Platelet transfusion in neonatal intensive care units of 22 European countries: a prospective observational study

Abstract

Background: Platelet transfusions are given to preterm infants with severe thrombocytopenia aiming to prevent haemorrhage. The PlaNeT2/MATISSE trial revealed higher rates of mortality and/or major bleeding in preterm infants receiving prophylactic platelet transfusions at a platelet count threshold of 50 × 10⁹/L compared to 25 × 10⁹/L. The extent to which this evidence has been incorporated into clinical practice is unknown, thus we aimed to describe current neonatal platelet transfusion practices in Europe.

Methods: We performed a prospective observational study in 64 neonatal intensive care units across 22 European countries between September 2022 and August 2023. Outcome measures included observed transfusion prevalence rates (per country and overall, pooled using a random effects Poisson model), expected rates based on patient-mix (per country, estimated using logistic regression), cumulative incidence of receiving a transfusion by day 28 (with death and discharge considered as competing events), transfusion indications, volumes and infusion rates, platelet count triggers and increment, and adverse effects.

Findings: We included 1143 preterm infants, of whom 71 (6.2%, [71/1143]) collectively received 217 transfusions. Overall observed prevalence rate was 0.3 platelet transfusion days per 100 admission days. By day 28, 8.3% (95% CI: 5.5-11.1) of infants received a transfusion. Most transfusions were indicated for threshold (74.2%, [161/217]). Pre-transfusion platelet counts were above 25 × 10⁹/L in 33.1% [53/160] of these transfusions. There was significant variability in volume and duration.

Interpretation: The restrictive threshold of 25 × 10⁹/L is being integrated into clinical practice. Research is needed to explore existing variation and generate evidence for various aspects including optimal volumes and infusion rates.

Funding: Sanquin, EBA, and ESPR.

Keywords: Epidemiology; Europe; Neonatal intensive care unit; Neonatology; Platelet transfusion; Preterm infants; Thrombocytopenia.

Fig. 1

Overall observed platelet transfusion day prevalence rates, and observed and expected (based on patient mix) prevalence rates per country. Observed prevalence rates were calculated using random effects Poisson models to pool transfusion day prevalence rates from the individual centres into country subgroup estimates and subsequently to derive the overall estimate. Confidence intervals were in part determined by the sample size per country as well as the practice variation between centres within one country. Upper limit confidence intervals of Hungary and United Kingdom were outside of axis limits. Expected prevalence rates as predicted based on patient-mix using a logistic regression model which included the following variables: sex assigned at birth, GA at birth, birth weight, congenital anomalies, SGA, major bleeding, NEC, sepsis, mechanical ventilation, surgical procedure, postnatal day.

Fig. 2

Cumulative incidence of receiving at least one platelet transfusion during the first 28 postnatal days of life. Adjusted for the competing risks of death and discharge without platelet transfusion (presented as cumulative incidence, with corresponding 95% confidence intervals). Figure is based on 468 out of 1143 (40.9%) infants that were followed from birth. The number at risk and the number of evenst per postnatal day are available in Supplementary Table S5.

Fig. 3

Primary indications for 217 platelet transfusions. Indications for transfusion were platelet count threshold (161/217, 74.2%), active bleeding (22/217, 10.1%), surgical procedures (14/217, 6.5%), prevention of major bleeding (based on other factors besides platelet count; 8/217, 3.7%), critically ill conditions (such as NEC and sepsis; 7/217, 3.2%), and other indications (5/217, 2.3%). Other indications included lumbar puncture, urinary catheter placement, low molecular heparin weight treatment for inferior vena cava thrombosis, nonactive bleeding, and petechiae.

Fig. 4

a. Transfusion volume, duration and infusion rate of 161 transfusions given based on threshold. Transfusion volumes were 5 mL/kg or less in 1.9% (3/161), 10 mL/kg in 13.0% (21/161), 15 mL/kg in 64.6% (104/161), 20 mL/kg in 18.6% (30/161), and 25 mL/kg in 1.9% (3/161). Transfusion durations were 30 min or less in 47.8% (77/161), 1 h in 30.4% (49/161), 1.5 h in 3.1% (5/161), 2 h in 13.7% (22/161), 2.5 h in 0.6% (1/161), and 3 h in 4.4% (7/161). Colors reflect most common infusion rate within each dot, infusion rates of individual transfusions within one dot may vary as infusion rates were calculated using unrounded transfusion volumes and durations. Percentages in figure represent the proportion of transfusions indicated for threshold given at this volume and duration. b. Transfusion volume, duration and infusion rate of transfusions given for active bleeding, surgery, prevention of major bleeding, critically ill conditions or any other indication. Volume and duration were available in 50 out of 56 transfusions (89.3%). Colors reflect most common infusion rate within each dot, infusion rates of individual transfusions within one dot may vary as infusion rates were calculated using unrounded transfusion volumes and durations. Percentages in figure represent the proportion of transfusions indicated for active bleeding, surgery, prevention of major bleeding, critically ill conditions or any other indication given at this volume and duration.

Fig. 5

Platelet count prior to transfusion. Platelet counts prior to transfusion were available in 211 out of 217 transfusions (97.2%). The dotted lines represent the liberal transfusion threshold of 50 × 10⁹/L and restrictive transfusion threshold of 25 × 10⁹/L as evaluated in the PlaNeT2/MATISSE trial. The middle line of each box denotes the median, the ends of the box denote the 25th to 75th percentile, the whiskers denote the minimum to maximum range. Platelet count prior to transfusion given for active bleeding (n = 1 transfusion) and surgical procedure (n = 3 transfusions) outside axis limits, respectively.