Transfusion strategies in patients with traumatic brain injury: which is the optimal hemoglobin target?

Robertson et al. (JAMA 2014; 312:36-47) investigated the effects of two different thresholds of hemoglobin (Hb) to guide red blood cells transfusions (RBCT; 7 g/dL vs. 10 g/dL) in patients suffering from traumatic brain injury (TBI). In a two-center, controlled, open-label trial (from May 2006 and August 2012), comatose patients with a closed TBI were randomized within 6 hours since initial resuscitation to one of the two RBCT strategies and, in a factorial design (2x2), to receive erythropoietin (EPO) or placebo. Patients were excluded if they had a Glasgow Coma Scale (GCS) score of 3 with fixed and dilated pupils, penetrating trauma, pregnancy, life-threatening systemic injuries and severe preexisting diseases. A total of 200 patients (7 g/dL with [N.=49] or without EPO [N. =50]; 10 g/dL with [N.=53] or without EPO [N.=48]) were enrolled among 598 who were screened. There was no interaction between EPO and Hb thresholds on the primary outcome, which was the occurrence of favorable neurological outcome, assessed using the Glasgow Outcome Scale (GOS) at 6 months after the injury (favorable=GOS 4-5). Favorable outcome was similar between patients included in the 7 g/dL (37/87-43%) and the 10 g/dL group (31/94-33%) as if receiving EPO or placebo, even after adjustment for several covariates. Thromboembolic events were significantly more frequent in the group transfused at 10 g/dL (22/101 [22%] vs. 8/99 [8%]; P=0.009). We discussed how theses results might influence the management of such patients as well as the methodological limitations that underline the need for further investigations.