Access site hematoma requiring blood transfusion predicts mortality in patients undergoing percutaneous coronary intervention: Data from the National Heart, Lung, and Blood Institute Dynamic Registry

Publisher Website
Google Scholar
Abstract
Objective: To determine both the etiology of and outcomes associated with access site hematoma requiring transfusion (HRT) in patients undergoing percutaneous coronary intervention (PCI). Background: Access site hematoma in the setting of PCI is the most frequent periprocedural complication (2–12%). Antiplatelet and antithrombin therapy is designed to lower the incidence of adverse ischemic events while maintaining an acceptable rate of hemorrhagic complications. Methods: This was a prospective, multi‐center, cohort study of consecutive patients undergoing PCI during 3 NHLBI Dynamic Registry recruitment waves (1997–2002). The primary endpoints included the incidence of HRT, in‐hospital death, and death at 1‐year. Results: The incidence of HRT was 1.8% and femoral access was common. Older age, lower BMI, female sex, concomitant renal, cerebrovascular, peripheral vascular, and pulmonary disease were significantly associated with HRT. Glycoprotein IIb/IIIa inhibitors, thrombolytic therapy, and postprocedure heparin were more commonly used in HRT patients, but there was no difference in thienopiridiene use. Attempted lesions in patients developing HRT were more often calcified, thrombotic, located in an ostial location, or class B2 or C. In‐hospital mortality and 1‐year death rate was 9 and 4.5 times higher in HRT patients respectively. Following adjustment, HRT remained independently associated with in‐hospital mortality (OR 3.59, 95% CI 1.66–7.77) and 1‐year death (hazard ratio [HR] 1.65, 95% CI 1.01–2.70, P = 0.048). Independent predictors of HRT included age, female sex, IIb/IIIa inhibitors, thrombolytic agents, and concomitant conditions. Conclusions: Access site complications, especially HRT, remain a very important predictor of adverse procedural success and patient outcome.