The Machine To Build the Station
Chris Hadfield was travelling at eight kilometres a second, 400 kilometres above the Earth. Beneath him, the planet raced silently past. Beside him waited fifteen years of engineering that would determine the future construction of the International Space Station.
In April 2001, the International Space Station was still a construction site. The great orbital laboratory familiar to us today did not yet exist. Major laboratories, habitation modules, trusses, solar arrays and scientific equipment still had to be delivered and assembled in space. None of it could be lowered into a hangar, placed on a workbench or returned to the factory when something went wrong.
The Station needed a permanent construction crane.
Canadarm2 was that machine. Seventeen metres long, capable of handling large payloads and designed to move end over end across the exterior of the growing Station, it had been designed, built and tested between 1986 and 2001. It would eventually assemble the majority of the ISS’s components and modules, move tonnes of equipment, support astronauts during dangerous spacewalks and capture visiting cargo spacecraft.1
But on April 22nd, 2001, Canadarm2 could do none of those things.
It was still cargo: folded in two, bolted to a metal pallet and secured outside the Destiny laboratory. Before the arm could contribute anything to the Station, it had to be unpacked, unfolded, connected and made operational. The future work of constructing and maintaining the ISS depended on first completing this work.
Hadfield and NASA astronaut Scott Parazynski emerged from the airlock of Space Shuttle Endeavour. Their immediate task was to install an ultra high frequency antenna, release Canadarm2 from its launch restraints, manually unfold its two long booms, secure its hinges, attach one end to Destiny and connect the cables that would allow the arm to communicate with the laboratory.1, 4
Every movement had been rehearsed underwater and in simulators. Every tool had been selected in advance. Every task had to be completed in a pressurised suit, wearing thick gloves, with no possibility of setting the machinery down and trying again later.
For Hadfield, it was the first spacewalk of his career—and the first ever performed by a Canadian.
The Station was moving at approximately 28,000 kilometres per hour, circling the Earth every 90 minutes.2 On one side of Hadfield’s visor, the world streamed past in colour and texture. On the other was an unfathomable blackness that appeared almost solid.
Outer space was not somewhere in the distance. It began a few centimetres beyond his face.
His spacesuit was not clothing. It was a one-person spacecraft. It supplied oxygen, removed carbon dioxide, controlled temperature and maintained the pressure necessary to keep him alive. Immediately beyond its protective layers lay death.
Hadfield steadied himself against the structure and continued the intricate work of bringing Canadarm2 to life.
Then something strange and completely unexpected happened.
His left eye began to sting.
The irritation intensified rapidly. His eye watered. Pain drove the eyelid shut. Within moments, Chris Hadfield could no longer see through his left eye—and he had no idea why.3
He had lost half his vision while performing delicate construction work, with only his spacesuit separating him from the vacuum of space. An unexplained problem had developed inside the sealed system keeping him alive. The mission was now in jeopardy, and Hadfield was in grave danger.
Nothing to Lose
On Earth, a cardiac arrest creates a different landscape but the same compression of time.
The patient is unresponsive. There is no pulse. Chest compressions begin. Defibrillator pads are applied. A rhythm is analysed. Shocks are delivered when indicated. Adrenaline is given. Every two-minute cycle is a wager against progressive cerebral and myocardial injury.
A life hangs in the balance. Decisions made under pressure may determine whether the patient regains circulation, survives to hospital discharge or returns to anything resembling their previous life.
Yet cardiac arrest encourages a peculiar distortion of reasoning. Because the patient has no beating heart, the team may begin to speak as though death has already completed its work.
“They are already dead. They cannot get any more dead.”
From that sentence, an entire therapeutic philosophy can follow. Give the extra drug. Try the untested intervention. Add something outside the algorithm. It might help—and what possible harm could it do?
This is the nothing-to-lose fallacy: the belief that a catastrophic starting point has removed all downside from the decisions that follow.
The error lies in the comparator. An intervention during cardiac arrest should not be compared with inevitable death. It should be compared with what might have happened under the best available evidence-based care.
Cardiac arrest is not irreversible death in every patient. Some regain spontaneous circulation. Some survive. Some return to their families with favourable neurological function. Resuscitation exists precisely because, for a proportion of patients, the boundary between life and death remains unsettled.
That remaining probability of recovery—however small—is what can be lost.
