Two episodes of delayed emergence in a healthy young man

Delayed emergence from anaesthesia is rare in healthy patients. We describe two episodes in a fit 44-year-old man: he did not speak and barely moved for 3 h after elective gastrointestinal endoscopy and for 8 h after elective inguinal hernia repair. Both episodes responded to unconventional use of reversal agents. Each required admission to our intensive care unit (ICU) overnight with full recovery at 24 h. The patient consented to this report.

Case 1

An 85 kg 44-year-old man presented for gastroscopy and colonoscopy. This was his first anaesthetic. He took no medication, did not smoke or vape, drank minimal alcohol, and had no family history of anaesthetic problems.

After connection to oxygen, intravenous fluids, and monitoring (non-invasive blood pressure, pulse oximetry and capnography), he was sedated with alfentanil boluses (0.45 mg total) and propofol by Marsh target-controlled infusion (initial target plasma concentration 6 μg/mL, rapidly weaned to 2.5 μg/mL during colonoscope withdrawal). No other drugs were given. The procedure took 35 min, the total propofol dose was 450 mg, and observations were stable throughout.

He was slow to wake in the recovery room. Observations remained stable and he maintained his own airway but still appeared deeply sedated after one hour—moving minimally and not talking. Naloxone (400 μg incrementally) had no effect. Temperature, blood sugar level, and a venous blood gas were normal.

A neurologist attended and gave flumazenil (0.5 mg incrementally). This improved arousal but did not resolve symptoms (at 2 h he was still not talking). Given the flumazenil response and absence of focal signs, the neurologist recommended against imaging but advocated observation in ICU overnight. At 3 h the patient began to talk and the next morning he was back to baseline.

Case 2

The patient returned 3 months later for laparoscopic inguinal hernia repair. He remained medically well.

Drugs used previously were avoided. Standard monitors, bispectral index and a neuromuscular transmission monitor were applied. Anaesthesia was induced with fentanyl 2 μg/kg, thiopentone 5 mg/kg and maintained with 0.8 minimum alveolar concentration sevoflurane in oxygen and air. Vecuronium 10 mg was used for muscle relaxation. Other drugs given were dexamethasone 8 mg, ondansetron 4 mg, parecoxib 40 mg, paracetamol 1 g, and sugammadex 200 mg; the surgeon also infiltrated 200 mg of ropivacaine under direct vision.

Surgery lasted 65 min and was uncomplicated. Bispectral index readings were between 30 and 50 with an unremarkable trace. After reversal (train-of-four ratio 1.0), he was extubated when appearing to reach for the tracheal tube.

The course in the recovery room then was similar to case 1: observations were stable, he opened eyes slightly to voice but did not verbalise or move limbs. Temperature, blood gas and glucose results were normal. Naloxone (400 μg) and flumazenil (0.5 mg incrementally) had no effect.

He was transferred to the ICU for monitoring. Thyroid and renal function tests done on arrival were normal. Over 6 h he became more alert and began to answer yes/no questions by way of blinks but still did not move or vocalise. We then conducted a therapeutic trial of physostigmine. Within 5 min after slow incremental administration of 2 mg, he spoke (his first word was ‘Finally!’). Recovery continued overnight and by the next morning he had returned to baseline.

The patient stated ‘For case 1 I felt relaxed coming out of the surgery but couldn’t move. I could see and hear and knew where I was, and people helped me remain calm. Case 2 was similar but coming out of anaesthetic took much longer. I got upset and frustrated that my body didn’t cooperate. It helped that my wife was present as I communicated by blinks (one for ‘yes’ and two for ‘no’).’

The patient’s wife also commented: ‘It was hard to see my husband unable to respond and a shock for him to be in ICU. The ICU doctor seemed to assume it was psychological. I was glad that the anaesthetist tried a new medicine which worked.’

We approached authors of recent relevant papers^1,2 for advice but they had no additional tests to suggest. We also consulted a pharmacogenomics expert. He considered a pharmacokinetic cause unlikely given the range of drugs used across the two episodes (T. Polasek, personal communication).

The patient has a detailed letter for future anaesthetists. If general anaesthesia is unavoidable, our recommendations are to warn about recurrence, book an ICU bed, operate early in the day, and have physostigmine readily available. It may also help to minimise anaesthetic doses with adjuvant regional blocks, nitrous oxide and remifentanil or other propofol-sparing alternatives. ³ Finally, all recovery room staff should be informed that the patient may be aware even when unresponsive so that they communicate with him accordingly.

Two points about these cases are noteworthy: the unconventional application of reversal agents and the patient’s recall of events when he was still unable to respond.

In case 1 the attending neurologist suggested flumazenil even though no benzodiazepine had been given. Flumazenil is recommended for delayed emergence after benzodiazepines,^1,4 but it can also accelerate recovery after unpremedicated propofol ⁵ and inhalational ⁶ anaesthesia.

In case 2 the apparent antidote was physostigmine. This drug has been shown to accelerate recovery after propofol, ketamine and halothane ⁷ but not sevoflurane^8,9—although the sevoflurane studies were small and not in the context of delayed emergence. Potential side-effects include nausea and salivation (which our patient had), and bradycardia, seizures and late symptom recurrence. ¹⁰ The time course of administration (slow and incremental) and response (minutes after the final increment) appeared unlikely to be a coincidence or placebo effect.

Finally, we note our patient’s account. It seems that elements of consciousness were disconnected during his emergence: ² awareness (receptive function) recovered earlier than responsiveness (verbal and motor function). This state was temporary but caused a distressing experience akin to a ‘locked-in’ syndrome. We should communicate directly with unresponsive patients in such cases: they may be aware and in need of our reassurance.