Delayed emergence is worked through in order

Why it matters

A patient who does not wake is the PACU emergency that is most often mishandled by being handled in the wrong order. The list of causes is long and the dangerous ones are not the common ones, so the sequence matters more than the list. The simulation case used to teach this groups the causes as residual anesthetics, pharmacological actions, surgical complications, neurological events, endocrine disturbances and patient factors ellis-2017. The two drugs reached for first, naloxone and flumazenil, each carry a label warning that bears directly on the PACU: naloxone can be outlasted by the opioid it reverses, and flumazenil can provoke seizures in exactly the patients most likely to be slow to wake from a benzodiazepine fda-naloxone-label-2023 fda-flumazenil-label-2025.

The consequence of getting this wrong runs two ways: a reversal agent given for a cause it does not treat, or a stroke found after an hour of waiting for the anesthetic to wear off.

Mechanism

Emergence requires that the airway is open and gas is moving, that the circulation is delivering it, that the brain is fuelled, that the muscles can obey, that the drug effect at the brain has fallen below the threshold for consciousness, that the brain is warm enough and its electrolytes are within the range it tolerates, and that the brain itself is intact. Each of those is a separate mechanism with a separate test, and each can look like the others from the end of the bed.

The order the day teaches follows danger first, then frequency and reversibility. First, the things that kill in minutes and are already on the monitor: oxygenation, ventilation, and circulation. Second, glucose, because a meter answers in seconds. Third, residual paralysis, because a quantitative monitor answers in seconds and the patient cannot tell you. Fourth, residual drug: volatile, opioid, benzodiazepine, and the anticholinergic syndrome. Fifth, temperature, sodium and the other electrolytes. Last, because it is the rarest and the least reversible, the intracranial event. Putting oxygenation first is not a claim about what is commonest; it is a claim about what a delay costs, and it is the same order every resuscitation framework teaches. The order is expert consensus, not a trial result, and the status chip says so.

Central anticholinergic syndrome sits inside the drug tier: a central effect of antimuscarinic drugs that presents as somnolence or agitation and responds to physostigmine link-1997.

Evidence

The evidence for the differential is descriptive. Link 1997 followed 962 recovery-room inpatients for two months and, after excluding prolonged anesthetic or relaxant effect, respiratory depression and metabolic disorder, found a distinct central anticholinergic syndrome in 18 (1.9%), with 6 of 60 after hysterectomy (P=0.003); untreated somnolence had lasted more than 2 h in 6 of the 18, all woke after physostigmine, and 6 relapsed and needed a second dose, one a third link-1997.

A 2023 cross-sectional study describes the distribution of normal emergence, delayed awakening, hypoactive emergence and emergence delirium and the factors associated with delay; several of its reported confidence intervals do not contain their own point estimates, so this page uses it for the shape of the differential and quotes none of its numbers bayable-2023.

The naloxone label states that, since the duration of action of some opioids may exceed that of naloxone, the patient should be kept under continued surveillance and repeated doses given as necessary; in one study the adult serum half-life ranged from 30 to 81 min (mean 64 min); for postoperative opioid depression the dose is increments of 0.1 to 0.2 mg intravenously at two to three minute intervals to the desired degree of reversal; and abrupt reversal may result in nausea, vomiting, sweating, tremulousness, tachycardia, increased blood pressure, seizures, ventricular tachycardia and fibrillation, pulmonary edema and cardiac arrest fda-naloxone-label-2023.

The flumazenil label carries a boxed warning: the use of flumazenil has been associated with the occurrence of seizures, most frequent in patients who have been on benzodiazepines for long-term sedation or in overdose cases showing signs of serious cyclic antidepressant overdose fda-flumazenil-label-2025. For reversal of sedation in adults the dose is 0.2 mg intravenously over 15 s, with further 0.2 mg doses at 60-s intervals up to four additional doses and a maximum cumulative dose of 1 mg; re-sedation occurred in 3% to 9% of conscious-sedation studies and the terminal half-life is 40 to 80 min fda-flumazenil-label-2025.

