Retrospective Review of Transfusion Practices in a Single Neonatal Intensive Care Unit (NICU) Compared to Recent Transfusion Guidelines

Background

Platelet and packed red blood cell (pRBC) transfusion practices for preterm infants vary significantly worldwide. Recent randomized trials of preterm infants receiving platelet or pRBC transfusions have not shown any benefit with higher compared to lower transfusion thresholds. In the recent PlaNeT2 study published in January 2019, infants receiving platelets at higher thresholds had significantly higher rates of death and major bleeding. These findings raise concern about transfusing adult platelets into a delicately balanced neonatal hemostatic system with relatively hypofunctional platelets as a potential etiology of this multifactorial outcome. In this retrospective review, we aimed to review transfusion practices and outcomes, including thrombosis, in the neonatal intensive care unit (NICU) at the University of Arizona, Tucson in comparison to the PlaNeT2 study.

Methods

After Institutional Review Board approval, we reviewed the electronic medical records of all neonates born at < 32 weeks gestation in the NICU or transferred to the NICU within 48 hours of birth between January 1, 2018 and December 31, 2019, corresponding to one year pre- and one year post-publication of the PlaNet2 study. For each patient record reviewed, we collected demographics, reviewed whether transfusions occurred, and documented pre-transfusion hemoglobin, platelet count, international normalized ratio, fibrinogen activity level, and clinical outcomes. We recorded the number of transfusions of each product, clinical central lines placed, and imaging studies performed to detect thrombosis. These data were analyzed using descriptive analysis.

Results

There were 127 neonates born <32 weeks gestation, all of whom were analyzed. Twenty-eight (35%) of neonates were born at <27 weeks' gestation and twenty-two (17%) were <1500 g at birth and were considered high risk infants. Seventy three percent of these patients were intubated at some point during their NICU admission. The majority of patients had bronchopulmonary dysplasia (63%). Ten patients developed necrotizing enterocolitis (7.8%), 28 had presumed or confirmed sepsis (22%), 52 had congenital heart disease (41%), and 12 had retinopathy of prematurity (9.4%).

During the two-year period of this review, 60% of patients had at least one pRBC transfusion, 13% of patients had at least one platelet transfusion, and 6% of patients had at least one fresh frozen plasma (FFP) transfusion. A total of 35 platelet, 370 pRBC, and 11 FFP transfusions were administered with a mean total of 3.3 total transfusions per patient. The mean and median pre-transfusion platelet counts were 79,300/uL and 52,000/uL, respectively. The mean and median pre-transfusion hemoglobin values were 12.1 g/dL and 12.2 g/dL, respectively. Only two of 35 platelet transfusions were given after the publication of the PlaNeT2 study in January 2019; however, both transfusions were given for neonates with pre-transfusion platelet counts >50,000/uL.

There were no arterial or venous thrombosis events noted in our population even though 108 patients (85%) had a peripherally inserted central catheters for an average of 12 days, 32 (25%) had umbilical arterial catheters for an average of 7 days, and 66 (52%) had umbilical vein catheters placed for an average of 6 days.

Conclusions

The majority of transfusions given to neonates born at <32 weeks gestation at the University of Arizona, Tucson institution over a two-year retrospective study period were pRBC transfusions with an incidence of 0.6 followed by platelet transfusions (incidence of 0.13) and FFP. Platelet transfusion practices changed drastically following publication of the PlaNet2 study, with only 2 of 35 (5.7%) platelet transfusions given after the publication of this study. No thrombotic events were documented in the study population despite a high prevalence of central venous lines and a mean pre-platelet transfusion threshold of 79,300/uL. Transfusion practice guidelines can affect clinical practices in the NICU and can potentially minimize risks. Larger prospective studies are needed to evaluate risk of thrombosis related to adult platelet transfusions given to neonates.