Indications for Platelet Transfusion

Excerpt from online article in the New England Journal of Medicine, December 25, 1997: Editorials -- NEJM 1997; 337: 1914-1915

Despite a recent National Institutes of Health consensus conference on platelets, (8) the role of platelet transfusions, especially those given prophylactically to forestall bleeding, remains controversial. At one extreme, results of the bleeding-time test become abnormal when the platelet count falls below 100,000 per cubic millimeter; at the other, spontaneous small-vessel bleeding does not increase until the platelet count is less than 5000 per cubic millimeter. (9) The currently accepted threshold for platelet transfusion was derived from a 1962 study by Gaydos et al., in which gross hemorrhage in patients with leukemia became more frequent at platelet counts of less than 20,000 per cubic millimeter. (10) However, this report established no threshold value that prevented all bleeding; furthermore, many patients in the study were treated with aspirin, which would have increased the risk of bleeding regardless of the platelet count.

Full text of online article in the New England Journal of Medicine, December 25, 1997: Editorials -- NEJM 1997; 337: 1914-1915

Commentary

Dr. Landon asks:


Dr. Coates, do you agree? 10,000 as a threshold for transfusion?

Dr. Coates replies:



I generally agree with this, though it is a little non-committal. The original article is very good and the first real attempt to counter the 1962 data. We have been radiating all blood products for years and using leukocyte filters. In general, we do not transfuse platelets unless there is evidence of significant bleeding (bruising, petichiae, microscopic hematuria and mild mucosal bleeding don't count) or some other specific indication known to increase the chance of intracranial bleed ( post surgury brain tumor, concurrent sepsis, DIC), even if the platelet count is less than 10,000.