What the two people on the bank can see is a pair of skis.
Everything above the boots is under eight inches of stream ice, in moving water, and she cannot come back out the way she went in.
So they hold her legs and wait for help, seven minutes to call and over an hour to arrive.
Somebody chops a hole downstream and she comes out through that.
The useful question is not how long she was under. It is what order things happened in.
The air pocket is the whole reason this worked
Cold protects a brain by slowing it down. Cerebral metabolic rate falls about three and a half percent per degree Fahrenheit, and by 82 degrees oxygen use and pulse are down by half.
But that only helps if the cooling gets there first.
She went in head down and found a gap between the ice and the water. For roughly the first forty minutes she had an airway and was breathing.
Forty minutes of that, in water just above freezing, takes a body temperature a very long way down while the brain is still supplied.
Only then did her circulation stop.
Reverse those two and it ends differently. Asphyxia first means the brain loses oxygen while still warm and still hungry, which is why a drowning at the same temperature carries a far worse outlook.
The mountain and the flight
She was 29, a Swedish trainee doctor at the hospital in Narvik, skiing off piste on a May afternoon.
She lost control on a steep slope and went through into a gully with a stream under it.
Her two companions were doctors too, which mattered less than it sounds. There was nothing to do but hold on and keep her from going under.
She came out at 19:40, eighty minutes after the fall.
A rescue helicopter flew her 155 miles north to the university hospital in Tromsø, roughly an hour, arriving at 21:10.
Note the hospital. Tromsø treated her and employs her today. The place she worked in 1999 was the smaller one she was flown out of.
The timeline, and which parts are published
The figure everyone knows, 56.7 degrees Fahrenheit, is real.
It comes with a caveat. A recent case report on a comparable patient reads the original as giving that as the low point during afterdrop, with a rectal reading of 57.9 degrees on admission.
Bypass started at 21:40 and the first spontaneous heartbeats came at 22:15, so circulation had been absent over two and a half hours, about thirty five minutes of that on the machine.
Then 35 days on a ventilator, a tracheostomy, 28 days in intensive care and two months of it in total. She woke around ten days in and was back at work that October.
One note on sourcing. The temperature and the survival are in the journal. The minute by minute times, and every figure circulating about units of blood, come from press reconstructions, and the transfusion counts appear nowhere checkable.
What recovered turns out to mean
She was initially paralysed from the neck down, and what never fully came back is peripheral.
Lasting nerve damage in her hands and feet, reduced sensation and motor function, wasting in the hand muscles. A decade later she still had symptoms.
The consequence is specific. She had been training toward orthopaedic surgery and could not finish, because of the hands. She became a radiologist instead.
So no gross cognitive deficit, near complete function after years of rehabilitation, and a career changed by the injury, as the case report and the coverage after it describe. Brain intact is doing a lot of work.
Two other things get repeated wrongly. The nine hours belongs to the operating theatre effort, a hundred staff in shifts, not to the wait before a heartbeat returned. And hers is no longer the lowest on record. In 2014 a toddler in Poland came back from 53.2 degrees and was normally developed at a five year check.
Hers is the lowest in an adult. Extreme physiology gets flattened this way, whether it is a supervised fast or how long freeing it took.
What the case actually changed
It did not invent rewarming on a heart lung machine; a Swiss series came two years earlier.
What it destroyed was the lower bound as a reason to stop.
The clearest evidence is the hospital’s own record. Of 34 hypothermic cardiac arrest patients rewarmed on extracorporeal circulation there from 1985 to 2013, there were no survivors before 1999. After it, nine of 24.
Current European guidance calls the old potassium threshold less reliable and points to a six variable score, which beat potassium alone and would have avoided 27 futile rewarmings without losing a survivor.
The list of reasons not to try is short now, and none of them is a number on a thermometer.
Which is the useful version of the old rule. It is not that cold water is survivable. It is that a cold body cannot be assessed until it is warm.
