- Email: [email protected]
Age specific differences in pediatric obstructive sleep apnea
International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Age specific differences in pediatric obstructive sleep apnea§ Debra M. Don *, Kenneth A. Geller, Jeffrey A. Koempel, Sally Davidson Ward Division of Pediatric Otolaryngology and Pulmonology, Childrens Hospital Los Angeles, Keck School of Medicine, University of Southern California, Los Angeles, CA, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 4 September 2008 Received in revised form 1 April 2009 Accepted 5 April 2009 Available online 1 May 2009
Background: Some have suggested that younger children have a more severe form of obstructive sleep apnea than older children and therefore are at a higher risk for respiratory compromise after tonsillectomy and adenoidectomy. However, at present there are few studies that have identified any significant correlation between age and severity of obstructive sleep apnea. Objective: To determine if age specific differences in obstructive sleep apnea are present in children. Design: Retrospective chart review. Setting: Tertiary care children’s hospital. Patients: The records of children (1–18 years of age) with obstructive sleep apnea diagnosed by overnight polysomnography between January 1998 and January 2001 were reviewed. Children included in the study also had evidence of adenotonsillar hypertrophy and had no other co-existing medical problems. Main outcome measures: Overnight polysomnography was performed in all children. Apnea–hypopnea index (AHI), baseline and lowest O2 saturation, baseline and peak end tidal CO2, and total number of obstructive apneas, hypopneas, central apneas and mixed apneas were measured during each polysomnogram. Children were subdivided into the following age groups: 1–2, 3–5, 6–11 and 12–18 years. Polysomnograms were classified into normal, mild, moderate and severe categories. Results: Three hundred and sixty-three children were studied; 45 children were ages 1–2 years, 159 children were ages 3–5 years, 137 children were 6–11 years and 22 children were 12–18 years. Although there appears to be a trend towards a greater mean number of obstructive apneas, hypopneas, central apneas, mixed apneas, a higher mean AHI, lower mean SaO2 nadir, and a higher mean PETCO2 in the younger age groups when compared to the older groups, a Student’s t-test demonstrates that there is no statistical significance for most OSA parameters. An analysis of variance using the F-test reveals statistical significance (p < 0.01) when children ages 1–2 were compared to those 3–5, 6–11 or 12–18 years of age for the variables AHI, mean number of central apneas, hypopneas and mixed apneas. When comparing patients in the various severity categories, children ages 1–2 years show a distinct distribution with a larger percentage in the moderate to severe categories. Chi square analysis reveals a significant difference between the frequency distribution of children in age group 1–2 years and that of the other age groups (p < 0.01). Conclusion: There is a predilection for children less than 3 years of age to have more severe obstructive sleep apnea as documented by polysomnography. Central apnea also appears to be more common in this age group. These findings may be explained by anatomic and physiologic differences related to age and support a period of observation following adenotonsillectomy in younger children. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Obstructive sleep apnea Pediatric Sleep disordered breathing Tonsillectomy and adenoidectomy Polysomnogram Adenotonsillar hypertrophy
1. Introduction Adenotonsillectomy (T&A) is a very common surgical procedure performed in children. In recent years, there has been a greater
§ This paper was presented at the American Academy of Otolaryngology Meeting, San Diego, CA, September 2003. * Corresponding author at: Childrens Hospital Los Angeles, 4650 Sunset Blvd., MS #58, Los Angeles, CA 90027, United States. Tel.: +1 323 661 2145. E-mail address: [email protected] (D.M. Don).
0165-5876/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2009.04.003
awareness of the potential health risks associated with obstructive sleep apnea (OSA) and this condition has now surpassed recurrent tonsillitis as the most frequent indication for surgery [1]. In most cases, this surgery can be safely performed as an outpatient procedure. However, there are some children who are at higher risk for postoperative complications and should not be considered candidates for outpatient surgery. Of primary concern postoperatively for these children is respiratory compromise. Children with OSA can experience complications related to post-obstructive pulmonary edema, airway swelling, and poor ventilatory responses after the administration of
1026
D.M. Don et al. / International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028
anesthetic agents. Certainly, it would be beneficial to have the ability to identify those children who are most at risk for problems. Some reviews regarding postoperative complications seem to indicate that age is an important risk factor, with the youngest children being at greatest risk. Although a number of studies are available, they are based on small sample sizes and many did not evaluate the children with polysomnography preoperatively [2–7]. The purpose of this study, therefore, was to investigate age specific OSA abnormalities in children by polysomnography. If differences do exist across various age groups, it may be possible to improve our recognition of children at higher risk and expand our understanding of why select groups of children have greater morbidity. 2. Materials and methods The polysomnograms of children (1–18 years of age) with OSA diagnosed by overnight polysomnography between January 1998 and January 2001 were reviewed. Children included in the study had evidence of adenotonsillar hypertrophy but no other coexisting medical problems. Children with neuromuscular disorders, craniofacial syndromes, prematurity, cardiopulmonary disease, failure to thrive, hematological disorder, malignancy and obesity were excluded. The study was approved by Childrens Hospital Los Angeles Committee on Clinical Investigations (Institutional Review Board). Overnight polysomnography was performed in all children. All testing was performed in the Sleep Physiology Laboratory with continued observation of the child by a polysomnography technician skilled in pediatric polysomnography. Polysomnograms were performed in a quiet, dark room at an ambient temperature of 24 8C. The following parameters were measured and recorded continuously by the Healthdyne computerized polysomnography system (Alice III: Respironics; Marietta, GA): (1) chest