Airway obstruction is usually obvious, and the rest presents as desaturation

Why it matters

Airway obstruction in the recovering patient is the emergency the PACU exists for, and it does not always announce itself. In the Australian Incident Monitoring Study, laryngospasm was clinically obvious in 77% of reports; in the remainder it presented as airway obstruction in 14%, as regurgitation or vomiting in 5%, and as desaturation alone in 4% visvanathan-2005-laryngospasm. The companion analysis of natural-airway obstruction carries the same warning in a different setting: obstruction may present simply as desaturation visvanathan-2005-airway.

The consequence of getting this wrong is not the obstruction, which usually resolves. It is what obstruction does on the way to resolving: post-obstructive pulmonary edema, aspiration and severe hypoxemia are named by the AIMS authors as the morbidity of poorly managed laryngospasm visvanathan-2005-laryngospasm.

Mechanism

Laryngospasm is a reflex closure of the glottis, triggered by stimulation of the larynx during light planes of anesthesia, most often at emergence. The tongue obstructs by falling back against the posterior pharynx when pharyngeal muscle tone is lost, which is the common mechanism of obstruction in the recovering patient who is not yet awake. Both produce the same downstream physiology: a patient making strong inspiratory efforts against a closed or narrowed upper airway generates a large negative intrathoracic pressure, which draws fluid across the pulmonary capillary into the alveolus. That is negative-pressure pulmonary edema, and it can follow either mechanism. Meanwhile the obstructed patient desaturates, and the unprotected airway is exposed to whatever is in the pharynx.

Evidence

The AIMS laryngospasm analysis reviewed 189 reports among the first 4000 incidents reported to the registry visvanathan-2005-laryngospasm. Desaturation occurred in over 60% of cases, bradycardia in 6% (23% in patients under one year), pulmonary edema in 4% and pulmonary aspiration in 3% visvanathan-2005-laryngospasm. The authors state that, had a structured approach been used, earlier recognition or better management may have occurred in 16% of cases visvanathan-2005-laryngospasm. The Scandinavian denominator study supplies the incidence: 1232 laryngospasms in 156,064 anesthetics in 136,929 patients, about eight per thousand anesthetics, with rates above 50 per 1000 in children with asthma, an airway infection, a previous anesthetic complication, esophagoscopy or hypospadias repair, and above 25 per 1000 at age 1 to 3 months olsson-1984. The AIMS paper cites the same study at 0.78% visvanathan-2005-laryngospasm.

The companion AIMS analysis of natural-airway obstruction supplies the observation that obstruction may present simply as desaturation, which is why a falling saturation in a recovering patient is an airway finding until the airway has been looked at visvanathan-2005-airway.

A 2025 series describes five cases of negative-pressure pulmonary edema during recovery from general anesthesia at one institution: four attributed to glossoptosis after extubation and one to laryngeal mask displacement, all with typical clinical and imaging features, and none with a poor outcome bmc-nppe-2025. It carries the point that in the recovery period the precipitant is not always laryngospasm.

What this does not show

Every source on this page is an incident registry or a case series visvanathan-2005-laryngospasm visvanathan-2005-airway bmc-nppe-2025. Registry data describe how events presented and what followed them among reported cases; they do not give a true incidence, because reporting is voluntary, and they do not test any management. The 16% figure is the authors’ judgment about what a structured approach might have changed, not a measured effect visvanathan-2005-laryngospasm.

One figure that circulates from this series, that intubation was required in 43% of cases, does not appear in the paper’s abstract or results and is not used here visvanathan-2005-laryngospasm. The bradycardia figure sometimes quoted at 11% is also wrong for the whole series; the paper gives 6%, with 23% in infants visvanathan-2005-laryngospasm.

The management ladder itself has no guideline source, and that is now a finding rather than a gap. On 2026-09-06 the two society documents that would be expected to carry it were searched. The Difficult Airway Society extubation guideline was read in full and is now in the ledger. It treats laryngospasm only as a risk, describing it as a protective exaggeration of the normal glottic closure reflex produced by stimulation of the superior laryngeal nerve, and warning that if it is not relieved promptly it may result in post-obstructive pulmonary edema and hypoxic cardiac arrest das-extubation-2012. It states no sequence. It contains no jaw thrust, no laryngospasm notch and no succinylcholine for post-extubation laryngospasm, and its single pharmacological statement is that doxapram has been used to prevent or treat laryngospasm although robust evidence to support its use for this indication is lacking das-extubation-2012. The Association of Paediatric Anaesthetists of Great Britain and Ireland has no laryngospasm guideline at all, and the joint paediatric difficult airway guidelines do not mention it. The ladder residents are taught is a textbook construction, not a society recommendation, and neither of those documents has a ledger row yet. The paper usually cited for the Larson maneuver remains a letter without an abstract that a later letter calls unproven. What the succinylcholine label does supply is the dose by route: an average intravenous dose of 0.6 mg/kg to produce blockade and facilitate intubation, an intramuscular dose of up to 3 to 4 mg/kg not exceeding 150 mg in total, with intramuscular onset usually observed in about 2 to 3 min; the label carries no laryngospasm-specific dose and carries a boxed warning for hyperkalemic cardiac arrest in children with undiagnosed myopathy fda-succinylcholine-label-2022. The five-case series describes supportive management of negative-pressure pulmonary edema only in outline bmc-nppe-2025.

Until this day was cleared it also carried the commonly taught sequence itself, with propofol and succinylcholine doses, marked as a general recommendation. Those doses had no source in this ledger, and a dose is the last thing that should be asserted without one, so they were removed rather than relabeled. The succinylcholine doses the FDA label does give are quoted above. Follow your institution’s laryngospasm algorithm.

At the bedside

Treat a falling saturation in a recovering patient as an airway problem until you have looked at the airway. Both AIMS analyses say the same thing from different directions: obstruction can present as nothing but desaturation visvanathan-2005-laryngospasm visvanathan-2005-airway.

Distinguish the larynx from the tongue. In the recovery period the tongue is the likelier culprit, and the negative-pressure edema that follows either looks the same bmc-nppe-2025. A jaw thrust and an oral or nasal airway address one; they do not address the other.

Use a structured approach and know it before you need it, because the registry authors judged that recognition or management could have been better in about one case in six visvanathan-2005-laryngospasm. Local practice differs. the institutional laryngospasm algorithm and the dose and route of succinylcholine it specifies differ between hospitals; know yours.

After any significant obstruction, watch for the edema, the aspiration and the hypoxemia that the registry names as its downstream harm, and do not discharge on the strength of the obstruction having resolved visvanathan-2005-laryngospasm.

Check yourself

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