When to Resist the Temptation to Transfuse

Is it always in the patient's best interests to target a hemoglobin level of 10 g/dL?

Here is a common clinical scenario. Your patient was admitted to the hospital with an acute myocardial infarction (MI). Intervention has been successful, and the patient's condition is now stable. The cardiologist is concerned about the patient's hemoglobin value of 9 g/dL. Active bleeding, hemolysis, and other potential causes of anemia (iron or vitamin B₁₂ deficiency, for example) have been excluded. The decision is made to transfuse to raise the hemoglobin level above 10 g/dL.

Is this the right thing to do? A recent paper and an oral presentation at a national meeting question the accepted "magic" hemoglobin level of 10, at least in the setting of acute MI.

EFFECT OF TRANSFUSION ON OUTCOME

Aronson and colleagues¹ investigated a potential association between red blood cell transfusion and 6-month outcome after acute MI (mortality and the composite end point of death, another acute MI, and heart failure). Of 2326 patients with acute MI who were admitted to a coronary care unit, 192 received transfusions. Six-month mortality was higher in patients with acute MI who received transfusions than in those who did not (28.1% vs 11.7%; P < .0001). The other end points of death, recurrent acute MI, and heart failure were also significantly increased by transfusion.

Stratification based on the hemoglobin level before transfusion led to an interesting qualification to these results. If transfusion was triggered by a hemoglobin value of less than 8 g/dL, it was protective; however, in patients with acute MI whose level was higher than 8 g/dL, transfusion was associated with a significant increase in adverse end points.

Another study, whose results were presented at a recent American College of Chest Physicians meeting, used a different design but led to similar caveats. A retrospective analysis of 1496 patients with acute MI (148 of whom were given transfusions) was the substrate for the data.² Atrial fibrillation, heart block, lethal ventricular tachycardia, cardiac arrest, and other significant cardiac events; length of stay; and in-hospital mortality were all increased by transfusion during the course of hospitalization for acute MI.

Because these data have not yet been published in a peer-reviewed journal, there have been some criticisms. Patients who were given transfusions were sicker (eg, were hypertensive or had diabetes) at the start than the control group of those with acute MI who did not receive transfusion. In the peer-reviewed paper, adjustments were made for baseline variables, and the study was prospective.¹

Theories posited to explain the negative impact of transfusion on certain patients with acute MI include¹:

• Storage-related deficiencies known to occur in transfused blood that would affect oxygen delivery.
• Potential pro-thrombotic effects of transfused blood products.
• Cytokine release triggered by transfusion.

IMPLICATIONS FOR PRACTICE

The data presented here are not the final answer, although they should lead to sober reflection. At this time, we need to be cautious when we transfuse any patient, but especially in the setting of acute MI. If nothing else, a hemoglobin level approximating 9 g/dL in a stable patient may not be an appropriate trigger for transfusion. The temptation to transfuse is strong but must be tempered to some degree.