British Journal of Anaesthesia 91 (1): 319 (2003)

DOI: 10.1093/bja/aeg126

The upper airway during anaesthesia

D. R. Hillman1 2*, P. R. Platt2 and P. R. Eastwood1
1
West Australian Sleep Disorders Research Institute, Department of Pulmonary Physiology and
2
Department of Anaesthesia, Sir Charles Gairdner Hospital, Hospital Avenue, Nedlands 6009,
Western Australia
*Corresponding author. E-mail: hillo@it.net.au

Upper airway obstruction is common during both anaesthesia and sleep. Obstruction is caused
by loss of muscle tone present in the awake state. The velopharynx, a particularly narrow
segment, is especially predisposed to obstruction in both states. Patients with a tendency to
upper airway obstruction during sleep are vulnerable during anaesthesia and sedation. Loss of
wakefulness is compounded by depression of airway muscle activity by the agents, and
depression of the ability to arouse, so they cannot respond adequately to asphyxia. Identifying
the patient at risk is vital. Previous anaesthetic history and investigations of the upper airway

Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016

are helpful, and a history of upper airway compromise during sleep (snoring, obstructive
apnoeas) should be sought. Beyond these, risk identication is essentially a search for factors
that narrow the airway. These include obesity, maxillary hypoplasia, mandibular retrusion,
bulbar muscle weakness and specic obstructive lesions such as nasal obstruction or
adenotonsillar hypertrophy. Such abnormalities not only increase vulnerability to upper airway
obstruction during sleep or anaesthesia, but also make intubation difcult. While problems
with airway maintenance may be obviated during anaesthesia by the use of aids such as the
laryngeal mask airway (LMA), identication of risk and caution are keys to management, and
the airway should be secured before anaesthesia where doubt exists. If tracheal intubation is
needed, spontaneous breathing until intubation is an important principle. Every anaesthetist
should have in mind a plan for failed intubation or, worse, failed ventilation.
Br J Anaesth 2003; 91: 319
Keywords: airway, complications; intubation tracheal, difcult; sleep; sleep apnoea

Collapse of the upper airway is common during anaesthesia,
and airway maintenance is a fundamental anaesthetic skill.
Anaesthetists recognise an airway as `difcult' if it is
difcult to maintain airway patency without tracheal
intubation or intubation itself is problematic. Such problems
often coexist because the tendency of the upper airway to
collapse during anaesthesia is related to anatomical factors,
which also make instrumentation difcult. Retrognathia for
example both connes the soft tissues and changes their
conguration.
The skeletal and neuromuscular characteristics that
predispose to these difculties also predispose to upper
airway obstruction during sleep. A tendency to obstruction
during anaesthesia and during sleep is related12 and the
anatomical features that predispose to difcult intubation
also predispose to upper airway obstruction during sleep.16
During wakefulness, airway patency is protected by
pharyngeal muscle tone. With both sleep and anaesthesia
there is a loss of muscle tone as a result of decreased cortical
inuences and chemoreceptor drive, together with modula-
tion of mechanoreceptor input. These changes predispose to
partial or complete upper airway obstruction, particularly in
those with already narrow, compliant upper airways.
In sleep, the changes are particularly evident in rapid eye
movement (REM) sleep, when the loss of muscle tone can
be profound. Where partial or complete obstruction occurs,
the event is terminated by arousal, which is usually brief
(<15 s), as a result of the accompanying return of muscle
tone. Once patency is restored, sleep resumes, and along
with it the tendency to obstruct again. These cyclical
obstructions terminated by arousals provide the patho-
physiological basis of obstructive sleep apnoea (OSA), a
common complaint with substantial associated morbidity.48
In the case of anaesthesia, the decrease in muscle tone
associated with loss of wakefulness is compounded by
specic drug-induced inhibition of upper airway neural and
LMA is the property of Intavent Limited.

The Board of Management and Trustees of the British Journal of Anaesthesia 2003
 Hillman et al.
Fig 1 Starling resistor model of the upper airway with a collapsible
(pharyngeal) segment between two rigid tubes (nasal and tracheal). When
ow limitation is present, maximal inspiratory ow is dened by
(PupstreamPcrit)/Rupstream, where Pcrit is the critical closing pressure
(=tissue pressure); Pupstream=upstream pressure; Pdownstream=downstream
pressure; Rupstream=upstream resistance, and Rdownstream=downstream
resistance. See text for further details.
muscle activity20 36 38 and suppression of protective arousal
responses. This increases obstruction, and external inter-
vention may be needed to overcome the obstruction.
The similar effects of sleep and anaesthesia on the upper
airway mean that a sleep history can give important
information to the anaesthetist about the unprotected
upper airway during anaesthesia and recovery.
Conversely, the anaesthetist may identify patients at risk
of upper airway obstruction during sleep if they have
structural (difcult intubation) or functional (upper airway
obstruction) problems during anaesthesia. The relationships
also contribute to understanding of the pathogenesis of
upper airway obstruction since information from studies of
one state can be applied to the other.
