Otolaryngol Clin N Am

40 (2007) 745–759

Interpretation of the Polysomnogram

in Children
Mary H. Wagner, MD*, Daniel M. Torrez, MD
Department of Pediatrics, Division of Pediatric Pulmonary, University of Florida,
1600 SW Archer Road, Gainesville, FL 32610, USA

Polysomnography (PSG) is important in the evaluation of nocturnal

events in children as well as adults. Events that can be evaluated include
obstructive sleep apnea syndrome (OSAS), periodic leg movements
(PLM), nocturnal seizures, parasomnias, and issues related to nocturnal
gas exchange. This article discusses the use of PSG primarily in terms of
respiratory events, leg movements, and gas exchange problems in children.
PSG has been recommended to evaluate several conditions in children,
including [1]
 Differentiation of benign from pathologic snoring
 Disrupted sleep
 Excessive daytime sleepiness
 Unexplained failure to thrive
 Cor pulmonale
 Polycythemia
 Laryngomalacia in children when worsened with sleep
 Underlying disorders predisposing children to nocturnal hypoxemia or
hypoventilation, such as bronchopulmonary dysplasia, cystic fibrosis,
neuromuscular disorders (muscular dystrophy, spinal muscular atrophy,
cerebral palsy, or congenital muscle diseases)
 Suspected alveolar hypoventilation
 Confirmation of clinical diagnosis of airway obstruction suggested by
symptoms including apnea, paradoxical respirations, or increased
work of breathing
 Documentation of severity of obstructed breathing to guide therapeutic
intervention and identification of those at increased risk of postopera-
tive complications

* Corresponding author.
E-mail address: wagneam@peds.ufl.edu (M.H. Wagner).

0030-6665/07/$ - see front matter Ó 2007 Published by Elsevier Inc.

doi:10.1016/j.otc.2007.04.004 oto.theclinics.com
746 WAGNER & TORREZ

 Titration of positive pressure for medical treatment of OSAS

 Follow-up evaluation of children who have persistent symptoms postin-
tervention for OSAS
Sleep disordered breathing (SDB) is a common cause of morbidity in
childhood, with a spectrum ranging from benign snoring to complete airway
obstruction. Benign or primary snoring is reported in 3% to 12% of the
pediatric population, with OSAS affecting 1% to 3% [2]. Several authors
have demonstrated that clinical history and physical examination are not
accurate in the identification of children who have OSAS [3]. PSG is useful
in documenting the presence of obstructive sleep apnea (OSA) events as well
as their severity. PSG has been recommended as the test of choice to
evaluate SDB by consensus of a panel of experts [4]. PSG has also shown
to be useful in determining readiness for decannulation in children who
have tracheostomy [5].

Performance of polysomnography in children

PSG should be performed by a laboratory experienced in and comfort-
able with caring for children [1]. Technicians need to be experienced in deal-
ing with children of various age levels and developmental status. Personnel
should be certified in pediatric cardiopulmonary resuscitation. If a dedicated
pediatric facility is not available, an area in the adult sleep laboratory should
be designated for children. Children should be housed in an appropriate
environment, with accommodations for a caregiver to sleep near the child.
Caregiver availability to the child is important to minimize the child’s
anxiety or fears about the study, as well as to provide any necessary care.
The procedure should be explained to the child and the family by personnel
skilled in the presentation of medical information. A crib should be
available for small children. The person responsible for supervision of the
pediatric sleep facility or functions should be a pediatrician with training
and expertise in the area of sleep medicine. This person must assure that
the PSG performance, scoring, and interpretation are appropriate for the
age and condition of the child.
The study timing should be set to mimic the child’s bedtime as closely as
possible. Overnight studies are preferred because negative nap studies have
been shown not to exclude the possibility of OSAS during a night study [6].
Studies should be performed without sedation in order to most accurately
mimic the child’s normal sleep.

