ADVERTISEMENT

Issue Navigator

Volume 13 No. 05
Earn CME
Accepted Papers





Scientific Investigations

Upper Airway Reflexes are Preserved During Dexmedetomidine Sedation in Children With Down Syndrome and Obstructive Sleep Apnea

Mohamed Mahmoud, MD1; Stacey L. Ishman, MD, MPH2,3; Keith McConnell, MS3; Robert Fleck, MD4; Sally Shott, MD2; Goutham Mylavarapu, PhD3; Ephraim Gutmark, PhD, DSc2,5; Yuanshu Zou, PhD6; Rhonda Szczesniak, PhD3,6; Raouf S. Amin, MD3
1Department of Anesthesiology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio; 2Division of Otolaryngology – Head and Neck Surgery, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio; 3Division of Pulmonary Medicine, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio; 4Department of Radiology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio; 5Department of Aerospace Engineering and Engineering Mechanics, University of Cincinnati, Cincinnati, Ohio; 6Division of Biostatistics and Epidemiology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio

ABSTRACT

Study Objectives:

The assessment of pharyngeal collapsibility is difficult to perform in children under normal sleep. An alternative is to perform the assessment under an anesthetic, such as dexmedetomidine (DEX), that induces non-rapid eye movement (NREM) sleep. The objectives of this study were to compare critical closing airway pressure (Pcrit) obtained during natural sleep to that obtained under DEX in patients with Down syndrome (DS) and persistent obstructive sleep apnea (OSA) and determine whether Pcrit measured under sedation predicts the severity of OSA.

Methods:

The passive and active Pcrit, which represent airway passive mechanical properties and active dynamic responses to airway obstruction, respectively, were measured. Upper airway reflex activity was estimated by calculating the difference between active and passive Pcrit. Subjects underwent overnight polysomnography during which Pcrit was measured during normal sleep. Pcrit was also measured during DEX sedation at a dose of 2 μg/kg/h.

Results:

The study included 50 patients with median age of 11.4 years (interquartile range: 7.0–13.9) and median body mass index of 23.0 kg/m2 (interquartile range: 18.4–29.1), 66% male and 80% Caucasian. Passive Pcrit was significantly higher than active Pcrit when measured during normal sleep and DEX-induced sleep. There was a positive association between apnea-hypopnea index and passive Pcrit (Spearman r = 0.53, P = .0001) and active Pcrit (r = 0.55, P = .0002) under DEX-induced sleep. There were no significant differences between the Pcrit measurements during natural sleep and during DEX sedation.

Conclusion:

Patients with OSA can compensate for airway obstruction under DEX-induced sleep. The close association between Pcrit and apnea-hypopnea index suggests that airway responses with DEX sedation parallel those seen during natural sleep.

Clinical Trial Registration:

ClinicalTrials.gov identifier: NCT01902407

Citation:

Mahmoud M, Ishman SL, McConnell K, Fleck R, Shott S, Mylavarapu G, Gutmark E, Zou Y, Szczesniak R, Amin RS. Upper airway reflexes are preserved during dexmedetomidine sedation in children with Down syndrome and obstructive sleep apnea. J Clin Sleep Med. 2017;13(5):721–727.


INTRODUCTION

Obstructive sleep apnea (OSA) is increasingly recognized as a common disorder in children. It is characterized by recurrent partial or total collapse of the upper airway1 and is associated with cognitive, metabolic, and cardiovascular morbidity.2

BRIEF SUMMARY

Current Knowledge/Study Rationale: The assessment of pharyngeal collapsibility under anesthesia can provide important information about dynamic patterns of airway collapse in children with obstructive sleep apnea. Dexmedetomidine provides sedative properties that parallel the physiologic changes seen during non-rapid eye movement sleep through action on alpha 2-adrenergic receptors. These properties make dexmedetomidine an attractive drug for sedating children who need dynamic upper airway evaluation.

Study Impact: The responses to airway obstruction and the association between airway collapsibility and obstructive sleep apnea severity during dexmedetomidine sedation in children were unknown. This study demonstrated that dexmedetomidine is a reasonable anesthetic option to sedate children with obstructive sleep apnea, especially those presenting for upper airway evaluation.

Children with OSA respond well to adenotonsillectomy (T&A) as a first-line treatment. However, a large percentage of children have recurrence of OSA several months after T&A. In one study, 27% of lean children and 79% of obese children without genetic syndromes had recurrences of OSA 1 year after T&A.3 In children with genetic syndromes and craniofacial abnormalities, OSA either persists or recurs in a much larger percentage of children.4 A significant proportion of children with Down syndrome (DS) may have persistent or recurrent OSA after T&A. Prior to repeat interventions, clinicians must identify the level(s) of residual obstruction. This is often accomplished through examination of the patterns of dynamic airway collapse during sedation-induced sleep.5,6 Examination often reveals obstruction and upper airway collapsibility secondary to recurrent enlargement of adenoid tissue, lingual tonsillar hypertrophy, reduced muscular tone, hypopharyngeal collapse, glossoptosis, or a combination of these anatomic and dynamic findings factors.5,7,8 Because different sedatives and anesthetic agents may affect respiratory and upper airway control differently, upper airway evaluation during drug-induced sleep may overestimate the contribution of hypopharyngeal hypotonia to the presence of OSA.