Sites and genesis of upper airway obstruction
during anaesthesia and sleep
The upper airway has a framework of bone and cartilage
with attached soft tissue structures, beginning at the nose
and lips and ending at the larynx.15 Where not supported by
bone or cartilage, the upper airway may collapse, because
muscle tone provides rigidity and this tone can change, as
with sleep or anaesthesia. It is these state-related changes
and the inuence of dynamic factors related to airow
through narrow, potentially collapsible, segments that limit
the ability to predict possible collapse during sleep or
anaesthesia from examination of the airway in the awake
state.
Sites of upper airway obstruction
Collapse will occur at points of narrowing and/or accidity.
Tonsillar hypertrophy is a common predisposing factor to
OSA in children, but adults rarely have specic pathology to
32
explain the disease. In general, adults with OSA have
narrower airways than normal subjects, with the velophar-
ynx often the smallest and most compliant segment in
both.22 25 This is the most common site of collapse during
sleep,21 32 although collapse may occur at several sites25
depending on specic predisposing factors.
It is not surprising that the velopharynx is also the most
common site of collapse during anaesthesia. This has been
shown radiologically during enurane anaesthesia,34 by
magnetic resonance imaging (MRI) during propofol anaes-
thesia29 and by pharyngeal manometry during isourane
anaesthesia.11 In the latter study, 14 of 16 subjects had this
pattern, with only two obstructing retrolingually. The
principal site of collapse was not affected by depth of
anaesthesia. These ndings contrast with those of an earlier
radiological study (and the traditional anaesthetic view)
which attributed upper airway obstruction during anaesthe-
sia to retrolingual collapse.31 This difference may be
because the earlier study was of paralysed patients.
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
During spontaneous ventilation, the potentially collapsible
pharynx is affected by dynamic effects of negative
intraluminal pressures during inspiration, which favour
collapse. Such conditions more closely approximate those
of sleep.
Prediction of the site of collapse during sleep from
measurements during wakefulness has been largely unsuc-
cessful,19 presumably because of marked differences in the
level of resting central drive from cortical inuences,
chemosensitivity and the gain of upper airway reexes. It is
likely that these differences are less between sleep and
anaesthesia, suggesting that behaviour under anaesthesia,
including site of collapse, may predict behaviour during
sleep.12 Hence patients vulnerable to OSA could be
identied during anaesthesia, and locating the site of
collapse might predict those most likely to benet from
surgical treatment such as palatal surgery.23
The genesis of upper airway collapse under
anaesthesia
Structurally, the upper airway consists of a collapsible
segment (the pharynx) situated between two rigid tubes
(nasal and tracheal). These correspond to the basic elements
of a Starling resistor (Fig. 1) which is a useful model of
upper airway pressureow relationships.14 Flow through
the collapsible segment depends on how the pressures
upstream and downstream of it relate to the pressure
surrounding it (tissue pressure). Three different states can
apply. (i) When both upstream and downstream pressures
exceed the tissue pressure, the segment will remain
distended and ow will vary with pressure gradient from
upstream to downstream. (ii) When the tissue pressure is
greater than the pressure downstream of the collapsible
segment but is exceeded by upstream pressure, ow
limitation will occur. Flow rate will be independent of
downstream pressure, and is determined by the gradient
 The upper airway during anaesthesia

Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016

Fig 2 Record from a patient during anaesthesia with isourane showing responses when mask pressure was rapidly reduced from a maintenance
pressure (14 cm H2O) for ve breaths and then returned to the maintenance pressure. This procedure was repeated with a range of positive and/or
negative airway pressures, causing varying degrees of inspiratory ow limitation (a plateau in inspiratory ow while oesophageal pressure fell
progressively, panels AC, see also Fig. 3), including a pressure sufcient to cause complete airway collapse (no change in inspiratory ow despite a
changing oesophageal pressure, panel D). Note that negative intrathoracic pressure swings are not transmitted to the nasopharynx, indicating collapse
distal to this point. Pm=mask pressure; Pnp=nasopharyngeal pressure; Pop=oropharyngeal pressure; Php=hypopharyngeal pressure; Poes=oesophageal
pressure. (From Eastwood and colleagues,11 with permission.)

between upstream and tissue pressures. There is a near-
constant pressure at the site of ow limitation that approxi-
mates tissue pressure, hence the resistance of the pathway
between the point of application of the upstream pressure
and the site of ow limitation is important. (iii) Where tissue
pressure exceeds upstream pressure then the segment is
occluded and no ow occurs. The upstream pressure at
which ow ceases is the critical closing pressure (Pcrit) of
the collapsible segment and provides a measure of its
collapsibility.
Studies of the human upper airway during sleep44 and
anaesthesia11 demonstrate the validity of this model. In 16
subjects breathing spontaneously at end-tidal isourane
levels of 0.4%, 0.8% and 1.2%, airway patency was
maintained with continuous positive airway pressure
applied via a nasal mask with the mouth occluded.11 As
this upstream pressure was systematically reduced, a value
was reached where inspiratory ow limitation rst
appeared. This was recognized by independence from
downstream pressure (Figs 2 and 3) and therefore
inspiratory effort. Further decreases in upstream pressure
caused proportionate reductions in maximum inspiratory
ow rates, until a critical level was reached (Pcrit), at which
complete collapse (no ow) was apparent. Pcrit varied
between individuals, reecting the variable collapsibility of
the upper airway: in some it was greater than atmospheric
pressure, indicating that such an airway would obstruct
without support; in others a negative pressure had to be
applied to produce obstruction, indicating relative resistance
to collapse. Indeed, Pcrit was greater than atmospheric
pressure in 8 of 16 subjects. These patients did not have any
known predisposition to upper airway obstruction, indicat-
ing that vulnerability to collapse of the passive upper airway
is remarkably common. Such vulnerability is likely to relate
to anatomical factors that reduce pharyngeal calibre by
increased accidity of the pharyngeal walls or increased
pharyngeal tissue pressures (see `Patient factors contribut-
ing to upper airway obstruction' below). Even when Pcrit
was less than atmospheric pressure, indicating a relatively
stable airway, ow limitation was evident at atmospheric
pressure (see Fig. 3, left panel).
