Benefits and harms of red blood cell transfusions in patients with septic shock in the intensive care unit.

BACKGROUND: Transfusion of red blood cells (RBCs) is widely used for non-bleeding patients with septic shock in the intensive care unit (ICU). The evidence for effect and safety are limited showing conflicting results and transfused RBCs have the potential to harm subgroups of critically ill patients. Our aim was to assess the benefits and harms of RBC transfusion in patients with septic shock in a randomised clinical trial and to conduct an up-to-date systematic review with meta-analysis of all randomised clinical trials comparing different transfusion strategies.
METHODS: We planned and conducted a randomised, partly blinded, clinical trial assigning patients with septic shock in the ICU and a haemoglobin level of 9 g/dl (5.6 mM) or below to receive single units of pre-storage leukoreduced RBCs at a lower haemoglobin threshold level of 7 g/dl (4.3 mM) or below or a higher haemoglobin threshold level of 9 g/dl (5.6 mM) or below. The primary outcome was death by day 90 after randomisation. Secondary outcomes were need for life support, severe adverse reactions, ischaemic events in the ICU and days alive and out of hospital. Secondly, we conducted a systematic review of randomised controlled trials comparing benefits and harms of using restrictive (range of lower haemoglobin thresholds) versus liberal (range of higher haemoglobin threshold) transfusion trigger strategies to guide RBC transfusion and pooled results in meta-analyses and trial sequential analyses.
RESULTS: Of the 1005 patients that underwent randomisation 998 were included in analysis of the primary outcome of mortality. Ninety days after randomisation, 216 of 502 patients (43%) in the lower threshold group had died compared to 223 of 496 (45%) patients in the higher threshold group (relative risk (RR) 0.94, 95% confidence interval (CI) 0.78 to 1.09, p=0.44). The number of patients who required life support, who had ischaemic events, severe adverse reactions and number of days alive and out of hospital were similar in the two groups. Patients in the lower threshold group received 1588 units of RBCs compared to 3088 units in the higher group. A total of 176 (36%) patients in the lower threshold group never received RBCs in the ICU compared with six patients (1%) in the higher threshold group. The systematic review identified 31 trials with a total of 9813 patients in different clinical settings. In meta-analyses restrictive versus liberal transfusion strategies were not associated with the RR of death (0.89, 95% CI 0.76 to 1.05, 5607 patients in eight trials with lower risk of bias), overall morbidity (RR 0.98, 95% CI 0.85 to 1.12, 4517 patients in six trials with lower risk of bias), fatal or non-fatal myocardial infarction (RR 1.32, 95% CI 0.61 to 2.83, 4630 patients in six trials with lower risk of bias). Trial sequential analysis on mortality and myocardial infarction showed that required information sizes had not been reached but use of restrictive transfusion strategies was associated with reduced numbers of RBC units transfused (mean difference -1.43, 95% CI -2.01 to -0.86) and reduced proportion of patients transfused (RR 0.54, 95% CI 0.47 to 0.63).
CONCLUSION: The TRISS trial provided evidence for the safe use of 7 g/dl as transfusion trigger in patients with septic shock and reduced the number of units transfused with about half. In line with this, the updated systematic review including data from several recent trials showed no associations with mortality or other adverse events when comparing restrictive to liberal RBC transfusion strategies, however, restrictive transfusion strategies reduce the exposure of patients to RBC transfusions and reduce number of transfused RBC units. Given the fact that liberal transfusion strategies have not been proven beneficial, a more restrictive approach should be considered. Results from the TRISS trial together with other recent trials have the potential to alter the international guidelines for transfusing critically ill patients. Several guidelines have been updated the last years recommending the use of 7-8 g/dl as the ''universal'' trigger level. Patients with acute myocardial ischaemia and patients with acute brain injury may need special considerations.