and abdominal wall motion by uncalibrated respiratory inductance plethysmography; (2) heart rate, by ECG; (3) end-tidal carbon dioxide pressure (PETCO2), sampled at the nose or mouth by capnography; (4) combined oral nasal air flow, sampled with a three-pronged thermistor (Healthdyne Technologies; Marietta, GA) placed at the upper lip; (5) arterial oxygen saturation (SaO2), by pulse oximetry (model N 200: Nellcor); (6) oximeter pulse wave form; (7) electro-oculogram; (8) electro-encephalogram; (9) chin electromyogram; (10) actimeter (placed on the hand); and (11) sensor placed over neck to monitor snoring. The following parameters were evaluated: (1) obstructive apnea, defined as complete cessation of air flow at the nose and mouth for two or more respiratory cycles; (2) obstructive hypopnea, defined as a reduction in airflow on thermistor for 6 s (tracing to <50% of the baseline); (3) number and duration of central apneas >10 s and central apneas of any length associated with bradycardia or desaturation; (4) number of mixed apneas (apneas with both central and obstructive components); (5) obstructive apnea/ hypopnea index (number of obstructive apneas, mixed apneas and hypopneas per hour of sleep (AHI)); (6) hypoventilation (PETCO2 >45 mmHg) with the highest PETCO2 scored; (7) oxygen desaturations (SaO2 < 95%) and SaO2 nadir; and (8) presence or absence of snoring, paradoxical breathing, chest wall retractions, and gasping. The total length of each study was also recorded. The mean values of the measurements for each age were computed and the Student’s t-test was applied to test for significance between ages. Various combinations of F-ratios were calculated and compared to determine if a significant disparity in variance existed across age groups. The data was further distributed such that children were subdivided into the following age groups: 1–2, 3–5, 6–11 and 12– 18 years. Polysomnographic findings were classified as normal or
abnormal based on criteria previously established [8]. A normal polysomnogram was defined by an AHI <1, SaO2 nadir >94 or highest PETCO2 <46. Mild OSAS was defined by an AHI of 1–2, SaO2 nadir of 90–94% or highest PETCO2 of 46–49. Moderate OSAS was defined by and AHI 3–5, SaO2 nadir of 85–89% or highest PETCO2 of 50–54. Severe OSAS was defined by an AHI >5, SaO2 nadir <85 and highest PETCO2 >54. Normal studies or those classified as having no evidence of OSAS were identified when all three measurements fell into the normal category. Studies in which at least one of the three parameters was categorized as mild but with no other measurement worse than mild, were defined as mild OSAS. Studies in which at least one of the three measurements was categorized as moderate were defined as moderate OSAS. The moderate OSAS category was further subdivided into moderate OSAS 1 (at least one or more measurements in the moderate range but none in the severe group), moderate OSAS 2 (at least one measurement in the severe range) and moderate OSAS 3 (at least two measurements in the severe range). Finally, severe OSAS was characterized when all three measurements fell into the severe category. Analysis was performed to determine if the frequency distribution of these categories was significantly different between age groups using the chi square method. 3. Results The overnight polysomnograms of 363 children were reviewed. Forty-five (12.4%) children were ages 1–2 years, 159 (44%) children were ages 3–5 years, 137 (38%) children were 6–11 years and 22 (6%) children were 12–18 years. The age range of children studied was 1.2–18.5 years with a mean of 4.5 years. The average duration of the polysomnograms was 350 min. Although there is a trend towards a greater mean number of obstructive apneas, hypopneas, central apneas, mixed apneas, a higher mean AHI, lower mean SaO2 nadir, and a higher mean PETCO2 in the younger age groups when compared to the older groups, a Student’s t-test demonstrates that there is no statistical significance for most OSA parameters (Table 1). There are a few exceptions which include mean AHI, central apnea and mixed apnea, where significant differences (p < 0.01) between ages 1–2 are noted when compared to ages 6– 11 and 12–18. An analysis of variance using the F-test reveals statistical significance (p < 0.01) when children ages 1–2 were compared to those 3–5, 6–11 or 12–18 years of age for the variables Table 1 Comparison of mean values between age groups for OSA parameters. OSA parameter
Measured t ratio tA/B
tA/C
tA/D
AHI 2.32 3.13 2.72 CENT 2.01 3.20 5.31 HYPT 1.94 2.08 1.79 MXDT 1.31 1.69 2.64 OAT 1.40 2.72 2.24 OAL 1.87 1.90 1.25 HYPL 0.28 1.37 0.66 CSAL 1.70 0.92 1.44 MXDL 0.75 0.06 0.45 MSAT 1.00 1.06 1.14 LSAT 3.10 3.15 1.73 MPET 1.70 1.92 2.44 HPET 1.14 1.50 0.79 Student’s t-test at 1% significance level 2.41–2.42 2.41–2.42 2.41–2.42 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi t x=y ¼ ðx¯ yÞ= ðS2x =nx Þ þ ðS2x =ny Þ. A: age 1–2 (n = 45); B: age 3–5 (n = 159); C: age ¯ 6–11 (n = 137); D: age 12–18 (n = 20). AHI: apnea/hypopnea index; CENT: central apnea/h; HYPT: hypopnea/h; MXDT: mixed apnea/h; OAT: obstructive apnea/h; OAL: average duration of obstructive apnea; HYPL: average duration of hypopnea; CSAL: average duration of central apnea; MXDL: average duration of mixed apnea; MSAT: mean O2 saturation; LSAT: O2 saturation nadir; MPET: mean PETCO2; HPET: peak PETCO2.
D.M. Don et al. / International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028 Table 2 Variance ratios for OSA parameters with significant deviation (p < 0.01) between age groups. OSA parameter
AHI CENT HYPT MXDT F-Ratio at 1% significance level
Sampling variance ratio S2A =S2B
S2A =S2C
S2A =S2D
2.14 2.25 7.15 2.09 1.70
3.84 2.50 6.63 4.06 1.72
2.96 16.96 3.67 14.64 2.74
2 S2X ¼ eðx xÞ ¯ =ðnX 1Þ. A: age 1–2 (n = 45); B: age 3–5 (n = 159); C: age 6–11 (n = 137); D: age 12–18 (n = 20). AHI: apnea hypopnea index; CENT: central apnea/ h; HYPT: hypopnea/h; MXDT: mixed apnea/h.
Table 3 Variance ratios for OSA parameters without significant deviation (p > 0.01) between age groups. OSA parameter
OAT OAL HYPL CSAL MXDL MSAT LSAT MPET HPET F-Ratio at 1% significance level
Sampling variance ratio S2A =S2B
S2A =S2C
S2A =S2D
0.99 0.83 1.37 0.63 0.64 1.31 1.10 1.19 1.24 1.70
3.94 0.49 1.82 0.51 0.92 2.33 0.76 1.16 0.91 1.72
2.77 0.39 1.33 0.61 0.57 2.59 0.71 1.63 2.42 2.74
2 S2X ¼ eðx xÞ ¯ =ðnX 1Þ. A: age 1–2 (n = 45); B: age 3–5 (n = 159); C: age 6–11 (n = 137); D: age 12–18 (n = 20). OAT: obstructive apnea/h; OAL: average duration of obstructive apnea; HYPL: average duration of hypopnea; CSAL: average duration of central apnea; MXDL: average duration of mixed apnea; MSAT: mean O2 saturation; LSAT: O2 saturation nadir; MPET: mean PETCO2; HPET: peak PETCO2.
