Studies about the use of blood in patients in extracorporeal membrane oxygenation (ECMO) because of COVID-19 are scarce. This study aims to evaluate retrospectively the use of blood components by patients in ECMO diagnosed with COVID-19 from March 2020 to July 2022, considering the consumption of blood components over time and the clinical factors involved.

Retrospective study with 37 patients diagnosed with COVID with criteria for ECMO support (two in veno-arterial because of pulmonary emboly, and 35 in veno-venous). We excluded patients undergoing transplantation. This study was approved under the number: CAAE 62641822.7.0000.0071. p<0.05 was considered statistically significant.

The mean age of the patients was 55.76 (±11.26), 83.8% of males, 33.3% had hypertension previously diagnosed, 67.6% died during hospitalisation, and 75.6% received hemodialysis. The median SAPS3 was 51 (29‒73). Body mass index was 29,93 (±5.57). The mean length of hospitalisation were 43.46 (±33.67) and 29.94 (±22.09) for mechanical ventilation (five patients were discharged in rehabilitation and were not included in this calculation). 45.9% had thrombosis, of which 29.4% during ECMO. 33.3% had bleeding during ECMO. On the day of the circuit installation, the mean value of haemoglobin was 10.9 (±2.15) and TTPA 1.5 (±0.76), but it did not influence the number of transfusions (p>0.05). The mean of RBC transfused was higher and statistically significant in the period of ECMO support (9.15± 9.10 vs. 2.41±4.26). Only three units of fresh frozen plasma were transfused. All patients received heparin unfractionated as an anticoagulant during ECMO. There was no difference in the patient's use of platelets (apheresis or pool) during the support (0.43±1.05 vs. 0.16±0.72). Any cryoprecipitate was transfused. There was a correlation (Spearman test) between the use of RBC and length of ECMO support (p<0.001), length of hospitalisation (p<0.004), length of mechanical ventilation (p<0.004) and length of intensive care support (p<0.001).

The survival rate of ECMO is about 40%, diminishing with more days in support. Our data was similar to the literature regarding epidemiological characteristics (age, male, high body mass index, and hypertension being the most common chronic disease). Bleeding and thrombosis are frequently associated with ECMO-treated patients with COVID-19, which may lead to the need for transfusion of multiple blood products. Despite that, our results showed that RBC was the product most transfused, and it was correlated with the length of hospital stay, mechanical ventilation, and intensive care support.

There is an increase in the consumption of RBC for patients with COVID-19 in ECMO support. Blood banks should be prepared for an emergent virus with respiratory symptoms and ECMO support indication.