Should Hemoglobin Drop Be Added to Bleeding Classifications in ACS ?

New potent antithrombotic therapies and rapid involvement in revascularization strategies have led to a significant reduction in MI and death in acute coronary syndrome (ACS) patients. Along with this reduced rate of ischemic complications, patient exposure to hemorrhage risk has increased over the past decades. Tremendous efforts have been made to develop relevant bleeding scales to better assess their adverse prognostic impact. However, cross-comparisons between randomized clinical trials are difficult due to the use of different classifications with different biological thresholds. The magnitude of in-hospital hemoglobin (Hb) drop is a key variable of all bleeding scales in ACS, which correlates with cardiovascular outcomes. However, cut-off values and denomination vary widely; a fall in Hb level of 2 g/dl would be a major bleeding according the International Society on Thrombosis and Haemostasis definitions when a 3- to-5 g/dl drop would be a minor bleeding according to the TIMI (Thrombolysis In Myocardial Infarction) definitions. In the absence of overt bleeding, Hb drop is almost ubiquitous in ACS patients admitted in intensive coronary care unit (ICCU), occurring in roughly 88% of nonbleeders with an average 1.27-g/dl Hb decrease. In ACS, anemia either on admission or during hospitalization is strongly associated with mortality. Baseline anemia (Hb <11 g/dl) is a major criteria in the Academic Research Consortium for High Bleeding Risk consensus definition. But little is known about the prognostic impact of Hb drop per se. To note, the recent Bleeding Academic Research Consortium (BARC) classification that was built on consensus by merging both biological and clinical variables from the historical TIMI and GUSTO (Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries) scales in an attempt to homogenize bleeding definitions, does not take into account the drop of Hb. The determinants of the worsened prognosis associated with Hb drop per se are difficult to identify. We can assume that physician (over)reacting may, most of the time, withdraw 1 or several antiplatelet agents and use more often than needed packed red blood cell transfusion. We may also accept that some of these Hb drops just reflect active unrecognized bleedings (gastrointestinal in particular) that are pre-existing undiagnosed comorbidities.

[...]