This review of interventions for trauma patients within 24 hours of injury found no evidence of an association between transfusion reduction and improvements in survival. The conclusions of this well-conducted review appear reliable.

To evaluate interventions for trauma patients with haemorrhagic shock within the first 24 hours after injury.

MEDLINE and EMBASE were searched from inception to July 2010 as well as Cochrane Central Register of Controlled Trials (CENTRAL; issue 7, 2010). Several trial databases including Current Controlled Trials, ClinicalTrials.gov and the National Health Service Blood and Transplant Systematic Review Initiative were searched. Reference lists were also searched for additional studies. The search strategy was reported and there were no language restrictions.

Randomised controlled trials that compared treatment with an alternative injury and reported on bleeding, blood loss, coagulopathy or transfusion requirements, were eligible for inclusion. At least 75% of the patients had to be trauma patients with bleeding or haemorrhagic shock. Quasi-randomised trials were included but those that assessed isolated traumatic brain injury or burns were excluded.

Most trials administered the intervention in hospital and most included only trauma patients. Four trials included non-trauma patients that ranged from four to 25% of patients. All trials included only civilian patients. Most trials included both blunt and penetrating injuries, there were six trials of penetrating injuries only and one of blunt injuries only. For the 25 trials that reported injury severity scores the mean score was 24 (range 15 to 33). Full details of all trials were included in appendix files.

Studies were selected by two reviewers independently.

Assessments of the randomisation method, allocation concealment, blinding and incomplete outcome data were made to judge study quality.

Assessment was performed by one reviewer and checked by a second; disagreements were resolved by consensus.

As well as bleeding outcomes results for mortality and morbidity including multi-organ failure, acute respiratory distress syndrome and infection were extracted.

Data were extracted by one reviewer and checked by a second, disagreements were resolved by consensus.

Results were presently descriptively because of the heterogeneity of the interventions. Interventions were grouped by the four main types of intervention: blood and blood saving strategies; mechanical and surgical management; use of intravenous fluids for resuscitation; and pharmaceutical agents.

Thirty-five trials were included (participants ranged from 32 to 20,211). Most (23) trials were single centre studies and all except one had a parallel design (one was a cross-over trial). Twelve trials had adequate randomisation, 13 had adequate allocation concealment, four had blinded outcome assessment and 31 had loss to follow-up less than 10%. Only one trial satisfied all quality criteria.

Blood and blood saving strategies (seven trials, 1,374 participants)

Two trials evaluated blood product administration but found no significant differences in mortality (both trials), microvascular bleeding or coagulation (one trial each). Five trials evaluated methods of reducing allogeneic blood use. Cell salvage at 24 hours significantly reduced transfusion (one trial) and blood substitute significantly reduced red blood cell requirements (three trials). PolyHeme significantly increased prolonged prothrombin time and activated partial thromboplastin time, but there was an imbalance in these outcomes at baseline (one trial). Two out of four blood substitute trials reported mortality and found no significant differences.

Mechanical and surgical management (two trials, 257 participants)

One trial evaluated the use of pneumatic anti-shock garments for traumatic injury and found a trend towards increased mortality with the garments. The other evaluated vascular control of renal vessels during surgery for kidney injury and found no effect for intra-operative blood loss. Neither trial found any difference in transfusion requirements or reported on coagulation.

Intravenous fluids for resuscitation (18 trials, 3,394 participants)

Twelve trials compared different resuscitation fluids (full details reported in paper). Six reported fluid administration strategies. One trial found significant reductions in mortality at 24 hours and 30 days with hypertonic saline dextran but this was not seen in six other trials of the same intervention. One trial of continuous arteriovenous rewarming also found a significant reduction in 24 hour mortality but not at discharge. One trial reported a significant increase in acute respiratory distress syndrome after albumin administration. One trial found a significant difference in infection rates following Plasma Protein Fraction infusion. There were no differences for multi-organ failure (two trials). One trial found a significant reduction in red blood count use at one hour with pentastarch, another throughout resuscitation with hypertonic solutions, and a third in the first hour after use of a rapid infuser. Three trials reported significant improvements in clotting parameters. Ten out of 12 trials found no differences in transfusion requirements.

Pharmaceutical agents (eight trials, 21,689 participants)

Three trials evaluated antifibrinolytics (two aprotinin, one tranexamic acid), four publications (from three trials) evaluated recombinant factor VIIa and two trials evaluated novel drugs (a bactericidal protein and a monoclonal antibody). All trials reported mortality and one assessing tranexamic acid found significant reductions in all-cause mortality and death from bleeding. For recombinant factor VIIa, one trial found a significant reduction in multi-organ failure rates in blunt trauma patients surviving more than 48 hours and one trial found a significant reduction in acute respiratory distress syndrome. Two other trials also found significant reductions in acute respiratory distress syndrome. This treatment also led to significant reductions in red blood cell use in blunt trauma and coagulopathic patient subgroups and greater numbers of massive transfusions if their post study prothrombin time remained elevated at one hour.

Due to the multifactorial nature of haemorrhage, differences in interventions and difficulties with trial design and conduct, only limited conclusions could be drawn. There was no evidence of an association between reduction in transfusion requirements and increases in survival, large well-conducted studies with pragmatic outcomes are needed.

This review had a clear aim, the literature search was thorough and included unpublished studies and those not published in English. Study selection, quality assessment and data extraction were performed by two reviewers to reduce the chance of errors and bias. Given the amount of and variation in the evidence, a narrative summary was appropriate. Study quality was assessed and the influence of quality on the reliability of the results was discussed. This was a well-conducted review and the authors' cautious conclusions appear to be reliable.

Practice: The authors did not state any implications for practice.

Research: Large, well-conducted studies with pragmatic endpoints were needed to improve the understanding of the complex interplay between bleeding and coagulopathy, transfusion requirements and mortality.

National Institute for Health Research Programme Grant for Applied Research.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.