AHI, mean number of central apneas, hypopneas and mixed apneas (Table 2). Analysis of variance tests for other OSA parameters did not reveal significance except for comparison of ages 1–2 with 6– 11 and 12–18 for mean obstructive apnea (Table 3). Fig. 1 displays the percentage of patients by age in the various categories. Children ages 1–2 years show a distinct distribution with a larger percentage in the moderate to severe categories. With an increase in age, the distribution of patients within the different categories changes with a smaller percentage of children in the moderate to severe groups and a greater percentage in the mild to normal range. Chi square analysis reveals a significant difference between the frequency distribution of children in age group 1–2 years and that of the other age groups (p < 0.01). However, a comparison of the frequency distribution between age groups 3–5, 6–11 and 12–18 failed to show statistical significance.
Fig. 1. Bar graph representing the percentage of patients by age in the normal, mild, moderate and severe OSA categories.
1027
4. Discussion/conclusion Respiratory compromise after T&A has been reported in as few as 0.4% to as many as 27% of children [2–7,9,10]. Children with OSA are at risk for post-operative respiratory complications for multiple reasons. Pulmonary edema may occur following relief of severe upper airway obstruction. Anesthetic agents can lead to a decrease in neuromuscular tone and transient worsening of upper airway collapse. Impaired ventilatory responses to CO2 after removal of upper airway obstruction may occur [11]. Children with severe OSA may present with cor pulmonale and heart failure which can cause an increased risk for postoperative hypoxemia and ventilatory difficulties. This group may also have more pronounced respiratory depression with narcotic analgesia. Finally, anatomic factors including a smaller oropharyngeal diameter and mandible cause an increase in airway obstruction that are easily exacerbated by postoperative edema. Many authors have reported higher rates of respiratory compromise after T&A in children 3 years of age or younger. As a consequence, it has generally been recommended to observe these children with a planned post-operative hospital admission [2–7]. However, the preponderance of studies on post-operative T&A complications have been retrospective, did not evaluate children preoperatively with polysomnography, or were based on relatively small sample sizes. As a consequence, a consensus regarding post-operative admission for OSA patients has not been established. In an attempt to help resolve this dilemma and improve the recognition and management of children at greatest risk for post-operative complications, we wanted to determine if there are age specific differences in OSA as measured by polysomnography. Based on a large sample of polysomnographic data, this study confirms that children less than 3 years of age differ significantly from older children. This group demonstrates a marked difference in most all polysomnographic parameters when compared to children ages 3–18. In general, these younger children demonstrated sleep study parameters that allowed for classification into severe OSA. Interestingly, although there is a trend towards improvement in the severity of the polysomnographic parameters with increasing age, there is a striking number of children in the older age groups who continue to exhibit moderate to severe OSA. Unfortunately, a major limitation of this investigation is that information regarding the occurrence of post-operative complications in the study population is lacking. This information would have been imperative for evaluating the efficiency of polysomnography for predicting respiratory compromise and for identifying factors that precipitate complications. However, at least one previous study demonstrated the positive predictive value of polysomnography and confirmed that OSA severity correlated well with the incidence of respiratory compromise. These authors report that children with mild, moderate and severe OSA were associated with a 6, 14, and 31% incidence of respiratory compromise, respectively. They further note that a preoperative apnea and hypopnea index of 5 or more and a preoperative oxygen saturation nadir of 80% or less increased the chance of postoperative respiratory complications [11]. If polysomnography is a useful predictor of post-operative risk, there remain significant challenges for its routine use for all adenotonsillectomies. Although polysomnography is currently considered the gold standard for diagnosis of OSA, its expense and inaccessibility make it a test which may not be feasible in all cases. The cost of a polysomnogram at our institution is estimated to be approximately $700–900 and the average waiting period to obtain a study is 3 months. There also remain significant challenges regarding standardization of definitions, parameters and methodologies for polysomnograms. As a consequence,
1028
D.M. Don et al. / International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028
multiple investigators have attempted to evaluate more costefficient alternatives of diagnosing OSA through video recordings, home sleep studies, and nocturnal pulse oximetry [12–14]. A few investigations seemed to indicate that overnight pulse oximetry could estimate the severity of OSA and predict the risk of respiratory complications [11,13,15]. Nevertheless, further studies must be performed to assess the diagnostic and prognostic value of these methodologies in comparison to polysomnography. One unique finding of this study is the surprising incidence of central apneas observed in children with OSA. This is a discovery that has not been previously reported in the literature. Central apneas have been studied in detail in neonates and infants and are considered pathologic in this age group if they are greater than 20 s in duration or are of any length and associated with hypoxia or bradycardia [16]. Only a few studies have evaluated central apnea in children >1 year of age. Other authors indicate that normal children can have up to 5 central apneas of more than 10 s duration per night of sleep [17,18]. Children with OSA in the present study demonstrated a higher mean central apnea rate with an average of 2.5 central apnea/h. The higher incidence of central apnea in our patient population may be due to inhibitory mechanoreceptors arising in the upper airway that stimulate the respiratory center to produce central apnea when pharyngeal collapse and obstruction occurs. In this study, children less than 3 years of age also had an increased frequency of central apnea than their older counterparts. The significance of this finding is uncertain but most likely represents an immaturity of the central nervous system in younger children. This finding may be a principal factor contributing to the greater incidence of respiratory complications observed in this age group. Perhaps in young children, central apnea is further exacerbated by anesthetic agents resulting in a greater risk of respiratory compromise after T&A. In support of this notion are reports that children with OSA have reduced spontaneous ventilation and elevated CO2 in the presence of inhalational anesthesia compared with control subjects. Furthermore, opioid analgesics appear to cause an increase in central apnea in a majority of OSA subjects [19]. This study has clearly defined children less than 3 years of age as a distinct population based on polysomnographic criteria. Despite these findings, there are still some that would argue that a majority of young children may still be managed on an outpatient basis [20,21]. They contend that postoperative complications are usually apparent in the recovery room within the first few hours after surgery, making discharge home for these children safe after a period of observation. Although it may be true that most complications may be evident early on after surgery, some authors have reported delayed respiratory complications that began several hours after surgery. There is also some anecdotal experience with postobstructive pulmonary edema that indicates that its onset or recognition may occur over 8 h following relief of airway obstruction. Of course, the pathophysiology of postobstructive pulmonary edema is not completely known and is most likely multifactorial and not related exclusively to the severity of OSA [22–24]. Lastly, in discussing respiratory compromise one must also not discount the paramount contribution that variations in surgical technique, intraoperative anesthesia and recovery room care play in a patient’s overall post-operative outcome. Our review of a large series of pediatric polysomnograms indicates that there are clear differences in OSA between age groups. The present findings are in accordance with what is known thus far in the literature regarding the young child and respiratory complications. This polysomnographic data, furthermore, allows a
greater understanding of why these young children are at greater risk after surgery. With our present results and persistent inability to predict which children will fair poorly, we advocate that all children less than 3 years of age with obstructive sleep apnea continue to be treated as a high-risk population for post-operative respiratory compromise. This group of young children warrants extra vigilance and consideration should always be given for a planned hospital admission after T&A surgery.