Physostigmine has two current labels in the United States and they do not agree, which matters because this is the drug for the central anticholinergic syndrome above. The approved product names the indication in as many words, listing central anticholinergic syndrome, delayed postoperative awakening and shivering, and doses by weight: 0.04 mg/kg injected slowly at about 1 mg/min, a maximum single dose of 2 mg, and a repeat after 5 to 20 min at the earliest; it warns of bradycardia and convulsions if given too fast, and asks for continuous electrocardiographic monitoring where tricyclic overdose is the indication fda-physostigmine-anticholium-label-2023. The other product is marked an unapproved drug and its label carries the statement that it has not been found by the FDA to be safe and effective; its dose is 0.5 to 1.0 mg intramuscularly or intravenously at no more than 1 mg per minute, and it asks that atropine always be at hand as the antidote fda-physostigmine-akorn-label-2022. The 0.5 to 2 mg figure this page used to carry matched the unapproved label at its lower end and neither label at its upper end.

What this does not show

No source on this page tests the ordered differential against any other order, or against no order. The sequence is expert consensus about frequency and reversibility, and the simulation case that carries it is an educational resource, not a study ellis-2017. The labels establish the pharmacology of the two reversal agents; they say nothing about how often each cause occurs in a PACU.

The requirement that a neurological examination be performed and documented before the patient leaves the PACU is on the content list for this day and it is not in the sources. Both American Society of Anesthesiologists documents were opened and searched on 2026-09-08. The standards ask that the patient be observed and monitored by methods appropriate to their condition, with particular attention to oxygenation, ventilation, circulation, level of consciousness and temperature, and that an accurate written report of the recovery period be maintained, encouraging a scoring system to assist with management and discharge readiness asa-pacu-standards-2024. The practice guideline asks that mental status be periodically assessed during emergence and recovery asa-pacu-2013. Neither requires a documented neurological examination, neither ties any assessment to the discharge decision, and neither mentions delayed emergence at all. Level of consciousness appears as one parameter among several, and the discharge criteria themselves are delegated to the local department.

The anesthetic-specific numbers the list expects, how long each volatile agent or opioid delays emergence in a given patient, are not in the held sources, so this page gives none. A published study of physostigmine after desflurane has been retracted and is not cited here.

At the bedside

Work the order and say it aloud: oxygen and carbon dioxide, circulation, glucose, paralysis, drug, temperature and sodium, brain. Look at the saturation, the capnograph and the chest before anything else, and take a blood gas if the carbon dioxide is a question. Each step has a test that takes less time than waiting: the oximeter and capnograph, the blood pressure, the glucose meter, the twitch monitor, the drug chart, the thermometer, the electrolytes. Do not skip to the end, and do not stop before it.

If you give naloxone, you have committed to watching the patient for longer than the naloxone lasts; the label says so, and the re-narcotized patient is the one who has been handed to the ward in the meantime fda-naloxone-label-2023. Give it in the label’s small increments rather than a full ampoule, because abrupt reversal has its own list of harms fda-naloxone-label-2023.

Before you give flumazenil, ask what benzodiazepines the patient takes at home. The boxed warning is about chronic use, and the answer changes the decision fda-flumazenil-label-2025.

When the commoner causes are excluded and the patient remains somnolent or agitated after antimuscarinic drugs, think of central anticholinergic syndrome; it responds to physostigmine and it can relapse link-1997. Local practice differs. the availability of physostigmine, and the local threshold for imaging a patient who has not woken, differ between institutions.

General recommendations

These are the numbers residents are commonly taught and expected to know. This project has not yet read them from a primary source, so each is labeled and none carries a ledger key. Where one conflicts with your institution’s protocol, the protocol wins.

One taught figure is worth correcting here, because it is a dose. The small naloxone increment that residents learn for avoiding abrupt reversal is not in the labelling: the adult postoperative increment there is 0.1 to 0.2 mg, cited above, and the only smaller increments in that label are pediatric fda-naloxone-label-2023. A search for a primary source found exactly one, and it does not support the range as it is usually taught. It is a retrospective case series of 15 patients treated in an emergency department on toxicology advice, 12 of them after ingesting methadone, started at 0.04 mg intravenously and titrated to a median total of 0.08 mg with an observed range up to 0.12 mg, and 2 of the 15 developed acute opioid withdrawal kim-2016-naloxone. Its own conclusion is limited to emergency department patients with moderate to severe opioid-induced ventilatory depression kim-2016-naloxone. So the starting figure has a source, in the wrong population and the wrong setting, with no control group; the upper bound usually taught alongside it matches nothing; and no society guideline recommends an adult increment of this size at all.

Check yourself

Three items. Every option carries an explanation. Progress is not saved.