Components of polysomnography in children

Variables gathered during PSG in children are similar to those obtained
in adults, with some additional information. Children may often demon-
strate obstructive hypoventilation as a component of their OSAS, which
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 747

may only be detected by increased carbon dioxide (CO2) levels [7]. Thus,
PSG monitoring in children should include some method of determining
CO2 levels, such as end tidal CO2 or transcutaneous CO2 [8]. The many
channels of physiologic parameters and their purpose in the PSG are
described in the following sections [1].
Sleep stages are determined using electroencephalogram (EEG), chin
electromyogram (EMG), and electro-oculogram (EOG). EEG monitoring
includes two central and two occipital leads with references to the opposite
posterior auricular area. Additional EEG leads can be used to detect noctur-
nal events such as seizure activity. Three chin EMG leads are placed and
used to detect skeletal muscle activity as required for the identification of
rapid eye movement (REM) sleep. The chin EMG is also useful to detect
swallowing and sucking during the study. Right and left EOG leads are
used to detect eye movements essential to the identification of REM sleep.
Respiratory effort is detected using chest and abdominal belts. Different
styles of monitoring include strain gauges, chest wall impedance, inductance
plethysmography, intercostal EMG, and pneumatic transducers [9]. These
belts can assess qualitative respiratory effort, which is essential to distin-
guishing whether respiratory events are central or obstructive in origin.
Air entry is assessed using a thermistor, nasal pressure, and a capnograph
tracing. The thermistor detects airflow at the nose and mouth by detecting
a temperature change in expired gas. Nasal pressure is an additional method
of monitoring airflow by detecting pressure changes via a cannula placed in
the nose. A recent study showed that the nasal pressure transducer is more
sensitive in the detection of hypopnea, and suggested combining the use of
both the thermistor and nasal pressure transducer for optimal detection of
SDB in children [10]. During positive pressure titrations, flow is detected
using measurements from the positive pressure device.
Movements of extremities are monitored by EMG leads placed on the
legs and sometimes on the arms. These movements are important in docu-
menting PLM as well as movements that result from respiratory events.
Position sensors can be used to document patient position. Digital video
is also obtained. The video is useful in documenting patient position, move-
ments, and unusual episodes such as parasomnias. Snoring is assessed by
a snore microphone placed on the neck of the patient, by technician
observation, and by audio recording.
Gas exchange is assessed using monitoring for both oxygen and CO2.
Oxygenation is monitored by pulse oximetry, using a comfortable sensor
that can be left on the child for the duration of the study. It is essential
that the pulse oximeter have a short sampling time (2 to 3 seconds) to avoid
missing brief desaturations associated with events in children. The reliability
of the pulse oximeter tracing is improved with recording of the pulse ampli-
tude signal, allowing identification of desaturation events that are caused by
poor probe function [11]. Ventilation is monitored in children by either end
tidal (ET) CO2 or transcutaneous CO2 monitoring. The ET CO2 method is
748 WAGNER & TORREZ

more responsive to rapid changes, whereas the transcutaneous method is not

reflective of transient CO2 changes, and is most useful for a trend [8,12]. ET
CO2 is monitored by a probe placed at the nose and mouth. ET CO2 gener-
ates a flow tracing that can also be used to monitor airflow. The sensor for
the transcutaneous device must be changed every several hours to maintain
accuracy. In patients with persistent gas exchange abnormalities, access to
arterial blood gas measurement is useful in corroborating noninvasive mon-
itoring of gas exchange. This can be useful in a patient who has alveolar
hypoventilation, to accurately assess the extent of CO2 retention, or in
a patient who has sickle cell disease with abnormal hemoglobin, in whom
the pulse oximetry may not accurately reflect arterial oxygen level [13].

Begin the review

It is helpful to preview the physician note/orders before the study to
determine why the PSG is being performed. This will help ensure that the
question being posed by the ordering physician will be appropriately
addressed. For example, if a child is being studied to determine whether
he/she can tolerate having his/her tracheostomy tube capped, it is important
to make certain the child has the capping device, and that the technicians
know that the tracheostomy is to be capped during the study. Before begin-
ning review of the study, it is helpful to review any notes made by the tech-
nician during the course of the night. This alerts the physician to technical or
patient issues encountered over the course of the study. These issues might
include unusual events over the night, such as confusional arousals or arti-
factual desaturation related to patient compromise of the oximeter probe. In
general, polysomnograms (PSGs) should be reviewed and scored by an
experienced scoring technician before interpretation; however, all PSGs
should be examined page by page by the reviewing professional for the
most accurate interpretation of the nocturnal events.
First, the biocalibration should be evaluated. This is a series of tests con-
ducted by the technician at the beginning and end of the study to document
the normal function of the various channels of information recorded. The
biocalibration may be limited in children who are young, developmentally de-
layed, or uncooperative. The components of the biocalibration include hav-
ing the patient look up, down, and to both sides in order to assess detection of
eye movements by EOG, which facilitates scoring of REM sleep. The EEG is
evaluated with eyes open and closed in order to identify alpha EEG waves,
which aid in detection of wakefulness. The patient is asked to make a snoring
type noise to check the snore channel, and the patient is asked to grit his/her
teeth to detect bruxism. The patient moves both legs separately to assess the
integrity of the leg EMGs. The patient is asked to hold his/her breath to
detect cessation of chest wall movement, and the chest and abdomen belts
are tested to make certain they move together with respiratory effort.
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 749