The variability in upper airway collapsibility in anaes-
thetized patients is also found in sleep. In general, Pcrit
during sleep is greater in patients with sleep apnoea than in
those who simply snore, who, in turn, have greater values
than normal subjects.14 The tendency to upper airway

33
 Hillman et al.
Fig 3 Relationship between maximal inspiratory ow and nasal mask
(upstream) pressure for ow-limited breaths from two subjects breathing
spontaneously at end-tidal isourane levels of 1.2%. The left panel
demonstrates data from a subject with a relatively stable upper airway;
note that the critical closing pressure (Pcrit) is subatmospheric. The right
panel demonstrates data from a subject with an unstable airway; note that
Pcrit exceeds atmospheric pressure.
obstruction during sleep and during anaesthesia is related,
and sleep apnoea is more prevalent in those with greater
closing pressure under anaesthesia.12
Since vulnerability to collapse of the accid upper airway
is common, in many people airway patency during wake-
fulness must depend on activity of the upper airway dilator
muscles. These muscles (of which the genioglossus is the
most inuential) demonstrate tonic and phasic activity in the
awake state. The phasic activity is present during inspiration
and, together with background tonic activity, acts to resist
airway narrowing when pharyngeal intraluminal pressure
decreases during inspiration. This activity is greater during
wakefulness in patients with snoring and, particularly, OSA,
reecting the underlying vulnerability of their airways to
collapse.30
Upper airway muscle activity depends partly on centrally
mediated drive from the brainstem and partly on reexes
originating within the upper airway. For these reexes, the
most important local stimulus is negative intrapharyngeal
pressure.28 Topical anaesthesia of the pharynx increases
upper airway obstruction during sleep.6 With sleep and with
anaesthesia, upper airway muscle activity is reduced by
reduced central drive, because of decreased cortical inu-
ences and decreased chemosensitivity. A `doseresponse'
effect is evident during sleep, with some preservation of
upper airway muscle tone present in non-REM sleep. In
REM sleep, the loss of muscle tone can be profound, and
OSA is usually at its most severe during this sleep stage.18
In addition to its central effects, anaesthesia also directly
inhibits upper airway muscle and neural activity, including
inhibition of laryngeal respiratory-modulated mechanore-
ceptors, and therefore upper airway reexes,33 35 and
depresses protective arousal responses. In both sleep and
34
anaesthesia there appears to be greater depression of upper
airway muscles than of the diaphragm, so that breathing
efforts continue when upper airway muscle activity is
markedly reduced. This increases the vulnerability to upper
airway collapse during inspiration, as air is accelerated
across a disproportionately oppy upper airway. Because of
these sleep-associated effects, hypoglossal nerve stimula-
tion by an implanted nerve stimulator has been investigated
as a therapy for OSA.43
Effect of individual anaesthetic agents
Changes during anaesthesia appear to vary with the agent
used. Clearly, with neuromuscular blockade profound
relaxation of skeletal muscle activity is present. This effect
is dose dependent and a hierarchy of effect is observed, with
the diaphragm more resistant to neuromuscular blockade
than non-diaphragmatic respiratory muscles,8 including the
upper airway muscles.6 This indicates relative safety
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
margins for neuromuscular transmission between these
muscle groups, with a particularly high margin for the
diaphragm, consistent with its importance for survival.