References [1] R.M. Rosenfeld, R.P. Green, Tonsillectomy and adenoidectomy:changing trends, Ann. Otol. Rhino. Laryngol. 99 (1990) 187–191; L.W.C. Tom, R.M. DeDio, D.E. Cohen, R.F. Wetmore, S.D. Handler, W.P. Potsic, Is outpatient tonsillectomy appropriate for young children? Laryngoscope 102 (1992) 277–280. [2] B.J. Wiatrak, C.M. Myer, T.M. Andrews, Complications of adenotonsillectomy in children under 3 years of age, Am. J. Otolaryngol. 12 (1991) 170–172. [3] M.E. Gerber, D.M. O’Connor, E. Adler, C.M. Myer, Selected risk factors in pediatric adenotonsillectomy, Arch. Otolaryngol. Head Neck Surg. 122 (1996) 811–814. [4] G.M. Rosen, R.P. Muckle, M.W. Mahowald, G.S. Goding, C. Ullevig, Postoperative respiratory compromise in children with obstructive sleep apnea syndrome: can it be anticipated? Pediatrics 93 (1994) 784–788. [5] S.A. McColley, M.M. April, J.L. Carroll, R.M. Naclerio, G.M. Loughlin, Respiratory compromise after adenotonsillectomy in children with obstructive sleep apnea, Arch. Otolaryngol. Head Neck Surg. 118 (1992) 940–943. [6] M.J. Biavati, S.C. Manning, D.L. Phillips, Predictive factors for respiratory complications after tonsillectomy and adenoidectomy in children, Arch. Otolaryngol. Head Neck Surg. 123 (1997) 517–521. [7] M.M. Statham, R.G. Elluru, R. Buncher, M. Kalra, Adenotonsillectomy for obstructive sleep apnea in young children, Arch. Otolaryngol. Head Neck Surg. 132 (2006) 476–480. [8] C.L. Marcus, K.J. Omlin, D.J. Basinki, S.L. Bailey, A.B. Rachal, W.S. Von Pechmann, et al., Normal polysomnographic values in children and adolescents, Am. Rev. Respir. Dis. 146 (1992) 1235–1239. [9] W.E. Crysdale, D. Russel, Complications of tonsillectomy and adenoidectomy in 9409 children observed overnight, Can. Med. Assoc. J. 135 (1986) 1139–1142. [10] P.J. Nicklaus, F.S. Herzon, E.W. Steinle, Short-stay outpatient tonsillectomy, Arch. Otolaryngol. Head Neck Surg. 121 (1995) 521–524. [11] K. Wilson, I. Lakheeram, A. Morielli, R. Brouillette, K. Brown, Can assessment for obstructive sleep apnea help predict postadenotonsillectomy respiratory complications? Anesthesiology 96 (2002) 315–322. [12] S.V. Jacob, A. Morielli, M.A. Mograss, F.M. Ducharme, M.D. Schloss, R.T. Brouillette, Home testing for pediatric obstructive sleep apnea syndrome secondary to adenotonsillar hypertrophy, Pediatr. Pulmonol. 20 (1995) 241–252. [13] R.T. Brouillette, A. Morielli, A. Leimanis, K.A. Waters, R. Luciano, F.M. Ducharme, Nocturnal pulse oximetry as an abbreviated testing modality for pediatric obstructive sleep apnea, Pediatrics 105 (2000) 405–412. [14] A. Morielli, S. Ladan, F.M. Ducharme, R.T. Brouillette, Can sleep and wakefulness be distinguished in children by cardiorespiratory and videotape recordings? Chest 109 (1996) 680–687. [15] G.M. Nixon, A.S. Kermack, G.M. Davis, J.J. Manoukian, K.A. Brown, R.T. Brouillette, Planning adenotonsillectomy in children with obstructive sleep apnea: the role of overnight oximetry, 113 (2004) e19–e25. [16] American Academy of Pediatrics Task Force on Prolonged Apnea, Pediatrics 61 (1978) 651–652. [17] M.A. Carskadon, K. Harvey, W.C. Dement, C. Guilleminault, F.B. Simmons, T.F. Anders, Respiration during sleep in children, West. J. Med. 128 (1978) 477–481. [18] E. Tabachnik, N.L. Muller, A.C. Bryan, H. Levison, Changes in ventilation and chest wall mechanics during sleep in normal adolescents, J. Appl. Physiol. 51 (1981) 557–564. [19] K.A. Waters, McBrianF, P. Stewart, M. Hinder, S. Wharton, Effects of obstructive sleep apnea, inhalational anesthesia and fentanyl on the airway and ventilation of children, J. Appl. Physiol. 92 (2002) 1987–1994. [20] M.L. Lalakea, I. Marquez-Biggs, A.H. Messner, Safety of pediatric short-stay tonsillectomy, Arch. Otolaryngol. Head Neck Surg. 125 (1999) 749–752. [21] N.L. Shapiro, A.B. Seid, S.M. Pransky, D.B. Kearns, A.E. Magit, P. Silva, Adenotonsillectomy in the very young patient: cost analysis of two methods of postoperative care, Int. J. Pediatr. Otorhinolaryngol. 48 (1999) 109–115. [22] K.A. Brown, I.M. Morin, C. Hickey, J.J. Manoukian, G.M. Nixon, R.T. Brouillette, Urgent adenotonsillectomy:an analysis of risk factors associated with postoperative respiratory morbidity, Anesthesiology 99 (2003) 586–595. [23] G.M. Nixon, A.S. Kermack, C.D. McGregor, G.M. Davis, J.J. Manoukian, K.A. Brown, et al., Sleep and breathing on the first night after adenotonsillectomy for obstructive sleep apnea, Ped. Pulmonol. 39 (2005) 332–338. [24] J.R. Holmes, R.N. Hensinger, E.W. Wojtys, Postoperative pulmonary edema in young, athletic adults, Am. J. Sports Med. 19 (1991) 365–371.