A treatment can increase myocardial injury, worsen an electrolyte disturbance, provoke arrhythmia, intensify cellular damage, delay a more effective intervention or distract the team from high-quality compressions, defibrillation and the treatment of a genuine reversible cause.
A pulseless patient may not be capable of becoming “more dead” as a figure of speech.
They are entirely capable of becoming less rescuable.
The Second Eye
Outside the Space Station, Hadfield still had one functioning eye.
Perhaps, he thought, that was why humans had two. He continued working.
But tears behave differently in microgravity. They do not fall down the cheek. They do not drain in the familiar way. The fluid remained over Hadfield’s closed eye, mixing with the unidentified contaminant and forming an enlarging sphere held together by surface tension.
The sphere grew.
It spread across the surface of his eye. It reached the bridge of his nose. Then, like a tiny waterfall moving sideways, the fluid crossed into his right eye.3
His right eye began to burn.
Then it closed.
Chris Hadfield was now completely blind outside the International Space Station.
The Canadarm2 installation was in jeopardy. Hadfield had become a potential casualty, tethered to the exterior of a spacecraft travelling around the Earth at enormous speed. The planet was far below him. Vacuum lay immediately beyond his visor. He could not rub his eyes, rinse them or open his helmet.
The situation that had seemed serious when one eye failed had become catastrophic.
The Ampoule That Could Not Hurt
As a cardiac arrest continues, the clinical team experiences a similar narrowing of options.
The recognised treatments have been delivered. The rhythm remains unchanged. There is still no pulse. The pressure to move beyond the evidence becomes almost physical.
Calcium was perfectly suited to that moment.
Its physiological rationale was immediately attractive. Calcium is fundamental to excitation–contraction coupling. It permits actin and myosin to interact, increases myocardial contractility and maintains vascular tone. When the monitor shows organised electrical activity without a palpable pulse, giving calcium can feel less like speculation than the correction of an obvious deficiency.
The drug was familiar, biologically active and readily available. It offered the satisfaction of doing something when the accepted algorithm seemed to have run out of answers.
There was even a faint historical signal. Two small randomised, double-blind trials published in 1985 studied calcium in refractory pulseless electrical activity and asystole. Together, they reported return of circulation in 11 of 87 patients given calcium and three of 76 given placebo. The difference did not reach conventional statistical significance. Only one patient across both trials survived to hospital discharge—and that patient had received placebo.7, 8
The evidence was weak, but the story was compelling. Calcium was essential to normal cardiac contraction; therefore, more calcium might help a heart that was failing to contract.
Beneath its empirical use sat the nothing-to-lose assumption: even if calcium did not help, surely it could not make a pulseless patient worse.
The Calcium for Out-of-Hospital Cardiac Arrest trial—COCA—put that assumption to the test.
The COCA Trial
COCA was an investigator-initiated, randomised, double-blind, placebo-controlled trial conducted in the Central Denmark Region. Adults with out-of-hospital cardiac arrest were eligible after receiving at least one dose of adrenaline. Participants were assigned to receive up to two intravenous or intraosseous doses of 5 mmol calcium chloride or identical saline. The first dose was administered immediately after the first dose of adrenaline; a second was given after the second adrenaline dose if cardiac arrest continued.5
Importantly, this was a trial of routine, empirical calcium. Patients with a recognised clinical indication for calcium, such as suspected hypocalcaemia or hyperkalaemia, were excluded. COCA did not ask whether calcium should be used to correct a specific electrolyte or toxicological emergency. It asked whether calcium improved outcomes simply because an adult cardiac arrest was ongoing.
The investigators aimed to recruit 674 patients. At a planned interim analysis involving 383 participants, however, the independent data-monitoring committee identified a consistent signal of harm in the calcium group and recommended that the trial stop. Ultimately, 397 patients had been randomised and 391 were included in the principal analysis.5, 6
The results challenged the idea that there was nothing left to lose.
Sustained return of spontaneous circulation occurred in 37 of 193 patients assigned calcium (19%), compared with 53 of 198 assigned saline (27%). The risk ratio was 0.72, with a 95% confidence interval from 0.49 to 1.03.
Survival at 30 days occurred in 10 patients assigned calcium ( 5.2%), compared with 18 assigned saline (9.1%). The risk ratio was 0.57, with a 95% confidence interval from 0.27 to 1.18.