Sleep stage analysis

It is helpful to quickly review the patient’s sleep architecture by viewing
the hypnogram (Fig. 1). A hypnogram is a summary of the different sleep
stages achieved shown in graphic form. It is important to review the sleep
architecture in terms of what is to be expected for the patient’s age. Timing
and length of various sleep stages vary with patient age, and should be com-
pared with age-expected norms. For example, although it is normal for
infants to enter sleep through stage REM, entering sleep through stage
REM may suggest an underlying sleep disorder such as narcolepsy in an
older child or adolescent.
Components of sleep architecture that should be assessed include:
percentage of total sleep time (TST) spent in stage I/II, stage III/IV, stage
REM, and wakefulness. These percentages should be compared with age-
appropriate normals. Several authors have examined sleep architecture in
normal children ages 1 to 15 years [14–18]. In these studies, stage I sleep oc-
cupied 4% to 7.7% of TST, and stage II occupied 36% to 49% of TST, with
the combination of stage I and II in each study ranging from 41% to 53% of
the TST. Slow wave sleep (combining stages III and IV) occupied 14% to
32% of the TST, whereas stage REM occupied 17.4% to 21.1% of the TST.

Fig. 1. This is a typical hypnogram showing time across the bottom axis. The left axis shows the
different sleep stages. The horizontal bars indicate the time spent in the various sleep stages
during the course of the PSG.
750 WAGNER & TORREZ

Timing of sleep stages can be noted by review of the hypnogram. Chil-

dren usually have a short period of stage I/II after sleep onset, and then
enter stage III/IV (also known as delta or slow wave sleep). Stage III/IV
sleep will predominate early in the night, with regular cycling between the
stages I/II, stages III/IV, and REM. REM sleep will usually cycle every
60 to 120 minutes, with a wide range of timing between REM periods [14].
Sleep latency, the time after lights out until sleep is achieved, should also
be noted. Sleep latency is generally less than 25 minutes [14,16]. It may be
prolonged if the child has recently had a nap, and it may be shortened in
certain sleep disorders. Sleep efficiency is a measurement of the amount of
the total time in bed that the patient spends asleep, and should also be
noted. It gives a picture of whether the patient has a disrupted sleep pattern.
Sleep efficiency in children is usually greater than 89% [14–18].
REM latency, the time from onset of sleep to the first epoch of REM
sleep, is also noted. REM latency can be prolonged if the first REM period
is difficult to detect by the scoring technician. REM latency can be shortened
in certain conditions, such as depression or narcolepsy. The presence and
amount of REM sleep deserves careful attention. Length of time spent in
REM is short earlier in the night, with lengthening of REM episodes as
the night progresses [14]. During REM sleep, OSA may be worsened
because of loss of muscle tone. Thus, to make certain the most severe extent
of OSA is observed, patients should achieve REM sleep. REM sleep may be
decreased if the patient has a disrupted sleep pattern, with arousals out of
REM caused by obstructive events. This may cause the overall number of
obstructive events to be lowered.
EEG should be monitored for unusual complexes, such as nocturnal
epileptiform discharges (Fig. 2). These may be noted in children who have
a history of a seizure disorder as well as in those who do not have such a his-
tory. Whether these episodes are associated with clinical seizure activity or
respiratory events should be noted. Technician observation and the video
should be reviewed for evidence of clinical seizure activity in association
with these EEG changes. Consideration may be given to a more thorough
evaluation by a full sleep-deprived EEG if complexes are widespread during
the recording.