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Age specific differences in pediatric obstructive sleep apnea§ Debra M. Don *, Kenneth A. Geller, Jeffrey A. Koempel, Sally Davidson Ward Division of Pediatric Otolaryngology and Pulmonology, Childrens Hospital Los Angeles, Keck School of Medicine, University of Southern California, Los Angeles, CA, United States
A R T I C L E I N F O
A B S T R A C T
Article history: Received 4 September 2008 Received in revised form 1 April 2009 Accepted 5 April 2009 Available online 1 May 2009
Background: Some have suggested that younger children have a more severe form of obstructive sleep apnea than older children and therefore are at a higher risk for respiratory compromise after tonsillectomy and adenoidectomy. However, at present there are few studies that have identified any significant correlation between age and severity of obstructive sleep apnea. Objective: To determine if age specific differences in obstructive sleep apnea are present in children. Design: Retrospective chart review. Setting: Tertiary care children’s hospital. Patients: The records of children (1–18 years of age) with obstructive sleep apnea diagnosed by overnight polysomnography between January 1998 and January 2001 were reviewed. Children included in the study also had evidence of adenotonsillar hypertrophy and had no other co-existing medical problems. Main outcome measures: Overnight polysomnography was performed in all children. Apnea–hypopnea index (AHI), baseline and lowest O2 saturation, baseline and peak end tidal CO2, and total number of obstructive apneas, hypopneas, central apneas and mixed apneas were measured during each polysomnogram. Children were subdivided into the following age groups: 1–2, 3–5, 6–11 and 12–18 years. Polysomnograms were classified into normal, mild, moderate and severe categories. Results: Three hundred and sixty-three children were studied; 45 children were ages 1–2 years, 159 children were ages 3–5 years, 137 children were 6–11 years and 22 children were 12–18 years. Although there appears to be a trend towards a greater mean number of obstructive apneas, hypopneas, central apneas, mixed apneas, a higher mean AHI, lower mean SaO2 nadir, and a higher mean PETCO2 in the younger age groups when compared to the older groups, a Student’s t-test demonstrates that there is no statistical significance for most OSA parameters. An analysis of variance using the F-test reveals statistical significance (p < 0.01) when children ages 1–2 were compared to those 3–5, 6–11 or 12–18 years of age for the variables AHI, mean number of central apneas, hypopneas and mixed apneas. When comparing patients in the various severity categories, children ages 1–2 years show a distinct distribution with a larger percentage in the moderate to severe categories. Chi square analysis reveals a significant difference between the frequency distribution of children in age group 1–2 years and that of the other age groups (p < 0.01). Conclusion: There is a predilection for children less than 3 years of age to have more severe obstructive sleep apnea as documented by polysomnography. Central apnea also appears to be more common in this age group. These findings may be explained by anatomic and physiologic differences related to age and support a period of observation following adenotonsillectomy in younger children. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Obstructive sleep apnea Pediatric Sleep disordered breathing Tonsillectomy and adenoidectomy Polysomnogram Adenotonsillar hypertrophy
1. Introduction Adenotonsillectomy (T&A) is a very common surgical procedure performed in children. In recent years, there has been a greater
§ This paper was presented at the American Academy of Otolaryngology Meeting, San Diego, CA, September 2003. * Corresponding author at: Childrens Hospital Los Angeles, 4650 Sunset Blvd., MS #58, Los Angeles, CA 90027, United States. Tel.: +1 323 661 2145. E-mail address: [email protected] (D.M. Don).
0165-5876/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2009.04.003
awareness of the potential health risks associated with obstructive sleep apnea (OSA) and this condition has now surpassed recurrent tonsillitis as the most frequent indication for surgery [1]. In most cases, this surgery can be safely performed as an outpatient procedure. However, there are some children who are at higher risk for postoperative complications and should not be considered candidates for outpatient surgery. Of primary concern postoperatively for these children is respiratory compromise. Children with OSA can experience complications related to post-obstructive pulmonary edema, airway swelling, and poor ventilatory responses after the administration of
1026
D.M. Don et al. / International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028
anesthetic agents. Certainly, it would be beneficial to have the ability to identify those children who are most at risk for problems. Some reviews regarding postoperative complications seem to indicate that age is an important risk factor, with the youngest children being at greatest risk. Although a number of studies are available, they are based on small sample sizes and many did not evaluate the children with polysomnography preoperatively [2–7]. The purpose of this study, therefore, was to investigate age specific OSA abnormalities in children by polysomnography. If differences do exist across various age groups, it may be possible to improve our recognition of children at higher risk and expand our understanding of why select groups of children have greater morbidity. 2. Materials and methods The polysomnograms of children (1–18 years of age) with OSA diagnosed by overnight polysomnography between January 1998 and January 2001 were reviewed. Children included in the study had evidence of adenotonsillar hypertrophy but no other coexisting medical problems. Children with neuromuscular disorders, craniofacial syndromes, prematurity, cardiopulmonary disease, failure to thrive, hematological disorder, malignancy and obesity were excluded. The study was approved by Childrens Hospital Los Angeles Committee on Clinical Investigations (Institutional Review Board). Overnight polysomnography was performed in all children. All testing was performed in the Sleep Physiology Laboratory with continued observation of the child by a polysomnography technician skilled in pediatric polysomnography. Polysomnograms were performed in a quiet, dark room at an ambient temperature of 24 8C. The following parameters were measured and recorded continuously by the Healthdyne computerized polysomnography system (Alice III: Respironics; Marietta, GA): (1) chest and abdominal wall motion by