During anaesthesia with spontaneous ventilation, this
same hierarchy is seen, with more inhibition of upper airway
relative to diaphragm activity, again reecting the func-
tional reserve of the diaphragm. This has been found in
experimental animals with subanaesthetic concentrations of
halothane, pentobarbital,20 thiopental, diazepam36 and
ethanol.4 Studies with anaesthetic concentrations of ha-
lothane found differential suppression of phrenic and
hypoglossal activity, with a dose-related decrease in both,
but to a greater extent for the hypoglossal nerve.36 When the
activity of respiratory muscles is considered, halothane has
a greater depressant effect on genioglossus activity than on
the intercostals, with least effect on the diaphragm.38 With
opioids these differential effects are less apparent, and
hypoglossal and phrenic nerve activity are reduced by
similar amounts.3 With ketamine the situation is less clear,
one study demonstrating differential effects,20 whereas
another did not.36
In humans, sedation with midazolam powerfully inhibits
upper airway activity.9 37 This is less apparent with
ketamine.9 Propofol sedation is also associated with
pharyngeal dysfunction46 and upper airway obstruction.29
However, doseresponse relationships have not yet been
clearly dened for any of these agents in man, so that it is
not known to what degree changes in anaesthetic depth
affect upper airway collapsibility. Such a relationship has
been seen with thiopental, with a progressive decrease in
upper airway electromyographic activity with increasing
dose.10
With isourane there is no clear doseresponse relation-
ship at anaesthetic concentrations, since genioglossus
activity is absent even during light planes of anaesthesia,
returning abruptly with emergence.11 However, disturbance
of swallowing occurs at subanaesthetic concentrations,46
 The upper airway during anaesthesia
Table 1 Known and suspected predisposing conditions for obstructive sleep
apnoea (adapted from Loadsman and Hillman,26 with permission)
Condition Examples
Obesity Acquired obesity, PraderWilli syndrome
Genetic predisposition
Age
Male gender
Alcohol, sedatives,
analgesics, anaesthetics
Smoking
Nasal obstruction Septal deviation, chronic nasal congestion
Pharyngeal obstruction Tonsillar and adenoidal hypertrophy
Craniofacial abnormality Down's, PierreRobin, TreacherCollin's,
Apert's, Crouzon's, and BeckwithWiedemann
syndromes, achondroplasia, acromegaly, fragile-
X syndrome
Laryngeal obstruction Laryngomalacia
Endocrine/metabolic Hypothyroidism, androgen therapy, Cushing's
disorders syndrome
Neuromuscular disorders Stroke, cerebral palsy, head injury, ShyDrager
syndrome, poliomyelitis, muscular dystrophies,
myotonic dystrophy, tetraplegia
Connective tissue Marfan's syndrome
disorders
Storage diseases Mucopolysaccharidoses
Chronic renal failure
suggesting subtle pharyngeal impairment, so that there
could be a denable doseresponse effect at these low drug
levels. A differential effect on upper airway muscles relative
to the diaphragm is clearly found at anaesthetic concentra-
tions, with inspiratory efforts continuing when upper airway
activity is profoundly depressed.11
The presence of measurable effects at relatively low
concentrations of many anaesthetic drugs highlights the
susceptibility of the patient with a vulnerable upper airway
to obstruction in the recovery room after anaesthesia is
complete, or elsewhere where narcotic analgesics or
sedation are administered.
Patient factors predisposing to upper airway
collapse during anaesthesia or sleep
Factors that narrow the pharynx, increase pressure around it,
reduce pressure within it or make its walls more compliant
will increase upper airway obstruction during sleep and
anaesthesia. These factors may exist in the lumen (e.g. a
foreign body), in the pharyngeal wall (e.g. bulbar muscle
weakness), or outside the pharynx. Factors acting outside
the pharynx include direct actions (e.g. lateral pharyngeal
fat pads in obesity, narrow skeletal connes) or indirect
actions (e.g. loss of longitudinal traction on the pharynx
from neck exion or obesity-related reduction in lung
volumes). A narrow airway is particularly vulnerable to
collapse for three reasons. Firstly, Laplace's law requires
that with a small radius of curvature, a greater expanding
force (transmural pressure) is needed to maintain sufcient
wall tension to oppose collapse. Secondly, the absolute
change in calibre before airway closure occurs is less if the
calibre is already small. Thirdly, the greater resistance of a
35
narrow airway requires generation of more negative
intraluminal pressures during inspiration, increasing the
tendency to collapse. These considerations are relevant to
conditions that predispose to upper airway obstruction when
upper airway dilator muscle activity is reduced with sleep or
when under the inuence of sedative or anaesthetic drugs
(Table 1).
Characteristics that reect propensity to upper airway
collapse include obesity, macroglossia, micrognathia and
maxillary hypoplasia.15 These may be present in otherwise
normal individuals or as part of a disease syndrome, such as
acromegaly, Down's syndrome, PierreRobin syndrome or
other craniofacial abnormalities. Familial predispositions
are common,41 reecting common morphological charac-
teristics.
Neuromuscular disorders that decrease upper airway
muscle tone predispose to upper airway obstruction, as do
endocrine, connective tissue and storage diseases affecting
upper airway calibre.
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
Specic lesions in the upper airway provide other
predispositions to upper airway obstruction that are site
specic. Examples include nasal obstruction, tonsillar and
adenoidal hypertrophy, pharyngeal and laryngeal tumours,
foreign bodies, haematomas, oedema and laryngomalacia.
Patient habits such as smoking, and alcohol and sedative
consumption also predispose to collapse through their
effects on muscle tone or airway calibre. The supine posture
is a further predisposing factor, as the effects of gravity
increase the extraluminal compressive forces exerted by the
tongue, soft palate and related structures.
Assessment of the upper airway
Clinical ndings
Identication of patients at risk of upper airway obstruction
during sleep or anaesthesia or of difcult intubation is a
search for the factors referred to in the previous section.
Many pathological processes are predictably associated
with difcult airway maintenance or intubation, such as
epiglottitis and tumours of the larynx or pharynx. Not all
difcult airways are associated with major pathology, but
result from more subtle anatomical predisposition, modu-
lated by neuromuscular forces. Such difculties may only
become manifest with the state-related decreases in
neuromuscular activation that accompany sleep and anaes-
thesia. Examination of the awake patient may not reveal
these inuences or the dynamic effect of airow across
potentially vulnerable segments, reducing the capacity to
identify airways at risk.
