Comparison of Platelet Mass Index to Platelet Count as Transfusion Trigger in Neonatal Extracorporeal Membrane Oxygenation.

BACKGROUND

Platelet transfusions are routinely administered to neonates in intensive care units when there are concerns of bleeding, including high-risk situations like Extracorporeal Membrane Oxygenation (ECMO). Most platelets in ICUs are transfused prophylactically for thrombocytopenia based solely on the platelet count. Platelet Mass Index (PMI) has been proposed as an alternative to platelet count (PC) as a transfusion trigger. The objective of this study was to determine the relationship between PMI and platelet-specific maximal clot firmness (PMCF) in Rotational thromboelastometry (ROTEM), which gives an indication of platelet contribution to clot firmness and to investigate whether PMI may be a better choice as a trigger for platelet transfusions than PC.

METHODS

Retrospective review of medical records of neonates with congenital heart disease placed on ECMO support in the cardiovascular intensive care unit (CVICU) from 2015 to 2018 was conducted. Platelet count (PC), platelet mean volume (PMV), ROTEM parameters along with demographic data including gestation age, birth weight, gender and survival were collected. Mixed effects linear models with a first order autoregressive covariance structure were used to assess the associations of PMI, PC, and MPV against PMCF. In addition, generalized estimating equations with a first order auto-regressive covariance structure were used to compare odds of transfusion using PC versus PMI triggers.

RESULTS

A total of 92 tests on consecutive days were obtained for 12 ECMO patients (5 male, GA = 38.1 ± 1.6 weeks, BW = 3.1 ± 0.4 kgs, mean ± SD). A variation of 40.1% in PMCF was explained by platelet count (p < 0.001) while 38.5% of the variation in PMCF was explained by PMI (p < 0.001). If the platelet transfusion trigger was PC < 100 x 103 platelets/µL vs. PMI < 800. Using the PC trigger yielded significantly higher odds of transfusion compared to the PMI trigger (odds ratio = 1.31, 95% confidence interval: 1.18 - 1.45, p < 0.001).

CONCLUSIONS

While our study failed to demonstrate a superior correlation of PMI with PMCF than PC, our study did reveal that using PMI as transfusion trigger would result in significantly less platelet transfusions, when compared with the current practice of using PC as a trigger.