Sanne de Bruin

35 Transfusion practice in the non-bleeding critically ill; an international online survey – The TRACE Survey x10 9 cells/L for all world regions. However, prior to an invasive procedure more het- erogeneity exist (Additional file 1 figure S4). Prior to general surgery the largest differ- ences between regions was observed, in half of the regions the mean of the applied platelet threshold was 50 x10 9 cells/L while in Southern Asia and South-Eastern Asia a median platelet count of 72.5 and 100 x10 9 cells/L was applied, respectively. Lastly, also plasma transfusion practices differed between world regions (Additional file 1 figure S5). In Southern Europe only 17% would never correct a vitamin K induced INR >3 in the absence of an invasive procedure, whereas in South-Eastern Asia 50%would never correct this INR. Figure 4. More respondents would correct an vitamin K induced INR >3 prior to an invasive procedure than in the absent of a procedure(4A). Both, in the absence of an invasive procedure (4B) and prior an invasive procedure (4C), the majority would correct this with vitamin K. Discussion This is the largest survey on transfusion practice in non-bleeding critically ill patients among ICU physicians to date. The main findings of this study are: 1) a high Hb thresh- old variation between ICU subpopulations, 2) the platelet transfusion threshold prior to invasive procedures differs greatly between and within the procedures, 3) plasma is considered by a large number of physicians in non-bleeding patients even in the absence of an invasive procedure, 4) base specialty of physicians is associated with variation in transfusion practices and 5) worldwide, institutions lack local ICU specific transfusion guidelines. The reported Hb threshold for the general ICU population in this survey is in line with the finding of the TRICC study, which demonstrated the safety of a restrictive transfusion strategy in the ICU population 8 . However, when looking at different patient subpopu- 2