uncalibrated respiratory inductance plethysmography; (2) heart rate, by ECG; (3) end-tidal carbon dioxide pressure (PETCO2), sampled at the nose or mouth by capnography; (4) combined oral nasal air flow, sampled with a three-pronged thermistor (Healthdyne Technologies; Marietta, GA) placed at the upper lip; (5) arterial oxygen saturation (SaO2), by pulse oximetry (model N 200: Nellcor); (6) oximeter pulse wave form; (7) electro-oculogram; (8) electro-encephalogram; (9) chin electromyogram; (10) actimeter (placed on the hand); and (11) sensor placed over neck to monitor snoring. The following parameters were evaluated: (1) obstructive apnea, defined as complete cessation of air flow at the nose and mouth for two or more respiratory cycles; (2) obstructive hypopnea, defined as a reduction in airflow on thermistor for 6 s (tracing to <50% of the baseline); (3) number and duration of central apneas >10 s and central apneas of any length associated with bradycardia or desaturation; (4) number of mixed apneas (apneas with both central and obstructive components); (5) obstructive apnea/ hypopnea index (number of obstructive apneas, mixed apneas and hypopneas per hour of sleep (AHI)); (6) hypoventilation (PETCO2 >45 mmHg) with the highest PETCO2 scored; (7) oxygen desaturations (SaO2 < 95%) and SaO2 nadir; and (8) presence or absence of snoring, paradoxical breathing, chest wall retractions, and gasping. The total length of each study was also recorded. The mean values of the measurements for each age were computed and the Student’s t-test was applied to test for significance between ages. Various combinations of F-ratios were calculated and compared to determine if a significant disparity in variance existed across age groups. The data was further distributed such that children were subdivided into the following age groups: 1–2, 3–5, 6–11 and 12– 18 years. Polysomnographic findings were classified as normal or
abnormal based on criteria previously established [8]. A normal polysomnogram was defined by an AHI <1, SaO2 nadir >94 or highest PETCO2 <46. Mild OSAS was defined by an AHI of 1–2, SaO2 nadir of 90–94% or highest PETCO2 of 46–49. Moderate OSAS was defined by and AHI 3–5, SaO2 nadir of 85–89% or highest PETCO2 of 50–54. Severe OSAS was defined by an AHI >5, SaO2 nadir <85 and highest PETCO2 >54. Normal studies or those classified as having no evidence of OSAS were identified when all three measurements fell into the normal category. Studies in which at least one of the three parameters was categorized as mild but with no other measurement worse than mild, were defined as mild OSAS. Studies in which at least one of the three measurements was categorized as moderate were defined as moderate OSAS. The moderate OSAS category was further subdivided into moderate OSAS 1 (at least one or more measurements in the moderate range but none in the severe group), moderate OSAS 2 (at least one measurement in the severe range) and moderate OSAS 3 (at least two measurements in the severe range). Finally, severe OSAS was characterized when all three measurements fell into the severe category. Analysis was performed to determine if the frequency distribution of these categories was significantly different between age groups using the chi square method. 3. Results The overnight polysomnograms of 363 children were reviewed. Forty-five (12.4%) children were ages 1–2 years, 159 (44%) children were ages 3–5 years, 137 (38%) children were 6–11 years and 22 (6%) children were 12–18 years. The age range of children studied was 1.2–18.5 years with a mean of 4.5 years. The average duration of the polysomnograms was 350 min. Although there is a trend towards a greater mean number of obstructive apneas, hypopneas, central apneas, mixed apneas, a higher mean AHI, lower mean SaO2 nadir, and a higher mean PETCO2 in the younger age groups when compared to the older groups, a Student’s t-test demonstrates that there is no statistical significance for most OSA parameters (Table 1). There are a few exceptions which include mean AHI, central apnea and mixed apnea, where significant differences (p < 0.01) between ages 1–2 are noted when compared to ages 6– 11 and 12–18. An analysis of variance using the F-test reveals statistical significance (p < 0.01) when children ages 1–2 were compared to those 3–5, 6–11 or 12–18 years of age for the variables Table 1 Comparison of mean values between age groups for OSA parameters. OSA parameter
Measured t ratio tA/B
tA/C
tA/D
AHI 2.32 3.13 2.72 CENT 2.01 3.20 5.31 HYPT 1.94 2.08 1.79 MXDT 1.31 1.69 2.64 OAT 1.40 2.72 2.24 OAL 1.87 1.90 1.25 HYPL 0.28 1.37 0.66 CSAL 1.70 0.92 1.44 MXDL 0.75 0.06 0.45 MSAT 1.00 1.06 1.14 LSAT 3.10 3.15 1.73 MPET 1.70 1.92 2.44 HPET 1.14 1.50 0.79 Student’s t-test at 1% significance level 2.41–2.42 2.41–2.42 2.41–2.42 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi t x=y ¼ ðx¯ yÞ= ðS2x =nx Þ þ ðS2x =ny Þ. A: age 1–2 (n = 45); B: age 3–5 (n = 159); C: age ¯ 6–11 (n = 137); D: age 12–18 (n = 20). AHI: apnea/hypopnea index; CENT: central apnea/h; HYPT: hypopnea/h; MXDT: mixed apnea/h; OAT: obstructive apnea/h; OAL: average duration of obstructive apnea; HYPL: average duration of hypopnea; CSAL: average duration of central apnea; MXDL: average duration of mixed apnea; MSAT: mean O2 saturation; LSAT: O2 saturation nadir; MPET: mean PETCO2; HPET: peak PETCO2.
D.M. Don et al. / International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028 Table 2 Variance ratios for OSA parameters with significant deviation (p < 0.01) between age groups. OSA parameter
AHI CENT HYPT MXDT F-Ratio at 1% significance level
Sampling variance ratio S2A =S2B
S2A =S2C
S2A =S2D
2.14 2.25 7.15 2.09 1.70
3.84 2.50 6.63 4.06 1.72
2.96 16.96 3.67 14.64 2.74
2 S2X ¼ eðx xÞ ¯ =ðnX 1Þ. A: age 1–2 (n = 45); B: age 3–5 (n = 159); C: age 6–11 (n = 137); D: age 12–18 (n = 20). AHI: apnea hypopnea index; CENT: central apnea/ h; HYPT: hypopnea/h; MXDT: mixed apnea/h.
Table 3 Variance ratios for OSA parameters without significant deviation (p > 0.01) between age groups. OSA parameter
OAT OAL HYPL CSAL MXDL MSAT LSAT MPET HPET F-Ratio at 1% significance level
Sampling variance ratio S2A =S2B
S2A =S2C
S2A =S2D
0.99 0.83 1.37 0.63 0.64 1.31 1.10 1.19 1.24 1.70
3.94 0.49 1.82 0.51 0.92 2.33 0.76 1.16 0.91 1.72
2.77 0.39 1.33 0.61 0.57 2.59 0.71 1.63 2.42 2.74
2 S2X ¼ eðx xÞ ¯ =ðnX 1Þ. A: age 1–2 (n = 45); B: age 3–5 (n = 159); C: age 6–11 (n = 137); D: age 12–18 (n = 20). OAT: obstructive apnea/h; OAL: average duration of obstructive apnea; HYPL: average duration of hypopnea; CSAL: average duration of central apnea; MXDL: average duration of mixed apnea; MSAT: mean O2 saturation; LSAT: O2 saturation nadir; MPET: mean PETCO2; HPET: peak PETCO2.
