Where does anaesthesia take you?

Anaesthetic agents act on movement, memory and awareness through partly separate routes. A patient can be immobile and amnesic while some form of experience continues. Sleep shows the reverse: people who sleepwalk respond to their surroundings without being conscious.

The isolated forearm technique keeps one arm free of the paralysing drug and asks the patient to squeeze a hand. In the older studies that did this during surgery or laryngoscopy, a median of 37 per cent of patients responded. The range ran from none of them to all of them, and the review that collected them declined to pool them. Two multicentre studies since have found fewer: 4.6 per cent of 260 adults after intubation in routine practice, and 11 per cent of 338 aged 18 to 40. In the first, none of the 253 asked after surgery recalled anything from during it.

Two anaesthetists, a psychiatrist and a neurologist made this argument in an anaesthesiology journal in 2012, against the assumption that a patient who does not respond is a patient who is not there. Absence of response is absence of evidence about experience.

A magnetic pulse delivered to the cortex produces a pattern of activity that echoes across the brain. In a conscious person the echo is complex: widespread, differentiated, long-lasting. In dreamless sleep, or under propofol, midazolam or xenon, it either stays short and local or spreads as one stereotyped wave. Under ketamine, after which people report vivid experiences, it stays complex.

Compressing that response into a single number gives an index. In the 2013 study that introduced it, 32 healthy subjects were measured awake and 24 of them again asleep or anaesthetised: nothing unconscious scored above 0.31 and nothing awake below 0.44. Two locked-in patients, who could answer only by moving their eyes, scored in the waking range. That study called itself hypothesis-generating. In 2016 a benchmark of 150 people, each able to report at the time or afterwards, gave a cutoff, again 0.31, that separated reported experience from its absence without an error in either direction, and it detected 94.7 per cent of minimally conscious patients.

It is a research instrument, not a monitor. The depth gauges used in surgery read the spectral content of the EEG, which is a related quantity and not the same one. They vary between people and between agents, and they have not reliably told a conscious brain-injured patient from an unconscious one.

Reported awareness during general anaesthesia is rare, and how rare depends on how the question is asked. A structured interview of 19,575 patients at seven American academic centres found 25 cases, roughly one in 800. A British national audit that counted the reports made to it found about one in 19,600.

Memory is the confounder. An agent that prevents encoding produces a patient who reports nothing regardless of what occurred, which means the honest quantity being measured is remembered awareness.

The molecular targets are known in some detail, down to anaesthetic binding sites resolved by structural biology. The step from a receptor to the disappearance of experience is not, and different agents reach the same endpoint by different routes.

That gap is the reason anaesthesia is interesting rather than merely useful. It is a switch whose mechanism nobody can yet describe in full.

Dreamless sleep, anaesthesia and the gap between are usually described as nothing happening. The evidence for that rests on the absence of a report afterwards, which is exactly what an agent that blocks memory would produce. Even the complexity index was calibrated against those reports, and its authors warn that people can be conscious and then forget.

No report made afterwards can tell an experience that did not occur from one that left no trace, and that limit is permanent. A hand that squeezes shows that someone was there. Nothing yet shows that no one was.