Sanne de Bruin

36 Chapter 2 lations, a greater variety of applied Hb thresholds was found. For the septic patients respondents reported a significantly higher Hb threshold compared to the general population, what deviates from current evidence supporting a restrictive transfusion strategy also in septic patients 11 . For patients with ACS, the higher preferred Hb thresh- old of 9 g/dL (8-9.6) is in accordance with the transfusion guideline from the National Institute for Health and Care Excellence (NICE, 2018), in which a Hb threshold (8-10 g/ dL) for patients with symptomatic coronary disease is advised. Also, patients with trau- matic brain injury were transfused at higher Hb thresholds since these patients may be more sensitive to anaemia induced cerebral hypoxia. However, evidence to justify this more liberal transfusion practice is limited. Multiple large RCTs are currently studying whether these patients benefit from a liberal transfusion strategy (ClinicalTrials.gov, NCT02968654 and NCT02981407). This survey also showed a high variety in preferred Hb thresholds for patients with ARDS and patients on ECMO. Since the evidence for these subpopulations is limited it is expected to observe a high heterogeneity in transfusion practice. For ARDS patients it is hypothesized that the hypoxaemia should be compensated by increasing the oxygen carrying capacity of the circulating blood by transfusing at higher Hb thresholds. How- ever, there is no solid evidence to support this practice, and the downside of allogenic blood transfusion is not taken into account in this reasoning. The applied platelet threshold differed between patients with and without an upcoming invasive procedure. The majority of the respondents (72%) would transfuse non-bleed- ing critically ill patients at a platelet count of ≤ 20x10 9 cells/L. The potential harm of platelet transfusion is supported by two recent RCTs, in which it was shown that prophy- lactically platelet transfusionmight be particularly harmful in neonates 16 and in patients with a cerebrovascular accident 17 . These studies cannot be directly translated to the non-bleeding critically ill adult patients, but they do show that platelet transfusion is not an intervention without risk. Prior to invasive procedures physicians transfuse platelets at higher platelet counts, while the evidence for this is limited. Ameta-analysis has shown that complications prior to CVC placement in patients with coagulopathy, including thrombocytopenia and prolonged INR and APTT, are rare 18 . Thus, the need for any platelet transfusion prior to this procedure is questionable. A large RCT studying the need of platelet transfusion prior CVC placement in severe thrombocytopenic patients is now recruiting 19 . Multiple RCTs have failed to demonstrate beneficial effects of prophylactic plasma trans- fusion prior to an invasive procedure in critically ill patients with an prolonged INR 12,20,21 .