Arousal summary
Arousals are scored by the scoring technician based on the appearance of
the EEG tracing. An arousal is scored when there is an abrupt change in the
EEG lasting 3 seconds, following at least 10 seconds of continuous sleep
[19]. Arousals can be attributed to preceding events, including respiratory
events, leg movements, snore events, or technician presence in the room,
or may occur without an obvious trigger. Arousals are reported using the
arousal index, which is the number of arousals divided by the hours of sleep.
Normal values for arousal indices vary from laboratory to laboratory.
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 751

Fig. 2. Shown here is an episode of spike and wave activity occurring out of stage II sleep. The
tracing shows the EEG leads on the left axis with time on the bottom axis. A 10-second duration
screen is shown, with each rectangle along the bottom axis representing 1 second.

Studies of normal children have found mean arousal indices of 8.8 to 9.5
[14,15,17]. Arousals attributed to respiratory events give a measure of sleep
disruption caused by those respiratory events. It is also important to note
that children may not arouse to respiratory events as easily as adults do [7].

Heart rate/rhythm
The ECG should be reviewed for evidence of brady or tachy rhythms as
well as abnormal ECG rhythms. Respiratory events may be associated with
decrease in heart rate, with subsequent increase in heart rate after the event
has resolved, or in association with arousals. Some patients will have
evidence of premature ventricular or atrial contractions, which may or
may not be related to respiratory events.

Snoring
Determination of the presence or absence of snoring is important to note,
particularly in comparison to what is observed in the home environment. If
a child has loud snoring over the course of the night, typical to what parents
describe at home, with few documented respiratory events, a diagnosis of
752 WAGNER & TORREZ

primary snoring may be made. This may be reassuring to family members,

in that the snoring they have noted at home is not associated with docu-
mented respiratory events or gas exchange abnormalities; however, there
are reports in the literature that snoring without respiratory events has
been associated with poor academic performance [20,21].

Leg movements
The scoring technician will score leg movements that meet criteria for
PLM. Criteria for PLM include leg movements noted in either or both
legs that are at least one quarter of the amplitude noted during the biocali-
bration lasting from 0.5 to 5 seconds. Leg movements must be separated by
at least 5 seconds, but not more than 90 seconds, and must occur in clusters
of at least four to be considered PLM (Fig. 3). These leg movements should
not be related to other events, such as respiratory events or arousals [22].
Leg movements should be carefully reviewed to make certain they meet
criteria for PLM and are not related to subtle respiratory events. The total
number of PLM that meet these criteria are determined, and the PLM index
is calculated by dividing the total number of PLM by the number of hours
of sleep. A PLM index of five or greater is considered abnormal [22,23].

Fig. 3. This tracing shows a series of four periodic leg movements occurring over a time span of
2.5 minutes. The PLM are apparent as deflections in the two leg channels. LAT refers to left
anterior tibialis and RAT refers to right anterior tibialis. These leg movements are not associ-
ated with respiratory events or arousals.
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 753

Respiratory events
Several definitions are important for the classification of respiratory
events. The term ‘‘apnea’’ refers to an event with no flow. Apneas are scored
for events with a decrease in flow by 75% or greater from the baseline flow
observed before the event. If any portion of an event fits this definition, the
event can be scored as an apnea. Hypopnea refers to a decrease in flow by
30% to 50% from the usual baseline. Apneas and hypopneas can be further
classified as being central, obstructive, or mixed in nature.
Obstructive events are those caused by a decrease in flow associated with
persistent, sometimes increased respiratory effort noted in the chest and
abdominal belts. Obstructive apneas are scored when a decrease of 75%
or greater is detected in the airflow or nasal pressure channels associated
with continued respiratory effort (Fig. 4) [24]. Whereas in adults there is
a 10-second minimum event length noted, in children the events must be
two respiratory cycles [25]. Thus, in an infant breathing at 60 times per min-
ute, a significant event can be as brief as 2 seconds. An obstructive hypopnea
is scored when a decrease in the flow channel of 30% to 50% is noted with
persistent respiratory effort (Fig. 5) [1,25]. In order for an obstructive hypo-
pnea to be scored, there must be some consequence to the event, including
a 3% to 4% drop in oxygen saturation, a leg movement, or an EEG arousal.

Fig. 4. This figure demonstrates two episodes of obstructive apnea shown by cessation of flow
in the nasal pressure and airflow channels with continued respiratory effort. The events result in
oxygen desaturation as low as 80%.
754 WAGNER & TORREZ

Fig. 5. An episode of obstructive hypopnea is shown here, depicting two epochs or 1 minute of
a PSG. The event is marked by a decrease in the nasal pressure and airflow with continued re-
spiratory effort. The hypopnea results in a drop in oxygen saturation to 94% and an EEG
arousal.

