KK Women's and Children's Hospital Singapore, Singapore
Background/Case Studies: Patients requiring extracorporeal membrane oxygenation (ECMO) support are critically ill and typically receive substantial amounts of blood and blood products. These transfusions occur usually in response to specific transfusion thresholds which are mostly unstudied. Our centre is a 16-bedded quaternary multidisciplinary paediatric intensive care unit (PICU) with an average of 15 to 20 ECMO runs annually, with an established protocol for transfusion of blood and blood products. Our practice of transfusion of red cells (RC), adherence to the haemoglobin (Hb) transfusion threshold of 10g/dL and documentation of clinical indications were audited.
Study
Design/Methods: This study involved retrospective audits of RC transfusions during ECMO runs from January to December 2022. The data captured included the types of ECMO, indications for ECMO, mode of cannulation, Hb level at the time of transfusion, volume of RC transfusion, evidence of bleed and patient outcome. Descriptive data were presented as median and interquartile range (IQR), and inferential analysis was used to describe the difference in the volume of RC received by patients based on the type of ECMO, i.e. veno-venous (VV), veno-arterial (VA) and the ECMO indications. Transfusions with no evidence of bleeds and at Hb more than 10g/dL were classified as out-of-criteria transfusions.
Results/Findings: Eighteen ECMO runs were audited from January to December 2022 (Table 1). There were 194 occurrences of RC transfusion. The median Hb (IQR) at the time of transfusion and transfusion volumes were 9.4 (8.5 – 9.8) g/dL and 13 (10 – 16) ml/kg respectively. 16/194 (7%) transfusions occurred with no documented evidence of bleed and at Hb more than 10g/dL. Of these 16 out-of-criteria transfusions, 12 (75%) and 4 (25%) were transfused to patients who underwent peripheral and central cannulation for VA ECMO respectively. Patients who underwent peripheral cannulation for VA ECMO received a higher median volume of RC transfusion (15 (IQR 11 – 15.25) ml/kg) as compared to patients who underwent central cannulation (14 (IQR 12.5 -15.25) ml/kg) (p < 0.001). There was no difference in transfusion volume according to indications for ECMO. Conclusions: Sixteen RC transfusions (7% of all RC transfusions) were given when patients had no evidence of bleed and at a Hb higher than the transfusion threshold of 10g/dL. The reasons for such transfusions were not documented. Improved adherence to transfusion indications may reduce future anticipated transfusion-related adverse events and optimize efficiency of blood product usage in the ECMO setting.
Importance of research: Transfusions are often administered in response to specific, but largely unstudied thresholds. Given the association between transfusion and adverse outcomes, effective blood management strategies and adherence to transfusion indications may be beneficial in our pediatric ECMO patients.