Features that anaesthetists associate with difcult intuba-
tion are also associated with upper airway obstruction
during sleep.16 However, although the upper airway may
readily obstruct during sleep or anaesthesia, and these
propensities are related,12 difcult intubation is a much rarer
phenomenon. Hence measurements that reasonably reliably
 Hillman et al.
predict vulnerability to collapse, such as obesity and
increased neck circumference,15 or others referred to
below are much less accurate predictors of difcult
intubation.47
In the absence of obvious pathology, the most reliable
predictor of difcult intubation is previously documented
difculty, and previous anaesthetic records should always
be sought for guidance. Problems with airway maintenance
or intubation should always be recorded in the notes and in
emergency medical record systems such as MedicAlert, and
the patient should be advised.
Where such information is not available then the
anaesthetist depends on clinical indicators to suggest the
possibility of difculty, but should recognise their limita-
tions. Particular note should be made of obesity, increased
neck circumference, decreased neck length, limited head/
neck extension, impaired nasal patency, crowded pharyn-
geal appearance (high Mallampati score, macroglossia,
tonsillar hypertrophy), dental abnormalities, limited mouth
opening, maxillary hypoplasia, mandibular hypoplasia/
retrusion, decreased thyromental distance and increased
mandibular angle.
As many of the airway abnormalities that predispose to
difculty during anaesthesia also cause problems during
sleep, a sleep history is useful, paying particular attention to
the cardinal symptoms of OSA: snoring, witnessed apnoeas
and disrupted sleep with excessive daytime somnolence.
Investigations
Where concern exists regarding upper airway calibre and
function, further information may come from radiographic,
endoscopic and physiological evaluation. Usually these
investigations have been performed awake, and so the
potential effects of state-related changes in neuromuscular
function accompanying anaesthesia need to be considered
when interpreting the ndings.
CT and MRI. Studies using computed tomography (CT) and
MRI scanning have improved our knowledge of upper
airway function under anaesthesia29 and the pathogenesis of
sleep apnoea.42 While the chest radiograph can indicate
tracheal deviation or compression, both CT and MRI
scanning give accurate images of the upper airway, which
allow its calibre to be examined from end to end. Although
useful generally, such scans are particularly helpful to
anaesthetists in examining the inuence of clinically
identied abnormalities on upper airway conguration.
For example, the effect of tumours that are displacing or
constricting the airway can be quantied. This information
is usually available in such cases as they present for surgery
and provides invaluable insights for the anaesthetist
preparing for the challenges in airway management that
they may present.
36
Lateral cephalometry. Lateral cephalometry involves a true
lateral radiograph of the head and neck, with penetration
carefully determined to outline soft tissue and skeletal
outlines. Allowance is made for magnication.
Measurements are taken of craniofacial and soft tissue
structures. Patients with OSA have a small retroposed
mandible, narrow posterior airway spaces, enlarged tongue
and soft palate, inferior positioned hyoid bones and
retroposition of their maxillae.42 Patients who are difcult
to intubate share many of these characteristics.16 While
readily available, lateral cephalometry requires expert
evaluation and is used more to examine the site and cause
of upper airway narrowing, for example when craniofacial
surgery is proposed, than for clinical anaesthesia.
Endoscopy. Fibreoptic upper airway endoscopy allows the
calibre and conguration of the airway to be seen directly
and possible endoscope-assisted intubation to be assessed.
Familiarity with airway endoscopy and breoptic-aided
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
intubation is essential for good management of difcult
airways. Laryngoscopic evaluation of the upper airway may
already have been done during the surgical assessment,
particularly of patients presenting for surgery to the upper
airway, and this information should be obtained.
Maximum inspiratory and expiratory ow volume curves. In
substantial upper airway obstruction, reduction in maximum
respiratory ow rates occurs, particularly during inspiration
as the negative intraluminal pressures favour collapse. A
plateau in ow is often seen, indicating ow limitation.1 The
reduction in maximum ow rates indicates the severity of
the obstruction and the measurement is readily repeated,
allowing changes with time and treatment to be assessed.
This is not used for routine evaluation but is valuable for
following the progress of patients with chronic conditions
associated with substantial airway narrowing, usually
clinically evident during wakefulness, that require intermit-
tent treatment (e.g. subglottic stenoses).
Recognition of upper airway obstruction
The hallmark of upper airway obstruction is diminished or
absent airow in the presence of continued respiratory
effort. Complete upper airway obstruction is silent, while
partial obstruction will be accompanied by snoring (if
supralaryngeal), or inspiratory stridor (if perilaryngeal). A
consequence of continued inspiratory efforts against the
partially or completely obstructed airway is the develop-
ment of large negative intrathoracic pressure swings, which
are often seen as a combination of diminished chest wall
movement and distortion, with paradoxical (inward) motion
of the anterior rib cage. In the laboratory (e.g. during
polysomnography in a sleep clinic), obstruction is quantied
by relating measures of breathing effort (chest wall motion
or, more accurately, oesophageal pressure changes) with
those of ventilation (oronasal airow).
 The upper airway during anaesthesia
Prevention and treatment of upper airway
obstruction
Many pathological processes are predictably associated
with difcult airway maintenance or intubation, such as
epiglottitis and laryngeal or pharyngeal tumours. These can
be extremely challenging for the anaesthetist. Their man-
agement is well described in standard anaesthetic textbooks
and is beyond the scope of this article, which is primarily
concerned with airway difculty arising from more subtle
variations in upper airway structure and function.