The contour of the nasal pressure tracing is often flattened, indicating flow
limitation with respiratory effort during obstructive hypopneas [26].
Central events are those caused by an absence or severe decrease in respi-
ratory effort, as measured by chest or abdominal belts. Central apneas
(Fig. 6) are scored when there is an absence of respiratory effort associated
with a 70% or greater decrease in the airflow or nasal pressure. During
central apneas, there may be small deflections noted in the chest tracing cor-
responding to cardioblastic artifact from chest wall movement associated
with cardiac activity. Central hypopneas (Fig. 7) are events with decreased
respiratory effort associated with a 30% to 50% decrease in the airflow or
nasal pressure channels and desaturation or EEG arousal. The contour of
the nasal pressure tracing will be rounded during central hypopneas, com-
pared with the flattened contour noted with obstructive hypopneas [26]. It
is important to be certain that central hypopneas are not confused with
obstructive hypopneas. Obstructive hypopneas may appear to be associated
with an apparent decrease in respiratory effort caused by work against an
obstructed upper airway [26]. Central events are more likely to be noted
during REM sleep, particularly in patients who have underlying disorders
such as Prader-Willi syndrome or Arnold-Chiari malformation. Mixed
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 755

Fig. 6. This tracing demonstrates a central apnea lasting 11.5 seconds, resulting in a drop in
oxygen saturation to 91% and an EEG arousal. During this event there is cessation of airflow
with lack of respiratory effort.

respiratory events are those with a central and obstructive component.

Usually, these events begin with the central component, then progress to
an obstructed component with the resumption of respiratory effort.
After all events are reviewed and scored, several indices can be calculated.
These include apnea index (AI) and apnea-hypopnea index (AHI) which are
calculated for the entire sleep period, for non-REM (NREM) and REM
sleep. The term ‘‘index’’ refers to the number of events divided by the num-
ber of hours of sleep. This calculation allows the comparison of PSGs of
various lengths. The AI is determined using only apneas; the AHI includes
apneas and hypopneas.
Studies of normal children suggest that any obstruction is abnormal, and
that normal values for children are different from those for adults. Several
authors have investigated PSG findings in normal children [14,15,17]. There
is general agreement that AI and AHI are less than 1 in normal children.
Katz and Marcus [2] have suggested the following values for classification
of respiratory events in children:
 AHI less than 1 ¼ normal
 AHI 1 through 4 ¼ mild OSA
 AHI 5 through 10 ¼ moderate OSA
 AHI greater than 10 ¼ severe OSA
756 WAGNER & TORREZ

Fig. 7. This tracing demonstrates a 2-minute window in REM sleep with two episodes of cen-
tral hypopnea. These events are marked by a decrease of at least 30% in the airflow channel,
with a concomitant decrease in respiratory effort in the chest and abdomen channels. These
events result in significant oxygen desaturation as low as 78% and increase in ETCO2 to 53 torr.

They also propose grading of disease severity by gas exchange parame-

ters. The AHI should be considered in the context of the patient’s presenta-
tion. Children who have complications of OSAS or underlying medical
problems may require intervention for mild OSA in order to prevent wors-
ening of their medical disorder or worsened complications. For example,
children who have pulmonary hypertension might require aggressive inter-
vention, despite mild OSA, to prevent worsening of their pulmonary
hypertension.

Gas exchange
Gas exchange should be reviewed carefully for the entire tracing. The
pulse oximetry tracing should be reviewed for desaturation, with careful
attention to whether the desaturation is associated with a respiratory event,
arousal, or leg movement. Desaturations might be caused by respiratory
events of any duration. The actual desaturation will occur 1 to several
seconds after the respiratory event. If the respiratory event is prolonged,
the desaturation may begin before the event has terminated. Desaturations
unassociated with respiratory events should be reviewed for accuracy in
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 757

