Background: The long-term prognostic implication of post-procedural hematocrit drops in patients undergoing cardiac catheterization outside the clinical trial setting is not well defined.
Methods: Data was prospectively collected from 12,661 patients undergoing diagnostic or interventional cardiac catheterization between July 1998 and July 2006. Patients were divided into three cohorts based upon the degree of hematocrit change: drop greater than 6, drop between 3 and 6, and drop less than 3. In-hospital major adverse events, 30-day mortality, and long-term all-cause mortality were recorded.
Results: Patients with larger reductions in hematocrit were more likely to be older, female, and have a higher baseline hematocrit, present with acute myocardial infarction, develop cardiogenic shock, require emergent catheterization, develop retroperitoneal bleeds and large hematomas, receive transfusions, have longer index hospitalizations, develop subacute stent thrombosis, and have higher 30-day and long-term mortality. An increase in long-term mortality was observed with progressive hematocrit drop. This finding is largely driven by early (30 day) mortality, as trends were no longer significant after rezeroing mortality. Hematocrit drop was not an independent risk factor for 30-day mortality. Transfusion and low baseline hematocrit were identified as independent predictors of near and long-term mortality.
Conclusions: Periprocedural bleeding, defined by hematocrit drop, is associated with increased near-term and long-term mortality in patients undergoing diagnostic and therapeutic cardiac catheterization procedures. Long-term mortality is largely driven by up front 30-day mortality. Hematocrit drop was not an independent predictor for near-term mortality. Transfusion and low baseline hematocrit were independent predictors for near and long-term mortality.