Survival at 30 days with a favourable neurological outcome occurred in seven patients assigned calcium (3.6%), compared with 15 assigned saline (7.6%). The risk ratio was 0.48, with a 95% confidence interval from 0.20 to 1.12.5
The observed outcome was worse with calcium at every clinically important point. Fewer patients regained sustained circulation. Fewer survived. Fewer survived with favourable neurological function.
None of these three comparisons crossed the conventional threshold of P < 0.05 in the final analysis. The trial had also stopped before reaching its planned sample size, and the confidence intervals were wide. Those limitations matter. COCA cannot establish the exact magnitude of harm with precision, and an early-stopped trial can overestimate a treatment effect.
But “not statistically significant” does not mean “safe”. It certainly does not mean “nothing to lose”.
The prespecified Bayesian analyses make the evidential direction clearer. Using non-informative priors, the posterior probability that calcium caused harm was 96% for sustained return of circulation, 94% for 30-day survival and 96% for favourable neurological outcome at 30 days. The corresponding probabilities of any benefit were only 4%, 6% and 4%.5
Nor had the intervention simply failed to alter physiology. Among patients with measured calcium concentrations after return of circulation, hypercalcaemia occurred in 74% of those assigned calcium and 2% of those assigned saline. The first measured ionised calcium concentration averaged 1.41 mmol/L with calcium and 1.17 mmol/L with saline.5, 6
The drug had done what pharmacology predicted. It had raised calcium.
That may have been part of the problem.
Ischaemic myocardium is not healthy myocardium waiting to be stimulated harder. During cardiac arrest, depletion of adenosine triphosphate disrupts ionic gradients. Intracellular sodium increases, the sodium–calcium exchanger may operate in reverse, and calcium can accumulate within the cytosol and mitochondria. An additional calcium load may intensify calcium overload, oxidative stress, mitochondrial injury and myocardial hypercontraction—the phenomenon sometimes described as “stone heart”.5
These mechanisms are plausible explanations rather than proof of the exact pathway through which calcium might cause harm. But they expose the weakness in the original intuition. A substance essential to normal contraction need not be beneficial when delivered into cells already overwhelmed by ischaemia.
More of a necessary ion is not necessarily more life.
COCA did not prove that calcium harms every patient in every cardiac arrest. It did something more directly relevant to the nothing-to-lose fallacy: it showed that an intervention given to patients thought to be beyond further harm could plausibly consume their remaining chance of rescue.
It Can Always Get Worse
Outside the International Space Station, Chris Hadfield’s world was completely black.
This was the point at which panic might have seemed not only understandable but inevitable. Yet Hadfield had been trained within a culture organised around a brutally practical warning:
“There is no problem so bad that you can’t make it worse.”3
The maxim does not deny the seriousness of the original problem. It insists that even a catastrophic situation contains choices of different quality. Disaster does not make every response reasonable. Urgency does not transform improvisation into safety.
Hadfield did not thrash against the Station. He did not begin a blind, uncontrolled retreat. He did not try to interfere with the helmet protecting him from vacuum. The instinctive responses available on Earth—rub the eye, rinse it, lift the visor—were either impossible or potentially lethal.
Instead, his training took over.
He could not see, but he could hear. He could speak to Mission Control. Scott Parazynski was outside with him. They had repeatedly practised incapacitated-crew rescue. If Hadfield could not recover, Parazynski could guide him or physically return him to the airlock. Hadfield could also orient himself by touch and by his detailed knowledge of the Station’s structure.3
He took stock of what had been lost—and, just as importantly, what remained.
His calm was not passivity. It was controlled action. He communicated, maintained his position and allowed the team to work the problem without creating a second emergency.
The Discipline of Resuscitation
Cardiac arrest is precisely the circumstance in which clinicians are most tempted to invert Hadfield’s rule.
Because the patient has no pulse, the evidential threshold is allowed to collapse. Desperation becomes an indication. Physiological plausibility becomes proof. Visible action becomes indistinguishable from useful action.
The human impulse is understandable. Standing beside a patient in cardiac arrest while the monitor remains unchanged creates an almost irresistible need to do more. Repeating the established interventions can feel inadequate. Reaching for another ampoule feels decisive.
But the severity of the situation does not make an untested treatment safer. It makes the consequences of error more important.
A patient in cardiac arrest exists at the very limits of physiological tolerance. There is no reserve with which to absorb an avoidable metabolic insult, a pro-arrhythmic effect, additional myocardial injury or distraction from higher-value care. That absence of reserve should make the team more disciplined, not less.