AHI, mean number of central apneas, hypopneas and mixed apneas (Table 2). Analysis of variance tests for other OSA parameters did not reveal significance except for comparison of ages 1–2 with 6– 11 and 12–18 for mean obstructive apnea (Table 3). Fig. 1 displays the percentage of patients by age in the various categories. Children ages 1–2 years show a distinct distribution with a larger percentage in the moderate to severe categories. With an increase in age, the distribution of patients within the different categories changes with a smaller percentage of children in the moderate to severe groups and a greater percentage in the mild to normal range. Chi square analysis reveals a significant difference between the frequency distribution of children in age group 1–2 years and that of the other age groups (p < 0.01). However, a comparison of the frequency distribution between age groups 3–5, 6–11 and 12–18 failed to show statistical significance.
Fig. 1. Bar graph representing the percentage of patients by age in the normal, mild, moderate and severe OSA categories.
1027
4. Discussion/conclusion Respiratory compromise after T&A has been reported in as few as 0.4% to as many as 27% of children [2–7,9,10]. Children with OSA are at risk for post-operative respiratory complications for multiple reasons. Pulmonary edema may occur following relief of severe upper airway obstruction. Anesthetic agents can lead to a decrease in neuromuscular tone and transient worsening of upper airway collapse. Impaired ventilatory responses to CO2 after removal of upper airway obstruction may occur [11]. Children with severe OSA may present with cor pulmonale and heart failure which can cause an increased risk for postoperative hypoxemia and ventilatory difficulties. This group may also have more pronounced respiratory depression with narcotic analgesia. Finally, anatomic factors including a smaller oropharyngeal diameter and mandible cause an increase in airway obstruction that are easily exacerbated by postoperative edema. Many authors have reported higher rates of respiratory compromise after T&A in children 3 years of age or younger. As a consequence, it has generally been recommended to observe these children with a planned post-operative hospital admission [2–7]. However, the preponderance of studies on post-operative T&A complications have been retrospective, did not evaluate children preoperatively with polysomnography, or were based on relatively small sample sizes. As a consequence, a consensus regarding post-operative admission for OSA patients has not been established. In an attempt to help resolve this dilemma and improve the recognition and management of children at greatest risk for post-operative complications, we wanted to determine if there are age specific differences in OSA as measured by polysomnography. Based on a large sample of polysomnographic data, this study confirms that children less than 3 years of age differ significantly from older children. This group demonstrates a marked difference in most all polysomnographic parameters when compared to children ages 3–18. In general, these younger children demonstrated sleep study parameters that allowed for classification into severe OSA. Interestingly, although there is a trend towards improvement in the severity of the polysomnographic parameters with increasing age, there is a striking number of children in the older age groups who continue to exhibit moderate to severe OSA. Unfortunately, a major limitation of this investigation is that information regarding the occurrence of post-operative complications in the study population is lacking. This information would have been imperative for evaluating the efficiency of polysomnography for predicting respiratory compromise and for identifying factors that precipitate complications. However, at least one previous study demonstrated the positive predictive value of polysomnography and confirmed that OSA severity correlated well with the incidence of respiratory compromise. These authors report that children with mild, moderate and severe OSA were associated with a 6, 14, and 31% incidence of respiratory compromise, respectively. They further note that a preoperative apnea and hypopnea index of 5 or more and a preoperative oxygen saturation nadir of 80% or less increased the chance of postoperative respiratory complications [11]. If polysomnography is a useful predictor of post-operative risk, there remain significant challenges for its routine use for all adenotonsillectomies. Although polysomnography is currently considered the gold standard for diagnosis of OSA, its expense and inaccessibility make it a test which may not be feasible in all cases. The cost of a polysomnogram at our institution is estimated to be approximately $700–900 and the average waiting period to obtain a study is 3 months. There also remain significant challenges regarding standardization of definitions, parameters and methodologies for polysomnograms. As a consequence,
1028
D.M. Don et al. / International Journal of Pediatric Otorhinolaryngology 73 (2009) 1025–1028
multiple investigators have attempted to evaluate more costefficient alternatives of diagnosing OSA through video recordings, home sleep studies, and nocturnal pulse oximetry [12–14]. A few investigations seemed to indicate that overnight pulse oximetry could estimate the severity of OSA and predict the risk of respiratory complications [11,13,15]. Nevertheless, further studies must be performed to assess the diagnostic and prognostic value of these methodologies in comparison to polysomnography. One unique finding of this study is the surprising incidence of central apneas observed in children with OSA. This is a discovery that has not been previously reported in the literature. Central apneas have been studied in detail in neonates and infants and are considered pathologic in this age group if they are greater than 20 s in duration or are of any length and associated with hypoxia or bradycardia [16]. Only a few studies have evaluated central apnea in children >1 year of age. Other authors indicate that normal children can have up to 5 central apneas of more than 10 s duration per night of sleep [17,18]. Children with OSA in the present study demonstrated a higher mean central apnea rate with an average of 2.5 central apnea/h. The higher incidence of central apnea in our patient population may be due to inhibitory mechanoreceptors arising in the upper airway that stimulate the respiratory center to produce central apnea when pharyngeal collapse and obstruction occurs. In this study, children less than 3 years of age also had an increased frequency of central apnea than their older counterparts. The significance of this finding is uncertain but most likely represents an immaturity of the central nervous system in younger children. This finding may be a principal factor contributing to the greater incidence of respiratory complications observed in this age group. Perhaps in young children, central apnea is further exacerbated by anesthetic agents resulting in a greater risk of respiratory compromise after T&A. In support of this notion are reports that children with OSA have reduced spontaneous ventilation and elevated CO2 in the presence of inhalational anesthesia compared with control subjects. Furthermore, opioid analgesics appear to cause an increase in central apnea in a majority of OSA subjects [19]. This study has clearly defined children less than 3 years of age as a distinct population based on polysomnographic criteria. Despite these findings, there are still some that would argue that a majority of young children may still be managed on an outpatient basis [20,21]. They contend that postoperative complications are usually apparent in the recovery room within the first few hours after surgery, making discharge home for these children safe after a period of observation. Although it may be true that most complications may be evident early on after surgery, some authors have reported delayed respiratory complications that began several hours after surgery. There is also some anecdotal experience with postobstructive pulmonary edema that indicates that its onset or recognition may occur over 8 h following relief of airway obstruction. Of course, the pathophysiology of postobstructive pulmonary edema is not completely known and is most likely multifactorial and not related exclusively to the severity of OSA [22–24]. Lastly, in discussing respiratory compromise one must also not discount the paramount contribution that variations in surgical technique, intraoperative anesthesia and recovery room care play in a patient’s overall post-operative outcome. Our review of a large series of pediatric polysomnograms indicates that there are clear differences in OSA between age groups. The present findings are in accordance with what is known thus far in the literature regarding the young child and respiratory complications. This polysomnographic data, furthermore, allows a
greater understanding of why these young children are at greater risk after surgery. With our present results and persistent inability to predict which children will fair poorly, we advocate that all children less than 3 years of age with obstructive sleep apnea continue to be treated as a high-risk population for post-operative respiratory compromise. This group of young children warrants extra vigilance and consideration should always be given for a planned hospital admission after T&A surgery.
