A Multi-Centre Prospective Observational Study of Platelet Transfusion Practice in Neonates with Severe Thrombocytopenia.

Background: Platelet transfusion practice in neonates is not evidence-based and there is a lack of data relating transfusion to clinical outcome. To inform the design of clinical trials, we conducted a prospective multicentre observational study of platelet transfusion in thombocytopenic neonates to describe:

1. transfusion practice, including reason for transfusion;

2. clinically-related outcomes, including minor and major bleeding and mortality.

Methods: Neonates with platelets <60x10⁹/l were studied at 7 UK neonatal intensive care units Mar 05-Jun 06. With parental consent, daily data were collected on minor bleeding (blood staining of oral/nasogastric/endotracheal secretions and stool; microscopic haematuria; petechial rash; oozing from puncture sites - scored 0–7), and major bleeding (intraventricular (IVH), intra-abdominal, pulmonary or renal). Surviving neonates were studied for 7 days or until platelets were ≥60x10⁹/l. Blood counts and all transfusions were directed by the attending neonatologist, with reason for transfusion and its effect on bleeding prospectively documented.

Results: 145 of 167 (87%) eligible neonates were enrolled; 123 (85%) were < 37 weeks gestational age (GA), 89 (61%) were male. The study documented 186 episodes of thrombocytopenia (1606 study days) and 309 platelet transfusions given to 91 (63%) neonates. GA, birth weight and age at thrombocytopenic episode were: 27 (24–32) weeks; 825 (675–1370) grams; 5 (2–16) days, respectively [all data median (IQR)]. 21/145 babies (14%) had major IVH (Grade 3/4) at study entry. Platelets (x 10⁹/l) at study entry, and at platelet nadir, and duration of count <60x10⁹/l were: 44 (33–54); 31 (19–42); 2 (1–5) days respectively. In transfused neonates, platelets (x10⁹/l) pre- and post-transfusion, and number of transfusions were: 27 (19–36); 84 (46–138); 2 (1–4) respectively. The principal indications for transfusion were:- platelet count below threshold of unit guideline: 265/309 (86%); deteriorating clinical condition: 17/309 (6%); and significant bleeding: 7/309 (2%). At least 1 episode of minor bleeding was recorded on 622/1606 (39%) of study days. Minor bleeding scores recorded for 12 hours pre- and post each transfusion were [median (IQR)]: 1 (0–2) and 0 (0–1) respectively. New or extension of IVH bleeding to Grade 3/4 occurred in 7 neonates and other major haemorrhage in 9 neonates (4 pulmonary, 3 GI). A total of 20/145 (14%) neonates died, 13 with major IVH bleeding (of which 10 had Grade 3/4 IVH at study entry); all received platelet transfusions [total 87; median 3 (2 to 7).

Conclusion: This study confirms that most neonates with severe thrombocytopenia are preterm, most episodes develop after 72 hours of life (median 5 days) and are of short duration (median 2 days). Minor haemorrhage is common and may be reduced by platelet transfusion. Major haemorrhage is uncommon (11%), is associated with mortality (82%) and affected patients receive a large number of platelet transfusions. There is a clear distinction between the majority of thrombocytopenic neonates who receive one or two transfusions as prophylaxis with a good outcome, and the minority who suffer adverse outcomes despite transfusion. Many neonates are transfused with platelets at thresholds below those suggested in guidelines without apparent clinical detriment. These data will be invaluable for planning the randomised trials necessary for rationalising platelet transfusion in these vulnerable patients.