reclassified events and no events both for MACE (NRI 8.21%) and major bleeding (NRI 6.85%). Conclusions In patients without AF undergoing PCI and discharged on dual antiplatelet therapy, the HAS-BLED score performed better than the CHA2DS2-VASc for the prediction of MACE. Although both scores predict MACE, their discrimination was modest. Conversely, both scores did not significantly predict major bleeding in non-AF patients undergoing PCI.