Correlates of poor outcome among patients with bleeding after coronary interventions.

OBJECTIVES

The objective of this study was to identify factors associated with the increased risk of mortality among patients with bleeding complications during percutaneous coronary intervention (PCI).

METHODS

The study included 1510 patients with periprocedural bleeding complications of the 14 180 patients undergoing PCI. Bleeding was defined and scaled according to the Bleeding Academic Research Consortium criteria. The main outcome was all-cause mortality at 1 year after PCI.

RESULTS

Of the 1510 patients with bleeding within the first 30 days of PCI, 104 patients (6.9%) died during the first year after PCI. One-year survival according to the Bleeding Academic Research Consortium classes was as follows: 96.4% (22 deaths) in class 1, 94.7% (12 deaths) in class 2, 91.0% (42 deaths) in class 3a, 85.4% (23 deaths) in class 3b, and 58.3% (five deaths) in class 3c (P<0.001). The Cox proportional hazards model identified bleeding severity [hazard ratio (HR)=1.55, 95% confidence interval 1.25-1.93 for one class increase in bleeding scale, P<0.001], bleeding site [HR=1.79 (1.18-2.71) for nonaccess vs. access site, P=0.006], elevated troponin levels [HR=1.66 (1.09-2.53), P=0.018], multivessel disease [HR=2.15 (1.01-4.56), P=0.047], C-reactive protein [HR=1.04 (1.01-1.06) for 5 mg/l increase, P=0.002], anemia [HR=2.27 (1.39-3.71) for 4 g/dl decrease in hemoglobin, P=0.001], and platelet count [HR=1.11 (1.02-1.21) for 50×10/l increase in platelet count, P=0.013] as independent associates of mortality. Age (P=0.072) and reduced estimated creatinine clearance (P=0.066) were close to reaching statistical significance as associates of mortality.

CONCLUSION

Among patients with peri-PCI bleeding, the characteristics of bleeding itself and the cardiovascular risk profile underlie the increased risk for poor outcome after PCI.