terms of whether the pulse oximeter is tracing accurately, or whether the

pulse oximetry probe has been compromised by patient position. It is impor-
tant to note the baseline oxygen saturation before study onset. Patients who
have low saturation at baseline may be at risk for desaturation with sleep
onset, worsening with respiratory events. Those who have persistent oxygen
desaturation without respiratory events potentially require further investiga-
tion for underlying lung disease, hypoventilation, or abnormal hemoglobin
(ie, sickle cell disease). Hypoventilation can be evaluated by noting the CO2
associated with episodes of oxygen desaturation. In general, oxygen satura-
tion should be greater than 92% in normal studies [14,15,17,18,27]. Children
on oxygen supplementation should have their flow rate recorded at the
beginning of the study, with monitoring of response to changes in the
oxygen flow rate over the course of the study.
CO2 tracing should be reviewed as well. Baseline CO2 level before sleep
onset should be noted. Normal children can be expected to have a 5 to 7
torr rise in CO2 with sleep onset [27]. Children may have a pattern of ob-
structive hypoventilation with OSAS, resulting in increases in CO2 without
significant oxygen desaturation. Abnormal levels of CO2 vary, with Marcus
and colleagues [27] reporting greater than 10% of TST being spent with CO2
greater than 50 torr as abnormal, and Montgomery-Downs and colleagues
[14] reporting that in normal children, 2.8  11.3 of the TST was spent with
a CO2 greater than or equal to 50 torr. Marcus and colleagues [27] also
report CO2 levels of greater than 53 torr as being abnormal. Uliel and col-
leagues [18] found less time spent with CO2 greater than 45 torr in their
study of normal children, suggesting that ET CO2 greater than 45 torr for
greater than 10% of the TST, or any CO2 greater than 50 torr, is abnormal.
It is important to recognize potential issues associated with ETCO2 moni-
toring. Sampling tubing can become obstructed with nasal secretions or
moisture, and may become displaced. Other factors that affect the accuracy
of the ETCO2 readings include mouth breathing, airway obstruction,
supplemental oxygen delivery at the other nostril, and cyanotic heart disease
[28–30]. It is important to know the values obtained in your individual sleep
laboratory, and compare values during the study with the baseline values.
CO2 trends may be more helpful, with consideration given to obtaining
a blood gas for high values to determine the accuracy of noninvasive
measurements.

Summary
PSG in children represents an important and useful tool in the evaluation
of the multitude of sleep-related conditions, including OSAS, PLM disorder,
and those with an underlying predisposition toward gas exchange aber-
rancies. To obtain the most useful information, sleep studies in this unique
population should be performed in laboratories with staff experienced in
758 WAGNER & TORREZ

working with children of all ages and stages of development, in a setting sen-
sitive to both child and caregiver.
In performing pediatric PSG, physicians should keep in mind
PSG can be successfully performed in children.
Pediatric PSG should be performed in a laboratory experienced in and
comfortable with the care of children.
Recognize that PSG in children includes measurement of carbon dioxide
levels to detect obstructive or central hypoventilation.
Criteria for scoring and interpretation of PSG in children differ from
those for adults.

References
[1] Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit
Care Med 1996;153:866–78.
[2] Katz ES, Marcus CL. Diagnosis of obstructive sleep apnea syndrome in infants and children.
In: Sheldon SH, Ferber R, Kryger MH, editors. Principles and practice of pediatric sleep
medicine. USA: Elsevier Saunders; 2005. p. 197–210.
[3] Brietzke SE, Katz ES, Roberson DW. Can history and physical examination reliably diag-
nose pediatric obstructive sleep apnea/hypopnea syndrome? A systematic review of the liter-
ature. Otolaryngol Head Neck Surg 2004;131(6):827–32.
[4] American Academy of pediatrics, Section on Pediatric Pulmonology, Subcommittee on Ob-
structive Sleep Apnea. Clinical practice guideline: diagnosis and management of childhood
obstructive sleep apnea syndrome. Pediatrics 2002;109:704–12.
[5] Tunkel DE, McColley SA, Baroody FM, et al. Polysomnography in the evaluation of read-
iness for decannulation in children. Arch Otolaryngol Head Neck Surg 1996;122(7):721–4.
[6] Marcus CL, Keens TG, Ward SL. Comparison of nap and overnight polysomnography in
children. Pediatr Pulmonol 1992;13:16–21.
[7] Rosen CL, D’Andrea L, Haddad GG. Adult criteria for obstructive sleep apnea do not iden-
tify children with serious obstruction. Am Rev Respir Dis 1992;146:1231–4.
[8] Morielli A, Desjardins D, Brouillette RT. Transcutaneous and end-tidal carbon dioxide
pressures should be measured during pediatric polysomnography. Am Rev Respir Dis 1993;
148:1599–604.
[9] Krishna J, Sans-Capdevila O, Gozal D. Sleep studies: which technologies? Paediatr Respir
Rev 2006;7(Suppl 1):S202–5.
[10] Budhiraja R, Goodwin JL, Parthasarathy S, et al. Comparison of nasal pressure transducer
and thermistor for detection of respiratory events during polysomnography in children.
Sleep 2005;28(9):1117–21.
[11] Lafontaine V, Ducharme FM, Brouillette RT. Can we rely on pulse oximetry desaturation
events? [abstract] Am Rev Respir Dis 1994;149:69A.
[12] Hansen TN, Tooley WH. Skin surface carbon dioxide tension in sick infants. Pediatrics 1979;
64:942–5.
[13] Craft JA, Alessandrini E, Kenney LB, et al. Comparison of oxygenation measurements in
pediatric patients during sickle cell crises. J Pediatr 1994;124:93–5.
[14] Montgomery-Downs HE, O’Brien LM, Gulliver TE, et al. Polysomnographic characteristics
in normal preschool and early school-aged children. Pediatrics 2006;117:741–53.
[15] Wong TK, Galster P, Lau TS, et al. Reliability of scoring arousals in normal children and
children with obstructive sleep apnea syndrome. Sleep 2004;27:1139–45.
 INTERPRETATION OF THE POLYSOMNOGRAM IN CHILDREN 759

