The infrastructure provided by mature trauma systems has facilitated timely delivery of specialized care to the critically injured. What were once universally fatal injuries are now identified and treated rapidly in trauma centres, resulting in increased survival rates. However, opportunities exist to prevent more deaths through the optimization of prehospital care1. The trimodal distribution of death following major trauma, consisting of immediate (within 1h of injury), early (within hours of injury) and late (within several weeks of injury) deaths, has been replaced in mature trauma systems by a unimodal distribution. The majority of deaths occur in the first several hours after injury, predominantly from haemorrhage and brain injury2. These observations highlight the need to focus on immediate postinjury care. The elimination of immediate and early trauma deaths requires rapid identification of imminently life-threatening injuries, early initiation of blood product resuscitation, and swift deployment of therapeutic interventions. The prehospital component of individual trauma systems is variable, and based on needs and available resources. The overarching objectives of rapidly responding, triaging and transporting the critically injured remain the same. Compliance with existing protocols is critical to delivering the most severely injured to designated trauma centres immediately after injury3. Future trauma systems will employ emerging technologies to expedite the identification of life-threatening injuries. Mobile telephones have already decreased first-responder notification times after injury. Vehicular accident alert and event data recorders have the potential automatically to alert first responders to a collision and provide detailed mechanistic data. Advances in ultrasound technology have made it possible to bring sonography to the prehospital setting, facilitating early identification of life-threatening injuries that are likely to require surgical intervention, and providing an opportunity to alert trauma centres.

In this issue of BJS, Sewalt and colleagues4 used the Trauma Audit and Research Network (TARN) to investigate the ability of trauma models to predict mortality and major trauma in the prehospital setting. They found that most perform reasonably well in predicting in-hospital mortality, but are inadequate in identifying patients with major trauma. However, the early recognition of lifethreatening haemorrhage in individual patients allows for early initiation of blood product resuscitation. When haemorrhage control cannot be obtained in the prehospital setting, replacing ongoing blood loss during transport to the nearest trauma centre is invaluable. Prehospital massive transfusion prediction scores allow for early mobilization of blood products before arrival in the trauma centre and may determine the appropriateness of blood administration on route. These are important, as every 1-min delay in the arrival of initial blood products