Design: Retrospective cohort with expert panel review of cause of death
Population: 3,089 adult trauma patients attended by London's Air Ambulance, 2019–2020; 497 deaths
Primary outcome: Timing and cause of trauma death
Effect: 77.1% of deaths prehospital; median 12 minutes to arrest; 96.2% of haemorrhage deaths prehospital

Researchers reviewed every adult trauma death attended by London's Air Ambulance in 2019 and 2020. A multidisciplinary panel assigned cause of death using clinical, imaging and post-mortem findings.

Of 3,089 patients attended, 497 died. Over three-quarters of deaths (77.1%) were prehospital, with a median of 12 minutes from injury to traumatic cardiac arrest. Haemorrhage and brain injury caused 82% of all deaths, but they split by phase: 96% of haemorrhage deaths happened before hospital, while 84% of in-hospital deaths were from brain injury. Nearly all penetrating trauma deaths were prehospital.

Hospital-based registries miss most of these deaths, which is why they can suggest bleeding is a solved problem. For Indian settings with long transfer times, the lesson is stronger still: bleeding control must start at the scene and on the road.

Recognise that hospital trauma registries miss most haemorrhage deaths.
Support training in tourniquets, pelvic binders and direct pressure for first responders.
Push for early blood and tranexamic acid as close to the scene as possible.
Expect brain injury to dominate deaths among patients who reach hospital alive.

Why it matters

It moves the target for improving trauma survival from the operating theatre to the roadside.

The statistics, in plain English

These are simple proportions from one helicopter service in one city, so they describe a mature, rapid system; in slower systems even more deaths are likely to occur before hospital.