Platelet transfusion in the critically ill

Reduction in platelet numbers (thrombocytopenia) is associated with reduced hemostatic response to blood-vessel injury and therefore increased risk of bleeding. In health, the concentration of platelets in peripheral blood is maintained within the range of 150-400 × 10³/mL. The risk of spontaneous life-threatening hemorrhage increases dramatically if the platelet count falls below 5 × 10³/mL. 

Platelet transfusion is prescribed for the treatment, but much more commonly, for the prevention of hemorrhage in patients with severe thrombocytopenia. Current guidelines suggest that platelets should be used therapeutically as part of the treatment of a patient who is bleeding if the platelet count is less than 50 × 10³/mL. 

For those who are not bleeding, the platelet count trigger for (prophylactic) platelet transfusion is currently set at <20 × 10³/mL for infected patients and <10 × 10³/mL for those not infected. In common with all blood product prescriptions, platelet transfusion is associated with risk of serious adverse effect and there is a paucity of clinical evidence of the overall benefit of prophylactic platelet transfusion for some patient groups, including the critically ill. 

The issue is visited in a recently published retrospective study of platelet transfusion practice at the intensive care units of the Mayo Clinic in Rochester, Maryland. During a 5-month study period, 1417 medical and surgical patients were admitted to intensive care. Of these, 162 (11.4 %) had at least one platelet count <50 × 10³/mL and were thus potential candidates for platelet transfusion. 

Twenty-eight of these patients were actively bleeding and, in line with guidelines, were given platelet transfusion. A further 17 patients had to be excluded, leaving 117 study patients who were thrombocytopenic but not bleeding. The use of prophylactic platelet transfusion in this patient cohort was the focus of the study. Of the 117, 90 received prophylactic platelet transfusion. 

The median threshold platelet count for transfusion was 23 × 10³/mL (range 11-38 × 10³/mL). The platelet count of the 27 patients not transfused ranged from 18 to 33 × 10³/mL. Significant new bleeding occurred in just one patient, and this was paradoxically from the transfused group. 

Six (8 %) of the transfused patients suffered significant transfusion-related complications, including two cases of potentially fatal TRALI (transfusion-related acute lung injury). Any invasive procedure is a risk factor for bleeding that might prompt the use of platelet transfusion in thrombocytopenic patients. In this study, 28 % of thrombocytopenic patients undergoing an invasive procedure were not given a platelet transfusion. 

Overall, the study identified significant variation in transfusion practice. In a significant number (19 %) of cases, platelets were administered despite the fact that, according to current guidelines, transfusion was not indicated. 

The study provided no evidence that prophylactic platelet transfusion was associated with improved outcome, and the authors suggest that prospective studies examining the risk-benefit ratio of liberal and restrictive use of platelet transfusion in the critically ill are long overdue.