References [1] R.M. Rosenfeld, R.P. Green, Tonsillectomy and adenoidectomy:changing trends, Ann. Otol. Rhino. Laryngol. 99 (1990) 187–191; L.W.C. Tom, R.M. DeDio, D.E. Cohen, R.F. Wetmore, S.D. Handler, W.P. Potsic, Is outpatient tonsillectomy appropriate for young children? Laryngoscope 102 (1992) 277–280. [2] B.J. Wiatrak, C.M. Myer, T.M. Andrews, Complications of adenotonsillectomy in children under 3 years of age, Am. J. Otolaryngol. 12 (1991) 170–172. [3] M.E. Gerber, D.M. O’Connor, E. Adler, C.M. Myer, Selected risk factors in pediatric adenotonsillectomy, Arch. Otolaryngol. Head Neck Surg. 122 (1996) 811–814. [4] G.M. Rosen, R.P. Muckle, M.W. Mahowald, G.S. Goding, C. Ullevig, Postoperative respiratory compromise in children with obstructive sleep apnea syndrome: can it be anticipated? Pediatrics 93 (1994) 784–788. [5] S.A. McColley, M.M. April, J.L. Carroll, R.M. Naclerio, G.M. Loughlin, Respiratory compromise after adenotonsillectomy in children with obstructive sleep apnea, Arch. Otolaryngol. Head Neck Surg. 118 (1992) 940–943. [6] M.J. Biavati, S.C. Manning, D.L. Phillips, Predictive factors for respiratory complications after tonsillectomy and adenoidectomy in children, Arch. Otolaryngol. Head Neck Surg. 123 (1997) 517–521. [7] M.M. Statham, R.G. Elluru, R. Buncher, M. Kalra, Adenotonsillectomy for obstructive sleep apnea in young children, Arch. Otolaryngol. Head Neck Surg. 132 (2006) 476–480. [8] C.L. Marcus, K.J. Omlin, D.J. Basinki, S.L. Bailey, A.B. Rachal, W.S. Von Pechmann, et al., Normal polysomnographic values in children and adolescents, Am. Rev. Respir. Dis. 146 (1992) 1235–1239. [9] W.E. Crysdale, D. Russel, Complications of tonsillectomy and adenoidectomy in 9409 children observed overnight, Can. Med. Assoc. J. 135 (1986) 1139–1142. [10] P.J. Nicklaus, F.S. Herzon, E.W. Steinle, Short-stay outpatient tonsillectomy, Arch. Otolaryngol. Head Neck Surg. 121 (1995) 521–524. [11] K. Wilson, I. Lakheeram, A. Morielli, R. Brouillette, K. Brown, Can assessment for obstructive sleep apnea help predict postadenotonsillectomy respiratory complications? Anesthesiology 96 (2002) 315–322. [12] S.V. Jacob, A. Morielli, M.A. Mograss, F.M. Ducharme, M.D. Schloss, R.T. Brouillette, Home testing for pediatric obstructive sleep apnea syndrome secondary to adenotonsillar hypertrophy, Pediatr. Pulmonol. 20 (1995) 241–252. [13] R.T. Brouillette, A. Morielli, A. Leimanis, K.A. Waters, R. Luciano, F.M. Ducharme, Nocturnal pulse oximetry as an abbreviated testing modality for pediatric obstructive sleep apnea, Pediatrics 105 (2000) 405–412. [14] A. Morielli, S. Ladan, F.M. Ducharme, R.T. Brouillette, Can sleep and wakefulness be distinguished in children by cardiorespiratory and videotape recordings? Chest 109 (1996) 680–687. [15] G.M. Nixon, A.S. Kermack, G.M. Davis, J.J. Manoukian, K.A. Brown, R.T. Brouillette, Planning adenotonsillectomy in children with obstructive sleep apnea: the role of overnight oximetry, 113 (2004) e19–e25. [16] American Academy of Pediatrics Task Force on Prolonged Apnea, Pediatrics 61 (1978) 651–652. [17] M.A. Carskadon, K. Harvey, W.C. Dement, C. Guilleminault, F.B. Simmons, T.F. Anders, Respiration during sleep in children, West. J. Med. 128 (1978) 477–481. [18] E. Tabachnik, N.L. Muller, A.C. Bryan, H. Levison, Changes in ventilation and chest wall mechanics during sleep in normal adolescents, J. Appl. Physiol. 51 (1981) 557–564. [19] K.A. Waters, McBrianF, P. Stewart, M. Hinder, S. Wharton, Effects of obstructive sleep apnea, inhalational anesthesia and fentanyl on the airway and ventilation of children, J. Appl. Physiol. 92 (2002) 1987–1994. [20] M.L. Lalakea, I. Marquez-Biggs, A.H. Messner, Safety of pediatric short-stay tonsillectomy, Arch. Otolaryngol. Head Neck Surg. 125 (1999) 749–752. [21] N.L. Shapiro, A.B. Seid, S.M. Pransky, D.B. Kearns, A.E. Magit, P. Silva, Adenotonsillectomy in the very young patient: cost analysis of two methods of postoperative care, Int. J. Pediatr. Otorhinolaryngol. 48 (1999) 109–115. [22] K.A. Brown, I.M. Morin, C. Hickey, J.J. Manoukian, G.M. Nixon, R.T. Brouillette, Urgent adenotonsillectomy:an analysis of risk factors associated with postoperative respiratory morbidity, Anesthesiology 99 (2003) 586–595. [23] G.M. Nixon, A.S. Kermack, C.D. McGregor, G.M. Davis, J.J. Manoukian, K.A. Brown, et al., Sleep and breathing on the first night after adenotonsillectomy for obstructive sleep apnea, Ped. Pulmonol. 39 (2005) 332–338. [24] J.R. Holmes, R.N. Hensinger, E.W. Wojtys, Postoperative pulmonary edema in young, athletic adults, Am. J. Sports Med. 19 (1991) 365–371.