Anatomical positioning and posture
The supine posture can worsen upper airway obstruction
because gravity affects the tongue and soft palate position,
narrowing the retropalatal and retrolingual airways. Neck
exion, mouth opening and table tilt with the head down are
also unfavourable because of loss of longitudinal tension on
the upper airway.
Apart from avoiding these unfavourable circumstances
where possible and practical, two commonly practised
manoeuvres can be used to improve airway patency. Firstly,
placing the head in the `snifng the morning breeze'
position (lower cervical exion, upper cervical extension
with full extension of head on neck) increases longitudinal
tension on the upper airway and decreases its collapsibility.
Secondly, displacing the mandible forwards pulls the tongue
forward and increases the calibre of the retrolingual airway
and the retropalatal airway as the soft palate is also
displaced forwards because it is coupled to the tongue via
the fauces.24 The principle of mandibular advancement is
now also applied to the treatment of OSA by dental splints
designed for the purpose.39 Simply lifting the chin with
mouth closure can also be effective in increasing pharyngeal
dimensions by increasing longitudinal tension in the
pharyngeal muscles and anterior neck structures. This
action increases the anteroposterior distance between
tongue and posterior pharyngeal wall and counteracts the
tendency to collapse.40
Continuous positive airway pressure (CPAP)
Applying positive pressure to the upper airway in order to
splint it pneumatically is a widely used principle in the
treatment of OSA45 but is less systematically applied during
anaesthesia, even though effective in this circumstance.11
While this is understandable intraoperatively where other
more convenient techniques are available to the anaesthetist
for airway management, applying PEEP when assisting
ventilation with bag and mask is a strategy that can help
relieve upper airway obstruction. It is probable that CPAP is
underused in the recovery room, where the potential for
upper airway obstruction is high, and further into the
postoperative period, where many patients with the ten-
dency to upper airway obstruction remain at increased risk
37
because of the use of narcotic analgesic or sedative drugs. In
this regard it is important to note that OSA remains
notoriously under-diagnosed,48 and may rst be identied as
a problem under these circumstances.26
Airway aids
Oral and nasopharyngeal airways help by bypassing the
obstruction (e.g. the oral airway in the case of retropalatal
obstruction) or providing a conduit through it (e.g. the
nasopharyngeal airway in the same circumstance). Often
these prove to be an aid rather than a complete solution to
provision of an adequate airway, and neck extension and
forward traction on the mandible may also be needed to
maintain patency.
The laryngeal mask airway (LMA), of which there are
now several variants, can provide a conduit from the
exterior to the rima glottidis. Its ease of introduction and
reliable provision of a patent airway without the require-
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
ment of additional manoeuvres have made it an invaluable
addition to anaesthetic equipment in the 19 years since its
rst description.5
Tracheal tubes
Tracheal intubation remains the `gold standard' for control
of the airway. Once secured, the lower airway and gas-
exchanging regions of the lung are directly accessed and
control is complete. However, because of their common
anatomical factors, airways that are difcult to maintain
patent may also be difcult to intubate. Where the
intubation is known (from previous experience) or predicted
(from preoperative assessment see above) to be difcult,
then spontaneous ventilation should be preserved until
intubation is accomplished. This is best done during
wakefulness using a breoptic endoscope introduced
under local anaesthesia, providing the patient is cooperative
and trauma associated with bleeding does not obscure the
view. Alternatively, it may be done in the anaesthetized
spontaneously breathing patient either by direct laryngo-
scopy, often with the aid of a bougie, light wand or similar
device, or via an LMA designed to allow passage of a
breoptic scope (e.g. the LMA-Fastrach). Tracheotomy
under local anaesthesia is a more drastic solution that may
be necessary if all else fails, or in the case of an obstructing
lesion (such as carcinoma of the larynx) that does not allow
the passage of even a relatively small tracheal tube.
When un-anticipated difcult intubation is encountered
after neuromuscular blocking agents have been given, then
subsequent events depend on how well ventilation and gas
exchange can be maintained (ventilating with 100% oxy-
gen) while attempting to solve the difculty. Adequate
oxygenation is the rst consideration. If ventilation is well
maintained by bag and mask, it may be possible to secure
the airway using one of the methods specied above. The
intubating LMA (LMA-Fastrach) is often useful,13 and even
 Hillman et al.
more so if there is any difculty maintaining ventilation. If
problems with ventilation persist, this is an emergency and
extra help should be obtained. An emergency airway needs
to be established using transtracheal techniques, such as
transtracheal jet ventilation or tracheotomy. Several algo-
rithms that deal with such circumstances have been
published, one of the most authoritative being from the
American Society of Anesthesiologists.2
Heliox
Where the airway is extremely narrow, ow turbulence is a
prominent feature. This is a density-dependent phenomenon
and a gas of low density may relieve respiratory distress.
Heliumoxygen mixtures full this requirement.27 A bal-
ance has to be struck between decreasing density of the
carrier gas and reducing the inspired concentration of
oxygen, hence the respective concentrations of helium and
oxygen.17 This is determined from assessment of breathing
comfort and arterial blood gas analysis. The treatment is
expensive so this is a strategy for short-term use until other
therapies can relieve the underlying problem.