[16] Coble PA, Kupfer DJ, Taska LS, et al. EEG sleep or normal health children. Part I: findings
using standard measurement methods. Sleep 1984;7:289–303.
[17] Traeger N, Schultz B, Pollock AN, et al. Polysomnographic values in children 2–9 years old:
additional data and review of the literature. Pediatr Pulmonol 2005;40:22–30.
[18] Uliel S, Tauman R, Greenfeld M, et al. Normal polysomnographic respiratory values in chil-
dren and adolescents. Chest 2004;125:872–8.
[19] American Sleep Disorders AssociationdThe Atlas Task Force. EEG arousals: scoring rules
and examples. Sleep 1992;15:174–84.
[20] Gozal D, Pope DW. Snoring during early childhood and academic performance at ages thir-
teen to fourteen years. Pediatrics 2001;107(6):1394–9.
[21] Urschitz MS, Guenther A, Eggebrecht E, et al. Snoring, intermittent hypoxia and academic
performance in primary school children. Am J Respir Crit Care Med 2003;168(4):464–8.
[22] American Academy of Sleep Medicine. International classification of sleep disorders. 2nd
edition. Diagnostic and coding manual. Westchester (IL): American Academy of Sleep Med-
icine; 2005. p. 182–6.
[23] Picchietti DL, Walters AS. Moderate to severe periodic limb movement disorder in child-
hood and adolescence. Sleep 1999;22(3):297–300.
[24] Phillips B, Kryger MH. Management of obstructive sleep apnea-hypopnea syndrome: over-
view. In: Kryger MH, Roth T, Dement WC, editors. Principles and practice of sleep medi-
cine. 4th edition. USA: Elsevier Saunders; 2005. p. 1111–3.
[25] Kheirandish-Gozal L. Practical aspects of scoring sleep in children. Paediatr Respir Rev
2006;7(S1):S50–4.
[26] Berry RB. Monitoring respiration during sleep. In: Sleep medicine pearls. 2nd edition. Phil-
adelphia: Hanley and Belfus, Inc.; 2003. p. 84–5.
[27] Marcus CL, Omlin KJ, Basinki DJ, et al. Normal polysomnographic values for children and
adolescents. Am Rev Respir Dis 1992;146:1235–9.
[28] Tobias JD, Meyer DJ. Noninvasive monitoring of carbon dioxide during respiratory failure
in toddlers and infants: end-tidal versus transcutaneous carbon dioxide. Anesth Analg 1997;
85:55–8.
[29] Friesen RH, Alswang M. End-tidal pCO2 monitoring via nasal cannulae in pediatric pa-
tients: accuracy and sources of error. J Clin Monit 1996;12:155–9.
[30] Fukuda K, Ichinohe T, Kaneko Y. Is measurement of end-tidal CO2 through a nasal cannula
reliable? Anesth Prog 1997;44:23–6.