Improving NICU Platelet Transfusion Guidelines
Platelet Mass Shows Promise
By Bill Malone
Current transfusion guidelines for neonates in the intensive care unit take into account the risk of bleeding and the platelet count. But because platelet transfusion carries with it significant risks, clinicians have struggled with knowing when prophylactic platelet transfusions are appropriate and what benefit they bring. A new study described in this issue of Strategies investigated whether including mean platelet volume, a parameter available on almost all cell counting devices but often overlooked, would refine the guidelines.
Thrombocytopenia is common in neonatal intensive care units (NICU), with many neonates receiving several platelet transfusions prophylactically to prevent intracranial hemorrhage and other bleeding events. However, recent studies have cast doubt on the merits of platelet transfusions, with some research showing increased mortality with higher numbers of transfusions. Part of the reason for concern is that every platelet transfusion exposes the baby to increased risk of an infection or a transfusion reaction. A blood bank can only keep platelets for a few days, so unlike red blood cells, each platelet transfusion originates with a different donor. Compounding the problem, platelets themselves are uniquely difficult to control for bacterial infection, and cannot be frozen or refrigerated. The situation leaves neonatologists in a dilemma, with mounting evidence of risk, and somewhat questionable guidelines for using platelet counts to guide them.
Now, new research suggests that one way out of this predicament is to use platelet mass—taking mean platelet volume (MPV) into account instead of using only the count—which the authors found lead to fewer platelet transfusions (Transfusion 2009; DOI: 10.1111/j.1537-2995.2009.02253.x). "Platelet transfusions in the newborn intensive care unit lack an evidence base, and this is a step in the right direction toward having a reasonable evidence base on which to make decisions," said Robert Christenson, MD, director of neonatology research at Intermountain Healthcare in Ogden, Utah, and an author of the paper. "We really have gone almost exclusively with treatments out of what we thought was best reasoning, rather than what we’ve assembled as best data. It’s a hard situation to study tiny, sick newborns, so there haven’t been very many studies to get that evidence base."
Trying Something New
Platelet mass is the product of the platelet count multiplied by MPV, the idea being that a deficiency in the number of platelets should be overcome by the efficacy of larger-sized platelets, explained Christensen. "When platelets aggregate to stop bleeding, the ones with a larger mass can do more of the work. So babies who have a platelet count of 50 x 109 cells/L and who have normal-size platelets, might have some bleeding problems. But one with the same count but with much larger platelets might not have a bleeding issue at all. Previous platelet transfusion schemes have ignored this difference in platelet size, with bigger platelets being more functional and smaller platelets being less functional."
Christenson explained an example in the study. A neonate with a platelet count of 120 x 109 cells/L and an MPV of 7 fL should have platelet plug generation that is hemostatically similar to a neonate with a count of 70 x 109/L and an MPV of 12 fL (120 x 7 = 70 x 12). Platelet mass is not a totally novel concept, and has been borne out in transfusion studies of adults; however, this was the first time the hypothesis has been tested on neonates. The research team kept a close eye on data coming out of the study so that if bleeding events increased, they could quickly change course, Christensen said.
The study employed a before versus after design, examining differences in platelet transfusion usage and hemorrhagic events between two time periods. During the first period, conventional platelet count-based guidelines were followed: <100 x 109/L for patients in the high risk group, <50 109/L for patients in the intermediate risk group, and <20 109/L for patients in the lower risk group. The authors then initiated formal, in-service educations for all neonatologists, nurse practitioners, and bedside nursing staff at both hospitals involved. The second period of the study used platelet mass-based guidelines calculated using an average MPV of 8 fL, a figure drawn from a previous study by the same authors that included MPV values of more than 40,000 neonates at Intermountain Healthcare. The second period guidelines established a platelet mass of <800, <400, and <160 for the three risk groups.
With the platelet mass guideline, the rate of transfusions was cut almost in half, from 3.6% to 1.9% (p < 0.002). After reviewing demographic information, bleeding events, and mortality rates for both periods, the authors concluded that the platelet mass-based guideline had accounted entirely for the reduction in platelet transfusions because it reduced the number of transfusions given prophylactically. Most important, the platelet mass-based guidelines in the second period did not lead to more bleeding episodes. In fact, there were fewer bleeding episodes in the second period, a result Christensen said he can’t explain.
While the current paper used a straightforward method to calculate platelet mass, the applicability of this strategy in other locations might depend on other factors besides being confirmed in more extensive research. The study exclusively used a Coulter LH750 hematology analyzer, which uses the traditional platelet impedance method. However, many labs use other company’s instruments that employ different methods and won’t always provide an MPV, explained Edward Wong, MD, director of hematology and associate director of transfusion medicine at Children's National Medical Center in Washington, D.C., who was not associated with the study. In addition, he is concerned that some instruments give MPV values that may be misleading. "Because neonates are often so very sick in the NICU, they have platelet sizes that can range from very small to very large, which is why some instruments don’t give an MPV, especially with low platelet counts," he said. Another instrument-related factor is that each analyzer tends to have a different average MPV.
These and other questions may get worked out as Christensen and his colleagues work on their next project. "This particular paper is provocative, but it may not be conclusive. So we have six large NICUs willing to study this in a randomized crossover design. This kind of collaborative, large study is really needed to see if this is really useful or not."