160 - Pre-transfusion Platelet Counts and Bleeding in Very Low Birth Weight Infants

Publication Number: 2154.160

Elizabeth Stone, Columbia University Vagelos College of Physicians and Surgeons, New York, NY, United States; Rebecca J. Birch, Westat, Rockville, MD, United States; Jeanne E. Hendrickson, Emory University School of Medicine, DECATUR, GA, United States; Erika M. Edwards, Vermont Oxford Network, Burlington, VT, United States; Eldad A. Hod, Columbia University Vagelos College of Physicians and Surgeons, New York, NY, United States; Matthew Karafin, University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, NC, United States; Oliver Karam, Yale School of Medicine, New Haven, CT, United States; Naomi L.C.. Luban, Children's National Research Institute, Washington, DC, United States; Gisela Marrero-Rivera, UCSF, San Francisco, CA, CA, United States; Jeffrey J. VanWormer, Marshfield Clinic Research Institute, Marshfield, WI, United States; Nalan Yurtsever, Yale School of Medicine, Fairfield, CT, United States; Martha Sola-Visner, Harvard Medical School, Boston, MA, United States; Ravi M. Patel, Children's Healthcare of Atlanta and Emory University, Atlanta, GA, United States; Marianne Nellis, Weill Cornell Medicine, New York, NY, United States

Background: Platelet transfusion (Plt Tx) thresholds for very low birth weight (VLBW, < 1500g) infants vary widely. It not known if Plt Tx thresholds are associated with bleeding in VLBW infants.

Objective: We evaluated how bleeding may be associated with pre-Tx Plt counts in the first 3 postnatal weeks among VLBW infants.

Design/Methods: We performed an observational birth cohort study of VLBW infants born in 8 U.S. hospitals from 4/2019-12/2023 as part of the Recipient Epidemiology and Donor Evaluation Study-IV-Pediatric (REDS-IV-P). Excluding surgical patients, VLBW infants with at least 1 Plt Tx in the first 3 postnatal weeks were included. We used ICD-10 codes and Vermont Oxford Network clinical data to identify major types of bleeding in these infants, including intraventricular hemorrhage (IVH), gastrointestinal bleeding, pulmonary hemorrhage, and other bleeding. To account for within-infant correlation, we used mixed effects multivariable linear regression to test if mean pre-Tx Plt counts differed among infants with and without bleeding. Models included adjustment for gestational age (GA), postnatal age (PA), and hospital.

Results: We evaluated 210 VLBW infants who received at least 1 Plt Tx during the first 3 postnatal weeks with a pre-Tx Plt count within 24 hours of Tx, for a total of 649 Plt Tx events. Of these infants, 39% were female sex, and 63% were born at < 24 weeks gestational age. 76% had documented bleeding, with IVH being the most common (61%), followed by other bleeding (26%), pulmonary hemorrhage (19%), and GI bleeding (10%). Most Plt Txs (71%) were given in the 1st postnatal week. The overall mean pre-Tx Plt count was 56 x 109/L among infants with any bleeding compared to 59 x 109/L among infants without bleeding (adjusted mean difference -2 x 109/L; 95% CI -14, 10; P=0.71). No significant difference in pre-Tx Plt between groups was seen when using large volume RBC Tx of >20 ml/kg as a surrogate for bleeding (P=0.32) and when evaluating only the first platelet transfusion (P=0.69). Increasing GA (P=0.002) and PA (p=0.03) were both associated with lower pre-Tx Plt counts. Among bleeding types, only infants with severe IVH had higher pre-Tx Plt counts compared to infants without severe IVH (adjusted mean difference 15 x 109/L; 95% CI 3, 28; P=0.01, Figure 1).

Conclusion(s): We found similar pre-Tx Plt counts among VLBW with most types of bleeding, although those with severe IVH had modestly higher pre-Tx Plt counts. These findings suggest that pre-Tx Plt counts are not associated with bleeding, except for severe IVH, in VLBW infants.

Figure 1. Density plot of pre-Tx Plt count (x10^9/L)
Density plot of the pre-Tx Plt count (x109/L) in infants without bleeding or with different bleeding diagnoses; n is the number of platelet transfusions per group.