Objective: Several studies have indicated a correlation between the number of platelet transfusions received by newborn intensive care unit (NICU) patients and the mortality rate. The number of platelet transfusions might be a marker for level of illness, and thus predictive of mortality. However, an alternative hypothesis is that multiple platelet transfusions themselves are harmful in this population.
Study design: We evaluated data from all thrombocytopenic neonates cared for in the Intermountain Healthcare NICUs in the past 4 years, seeking associations between the lowest platelet count recorded, number of platelet transfusions received and mortality rate. We also conducted a sensitivity analysis to examine the hypothesis that platelet transfusions were responsible for some fraction of the mortality rate.
Result: Transfusion and outcome data were examined from 1600 thrombocytopenic NICU patients. At any level of platelet count, some patients received platelet transfusions but others did not. However, at all levels of platelet count, those that received platelet transfusions had a higher mortality rate. Neonates not given any platelet transfusions had a mortality rate of 2%, those with 1 or 2 transfusions had a mortality rate of 11% (P<0.001); those with >10 had a mortality rate of 35% (P<0.001); and those with > or = 20 had a mortality rate of 50% (P<0.001). A sensitivity analysis suggested that the platelet transfusions themselves were very likely responsible for some fraction of the increasing mortality rate.
Conclusion: The number of platelet transfusions administered in the NICU predicts the mortality rate. Some of this correlation is ascribable to unknown and unmeasured factors such as level of illness. However, the present data and the sensitivity analysis both suggest that some of this correlation is due to harmful effects of multiple platelet transfusions in this group of patients.