Summary
The difcult upper airway causes problems with airway
maintenance and/or instrumentation under anaesthesia. The
skeletal and neuromuscular characteristics that predispose
to these difculties also predispose to upper airway
obstruction during sleep.
The behaviour of the upper airway during sleep can
indicate to the anaesthetist its likely behaviour during
anaesthesia, where the effects of loss of wakefulness are
compounded by specic anaesthesia-induced inhibition of
upper airway neural and muscle activity and suppression of
protective arousal responses. Conversely, difcult airway
maintenance or intubation during anaesthesia suggest OSA,
a common condition with substantial associated morbidity.
Safe management of such cases involves identication of
risk and cautious preparation, securing the airway before
anaesthesia where doubt exists, and a systematic approach
in the case of failed intubation or, worse, failed ventilation.
Acknowledgements
This work was supported in part by the Australian Lung Foundation, the Sir
Charles Gairdner Hospital Research Fund, and grant no. 109903 from the
National Health and Medical Research Council of Australia.
References
1 Acres JC, Kryger MH. Clinical signicance of pulmonary function
tests: upper airway obstruction. Chest 1981; 80: 20711
2 American Society of Anesthesiologists Task Force Report.
Practice guidelines for management of the difcult airway.
Anesthesiology 1993; 78: 597602
3 Bartlett D, St John WM. Inuence of morphine on respiratory
38
activities of phrenic and hypoglossal nerves in cats. Respir Physiol
1986; 64: 28994
4 Bonora M, Shields GI, Knuth SL, Bartlett D, St John WM.
Selective depression by ethanol of upper airway respiratory
motor activity in cats. Am Rev Respir Dis 1984; 130: 15661
5 Brain AI. The laryngeal mask a new concept in airway
management. Br J Anaesth 1983; 55: 8015
6 Deegan PC, Mulloy E, McNicholas WT. Topical oropharyngeal
anesthesia in patients with obstructive sleep apnea. Am J Respir
Crit Care Med 1995; 151: 110812
7 D'Honneur G, Slavov V, Merle JC, et al. Comparison of the
effects of mivacurium on the diaphragm and geniohyoid muscles.
Br J Anaesth 1996; 77: 7169
8 DeTroyer A, Bastenier J, Delhez L. Function of respiratory
muscles during partial curarization in humans. J Appl Physiol 1980;
49: 104956
9 Drummond GB. Comparison of sedation with midazolam and
ketamine: effects on airway muscle activity. Br J Anaesth 1996; 76:
6637
10 Drummond GB. Inuence of thiopentone on upper airway
muscles. Br J Anaesth 1989; 63: 1221
11 Eastwood PR, Szollosi I, Platt PR, Hillman DR. Collapsibility of
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
the upper airway during anesthesia with isourane. Anesthesiology
2002; 97: 78693
12 Eastwood PR, Szollosi I, Platt PR, Hillman DR. Comparison of
upper airway collapse during general anaesthesia and sleep.
Lancet 2002; 359: 12079
13 Ferson DZ, Rosenblatt WH, Johansen MJ, Osborn I, Ovassapian
A. Use of the intubating LMA-Fastrach in 254 patients with
difcult-to-manage airways. Anesthesiology 2001; 95: 117581
14 Gold AR, Schwartz AR. The pharyngeal critical pressure. Chest
1996; 110: 107788
15 Goldberg AN, Schwab RJ. Identifying the patient with sleep
apnea: upper airway assessment and physical examination.
Otolaryngol Clin North Am 1998; 31: 91930
16 Hiremath AS, Hillman DR, James AL, Noffsinger WJ, Platt PR,
Singer SL. Relationship between difcult tracheal intubation and
obstructive sleep apnoea. Br J Anaesth 1998; 80: 60611
17 Ho AM, Dion PW, Karmakar MK, Chung DC, Tay BA. Use of
heliox in critical upper airway obstruction. Physical and
physiologic considerations in choosing the optimal
helium:oxygen mix. Resuscitation 2002; 52: 297300
18 Horner RL. Motor control of the pharyngeal musculature and
implications for the pathogenesis of obstructive sleep apnoea.
Sleep 1996; 19: 82753
19 Hudgel DW. Variable site of airway narrowing among
obstructive sleep apnea patients. J Appl Physiol 1986; 61: 14039
20 Hwang JC, St John WM, Bartlett D. Respiratory-related
hypoglossal nerve activity: inuence of anesthetics. J Appl
Physiol 1983; 55: 78592
21 Isono S, Morrison DL, Launois SH, Feroah TR, Whitelaw
WA, Remmers JE. Static mechanics of the velopharynx of
patients with obstructive sleep apnea. J Appl Physiol 1993; 75:
14854
22 Isono S, Remmers JE, Tanaka A, Sho Y, Sato J, Nishino T.
Anatomy of pharynx in patients with obstructive sleep apnea and
in normal subjects. J Appl Physiol 1997; 82: 131926
23 Isono S, Shimada A, Tanaka A, et al. Efcacy of endoscopic static
pressure/area assessment of the passive pharynx in predicting
uvulopalatopharyngoplasty outcomes. Laryngoscope 1999; 109:
76974
24 Isono S, Tanaka A, Sho Y, Konno A, Nishino T. Advancement of
the mandible improves velopharyngeal airway patency. J Appl
Physiol 1995; 79: 21328
 The upper airway during anaesthesia
25 Kuna ST, Bedi DG, Ryckman C. Effect of nasal airway positive
pressure on upper airway size and conguration. Am Rev Respir
Dis 1988; 138: 96975
26 Loadsman JA, Hillman DR. Anaesthesia and sleep apnoea. Br J
Anaesth 2001; 86: 25466
27 Lu TS, Ohmura A, Wong KC, Hodges MR. Heliumoxygen in
treatment of upper airway obstruction. Anesthesiology 1976; 45:
67880
28 Malhoutra A, Pillar G, Fogel RB, et al. Pharyngeal pressure and
ow effects on genioglossus activation in normal subjects. Am J
Respir Crit Care Med 2002; 165: 717
29 Mathru M, Esch O, Lang J, et al. Magnetic resonance imaging of
the upper airway. Effects of propofol anesthesia and nasal
continuous positive airway pressure in humans. Anesthesiology
1996; 84: 2739
30 Mezzanotte WS, Tangel DJ, White DP. Waking genioglossal
electromyogram in sleep apnea patients versus normal controls
(a neuromuscular compensatory mechanism). J Clin Invest 1992;
89: 15719
31 Morikawa S, Safar P, DeCarlo J. Inuence of the headjaw
position upon airway patency. Anesthesiology 1961; 22: 26570
32 Morrison DL, Launois SH, Isono S, Feroah TR, Whitelaw WA,
Remmers JE. Pharyngeal narrowing and closing pressures in
patients with obstructive sleep apnea. Am Rev Respir Dis 1993;
148: 60611
33 Mutoh T, Kanamaru A, Kojima K, Nishimura R, Sasaki N,
Tsubone H. Effects of volatile anesthetics on the activity of
laryngeal `drive' receptors in anesthetized dogs. J Vet Med Sci
1999; 61: 10338
34 Nandi PR, Charlesworth CH, Taylor SJ, Nunn JF, Dore CJ. Effect
of general anaesthesia on the pharynx. Br J Anaesth 1991; 66:
15762
35 Nishino T, Anderson JW, Sant'Ambrogio G. Effects of halothane,
enurane, and isourane on laryngeal receptors in dogs. Respir
Physiol 1993; 91: 24760
36 Nishino T, Shirahata M, Yonezawa T, Honda Y. Comparison of
changes in the hypoglossal and phrenic nerve activity in response
to increasing depth of anesthesia in cats. Anesthesiology 1984; 60:
1924
37 Nozaki-Taguchi N, Isono S, Nishino T, Numai T, Taguchi N.
39
Upper airway obstruction during midazolam sedation:
modication by nasal CPAP. Can J Anaesth 1995; 42: 68590
38 Ochiai R, Guthrie RD, Motoyama EK. Effects of varying
concentrations of halothane on the activity of the genioglossus,
intercostals and diaphragm in cats: an electromyographic study.
Anesthesiology 1989; 70: 8126
39 O'Sullivan RA, Hillman DR, Mateljan R, Pantin C, Finucane KE.
Mandibular advancement splint: an appliance to treat snoring and
obstructive sleep apnea. Am J Respir Crit Care Med 1995; 151:
1948
40 Reber A, Wetzel SG, Schnabel K, Bongartz G, Frei FJ. Effect of
combined mouth closure and chin lift on upper airway
dimensions during routine magnetic resonance imaging in
pediatric patients sedated with propofol. Anesthesiology 1999;
90: 161723
41 Redline S, Tosteson T, Tishler PV, Carskadon MA, Millman RP.
Studies in the genetics of obstructive sleep apnoea. Familial
aggregation of symptoms associated with sleep related breathing
disturbances. Am Rev Respir Dis 1992; 145: 4404
42 Schwab RJ, Goldberg AN. Upper airway assessment:
radiographic and other imaging techniques. Otolaryngol Clin
North Am 1998; 31: 93168
Downloaded from http://bja.oxfordjournals.org/ by guest on February 25, 2016
43 Schwartz AR, Bennett ML, Smith PL, et al. Therapeutic electrical
stimulation of the hypoglossal nerve in obstructive sleep apnea.
Arch Otolaryngol Head Neck Surg 2001; 127: 121623
44 Smith PL, Wise RA, Gold AR, Schwartz AR, Permutt S. Upper
airway pressure-ow relationships in obstructive sleep apnea.
J Appl Physiol 1988; 64: 78995
45 Sullivan CE, Issa FG, Berthon-Jones M, Eves L. Reversal of
obstructive sleep apnoea by continuous positive airway pressure
applied through the nares. Lancet 1981; 18: 8625
46 Sundman E, Witt H, Sandin R, et al. Pharyngeal function and
airway protection during subhypnotic concentrations of
propofol, isourane, and sevourane: volunteers examined by
pharyngeal videoradiography and simultaneous manometry.
Anesthesiology 2001; 95: 112532
47 Yentis SM. Predicting difcult intubation worthwhile exercise
or pointless ritual? Anaesthesia 2002; 57: 1059
48 Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive
sleep apnea: a population health perspective. Am J Respir Crit Care
Med 2002; 165: 121739