Stanworth et al previously showed that 30% of ICU patients have prothrombin time prolongation at some point. Here, they examined FFP transfusion In an 8-week-long prospective observational study in 29 ICUs in the UK. Fully 31% of FFP units were given to patients with INR < 1.5. About half the FFP units went to patients believed to be bleeding (although not massively so), and 36% as pre-procedure prophylaxis. Given that some of the trauma literature advocates high FFP:RBC transfusion ratios, this isn't entirely indefensible; authors rightly suggest better guidelines should be issued on FFP use. Critical Care 2011;15:R108