The history of cardiac-arrest pharmacology is, in part, a history of subtraction. Atropine was removed from the routine treatment of pulseless electrical activity and asystole in 2010. Vasopressin was removed from the adult cardiac-arrest algorithm in 2015 because it offered no advantage over adrenaline. Routine magnesium and sodium bicarbonate are not recommended, and routine calcium is now specifically recommended against.9, 10, 12
These interventions were not removed or demoted because resuscitation clinicians became less committed to saving lives. Some disappeared because trials failed to demonstrate meaningful benefit. Others carried evidence or credible signals of harm. In several cases, an attractive physiological effect failed to translate into a better patient-centred outcome.
The distinction is crucial. An intervention can raise a blood pressure, alter an electrocardiogram, increase a laboratory value or even produce a transient pulse—and still fail to improve survival or neurological recovery.
Some drugs remain appropriate for specific causes of cardiac arrest. Calcium may still be indicated when a relevant electrolyte disturbance or toxicological emergency is genuinely suspected. Sodium bicarbonate has defined roles in selected poisonings. Magnesium remains appropriate for particular arrhythmias. The problem is not targeted treatment of an identified cause. It is empirical treatment justified solely by desperation.
A systematic review following COCA found no evidence that routine calcium improves survival or neurological outcomes in adult or paediatric cardiac arrest, and contemporary guidelines advise against its routine administration.11, 12
Absence of reliable evidence is not evidence of harmlessness. An unknown treatment effect is not a zero treatment effect.
The answer is not therapeutic nihilism. It is disciplined prioritisation: high-quality chest compressions, timely defibrillation, effective oxygenation and ventilation, appropriate adrenaline, and the identification and treatment of a genuine reversible cause.
The team must do what is supported, do it well and resist filling every apparent gap in the algorithm with an ampoule.
Vision Returns
Hadfield remained outside the International Space Station.
As his eyes continued to water, the tears gradually diluted the contaminant. The fluid began to clear. His vision started to return—blurred at first, but sufficient for him to reorient himself and assess what was happening.
He remained in communication with his colleague and Mission Control. Once it was clear that he could see adequately and continue safely, Houston allowed the spacewalk to proceed.3
Hadfield and Parazynski completed every planned task. During a spacewalk lasting 7 hours and 10 minutes, they installed the ultra high frequency antenna, unfolded Canadarm2, connected its communications cables and secured its booms.4
When Hadfield returned inside, the residue around his eyes was examined. The cause was anti-fog compound—a mixture of oil and soap—that had entered his eye. It had been painful, incapacitating and frightening, but it had not been the catastrophic spacesuit failure that the sudden blindness might initially have suggested.3
Canadarm2 was successfully brought into service. It went on to assemble the majority of the International Space Station’s components and modules and became indispensable to its maintenance and resupply.1
Hadfield succeeded not because he did nothing, and not because he did everything. He succeeded because he continued to distinguish useful action from dangerous action after the situation had become frightening. He declined improvisation and stuck with protocol. He trusted what had already been tested.
That is the discipline required during cardiac arrest.
The patient without a pulse is not beyond harm. They possess a finite, fragile probability of recovery. Every intervention must be judged by what it does to that probability—not by how satisfying it feels to administer.
The nothing-to-lose fallacy mistakes a terrible prognosis for no prognosis. It assumes that because the patient is close to death, the downside of treatment has disappeared. COCA demonstrated the danger of that assumption. An intervention that was familiar, physiologically persuasive and emotionally irresistible was associated with fewer patients regaining circulation, fewer surviving and fewer recovering neurologically.
The lesson is not that clinicians should hesitate when action is required. It is not to avoid an indicated therapy for a specific pathology, which happens to be outside the cardiac arrest algorithm. It is that urgency must sharpen discrimination. Cardiac arrest is the clinical situation in which calm, discipline and fidelity to evidence matter most, because the patient has no reserve to absorb our mistakes.
Travelling at eight kilometres a second, on the edge of space, Chris Hadfield understood that there was still something to lose.
Beside a pulseless patient, we must understand the same.
There is no problem so bad that it cannot be made worse.
References
1. Canadian Space Agency. The history of Canadarm2. Government of Canada; 2018.
3. Hadfield C. What I learned from going blind in space. TED2014; March 2014.
4. National Aeronautics and Space Administration. STS-100. NASA.
Related Materials
- This blog was written with the assistance of AI