Casualty evacuation

Why does a medical system that treats fewer casualties near the front save more of them?

The intuition is almost impossible to argue with: a wounded man needs a surgeon, so put the surgeon where the wounded men are. Every instinct we have about emergencies says close is better than far, and sooner is better than later.

Yet military medical systems are built the other way. The people nearest the fighting are deliberately given the least capability, and are trained to do very little and then let the casualty go. The full operating theatre sits hours to the rear. That looks like an accountant's decision dressed up as doctrine — so what would have to be true about wounds for it to be the right one?

Begin by refusing to treat "a wound" as one thing, and ask instead about each intervention separately: how long can it wait before it stops working, and what does it cost to have available?

Do that and something striking appears — the two questions have opposite answers. The interventions that save a life in the next few minutes are almost all cheap: pressure or a tourniquet on a limb bleed, a needle through the chest wall for a tension pneumothorax, rolling someone onto their side so their airway stays open. No electricity, no sterile field, no specialist. The interventions that are expensive — imaging, a vascular repair, an intensive care bed, reconstructive work on a shattered face — are almost all things that can wait hours or days without much changing the outcome.

That asymmetry is the whole design. If the urgent set were also the expensive set, capability would have to travel to the casualty and nothing else would work. Because they separate, capability can be stratified in space: put the cheap urgent things everywhere, in the hands of whoever is present, and keep the expensive patient things in one well-equipped place behind.

Now add the second constraint, which is about the forward node rather than the patient. A team near the fighting is small, mobile, and holds almost nobody. Its value is measured in casualties passed through per hour, not in depth of care. The moment it commits to a long operation it has stopped being an evacuation node — it cannot move, and everyone arriving behind that case waits. So a forward unit doing more is not a forward unit doing better; it is a forward unit that has blocked itself. Fewer treated near the front, in the sense the question means, is exactly what a system optimised for flow looks like.

There is a third piece. The middle of the chain does not do abbreviated definitive surgery — it does a different operation. Damage control surgery deliberately stops at controlling bleeding and contamination and leaves the anatomy unrepaired, because the patient at that point is dying of physiology, not of anatomy: cold, acidotic, and unable to clot. Warming and resuscitating him and coming back a day later beats a long, correct, immediate repair on a body that cannot survive it. The short operation was invented because the chain exists.

Here is the test. If echeloning works by separating cheap-and-urgent from expensive-and-deferrable, then the right amount of capability at any node should be a function of one variable: the transport time to the next node. Lengthen it — contested airspace, weather, a maritime or Arctic theatre — and surgical capability should be pushed forward, accepting the loss of mobility, because the deferrable set is no longer deferrable within that delay. Shorten it and capability should retreat rearward again. That is in fact how forward surgical teams are argued about. The refuting observation would be forward units that performed more definitive surgery producing better outcomes at the same evacuation times, or a doctrine that placed capability by casualty numbers rather than by delay and did no worse.

The misreading to dislodge is that this is triage. It is not. Emergency triage is a sorting rule applied at one place, where capacity is fixed and the only free variable is the order of a queue. Echeloned evacuation adds no ordering at all; it is an allocation of capability across places separated by transport time, and the free variable is where each capability sits. Triage cannot conjure a surgeon; echeloning cannot decide who goes first. In a real system both run at once, and each solves a problem the other cannot touch.

A refinement that joins the steps: the chain is not a conveyor. Casualties skip nodes when the tactical situation and the aircraft allow, and the great gains of the last two decades are usually attributed as much to bypassing echelons — flying directly to a surgical facility — as to what happens inside them.

Now the honest limits, and they are substantial. Almost none of this is trial evidence; you cannot randomise a casualty to a longer evacuation. The "golden hour" is better read as a planning doctrine than a measured threshold, and its direct evidential support is thin and argued over. Improvements in survival across a modern campaign coincide with better body armour, better tourniquets, better resuscitation and shorter flights all at once, so attributing them to the chain's structure specifically is an inference rather than a finding. I have also treated the forward set as fixed when its boundary is exactly what changes — blood products moving forward is a live argument about whether something expensive has become cheap enough to push down the chain. And casualty statistics are published by the same institutions whose doctrine is being evaluated, which is a reason for caution in both directions.

• How the boundary moves: whether blood transfusion has become cheap and portable enough to belong in the forward set.
• Its neighbour Emergency triage solves the other half — ordering a queue at one node under scarcity — while this piece only ever asks where the capability should stand.

Every term the collection defines is gathered in the glossary.