Dexmedetomidine (DEX) is a selective alpha-2 adrenergic receptor agonist that provides sedation with minimal respiratory depression.9 Sleep architecture during DEX infusion is exclusively non-rapid eye movement (NREM) and is limited to stage 1 and stage 2 sleep. Neither slow wave nor rapid eye movement sleep are observed during DEX infusion.

The objectives of the current study were to compare airway collapsibility measured by the critical closing pressure (Pcrit) obtained during natural sleep to that obtained under DEX in patients with OSA and determine whether Pcrit measured under sedation predicts the severity of OSA. The overarching hypothesis is that airway collapsibility during normal sleep is not significantly different from collapsibility measured during sedation with DEX.

The patency of the upper airway during sleep is dependent on airway structure, shape, and neuromuscular function. The contribution of each of these factors to airway obstruction during sleep varies by age, sex, the degree to which the patient is overweight, and the presence of associated genetic syndromes.

In this study, we investigated the effect of DEX on airway collapsibility in children and young adults with DS who had persistent OSA after T&A.

METHODS

Study Group

The study enrolled children and young adults with DS, age 1 to 21 years, who have persistent OSA after T&A. Persistent OSA was defined as apnea-hypopnea index (AHI) of ≥ 4 events/h of sleep. Our institution routinely offers continuous positive airway pressure or additional therapy to those patients with an AHI ≥ 5 events/h regardless of symptoms. Thus, we determined that an AHI ≤ 4 events/h indicated they would not need further treatment in the absence of symptoms. Subjects were recruited from a multidisciplinary clinic and were receiving an airway evaluation by magnetic resonance imaging (MRI) under sedation. Children with dental hardware or devices that are incompatible with MRI were excluded from the study. The DYMOSA (Dynamic Computational Modeling of Obstructive Sleep Apnea in Down Syndrome) study is a registered clinical study (NCT01902407) that received approval by the institutional review board at our institution. The subjects' parents provided informed consent.

Eligibility for the study was based on an overnight polysomnography (PSG) during which Pcrit during normal sleep were measured. Pcrit during sedation with DEX was obtained when subjects were being evaluated by MRI.

Polysomnography

All patients underwent overnight 16-channel PSG to establish the diagnosis and severity of OSA. The signals included electroencephalograms, left and right electrooculograms, sub-mental electromyogram, tibial electromyogram, electrocardiogram, airflow, and oxyhemoglobin saturation. End-tidal carbon dioxide was obtained for all participants. Respiratory effort was assessed with thoracic and abdominal inductive plethysmography, and body position was monitored via infrared video camera. Studies were interpreted by board-certified pediatric sleep medicine physicians.

PSG was scored according to the American Academy of Sleep Medicine guidelines. An apnea was defined as complete absence of airflow for at least two breath cycles. Apneas were identified as obstructive when associated with continued or increased inspiratory effort. A mixed apnea was identified when absence of airflow was associated with periods both with and without inspiratory effort. A hypopnea was defined as a decrease in airflow of ≥ 50% for at least two breath cycles followed by a ≥ 3% decrease in oxygen saturation or an electrocortical arousal from sleep. Respiratory event-related arousals were not scored. The AHI was calculated as the sum of the obstructive apneas, mixed apneas, and hypopneas, divided by the total sleep time. OSA severity was stratified by AHI. Mild OSA was defined as 1 to < 5 events/h; moderate OSA was defined as 5 to < 10 events/h; and severe OSA was defined as ≥ 10 events/h. The saturation nadir was defined as the lowest oxygen saturation reading associated with an obstructive respiratory event.

Pharyngeal Critical Pressure Measurements

The critical pressure is defined as the airway pressure measured when there is cessation of airflow as a result of complete airway obstruction. In children with OSA, the Pcrit is higher compared to patients without OSA.10 Pcrit is measured under two different experimental conditions, passive and active states. Passive Pcrit (relative hypotonia) reflects the contribution of mechanical loads on upper airway collapsibility. Active Pcrit represents the combination of mechanical loads and neural reflex responses.

Signals were acquired on a PowerLab system (ADInstruments, Dunedin, New Zealand). Pressure was altered, using a modified Respironics BiPAP Harmony S/T Device (PCRIT 3000, version 1.0, Philips Respironics, Murrysville, Pennsylvania, United States) provided by the manufacturer. Two techniques, passive and active Pcrit, were applied to measure the critical closing pressure. For both techniques, the holding nasal pressure just above the pressure at which flow limitation first became discernible was determined.

Passive Pcrit

For this technique nasal pressure was decreased by 1 cm H2O from the holding pressure for 5 breaths, following which it was rapidly returned to the holding pressure. Nasal pressure was dropped repeatedly to incrementally lower levels, with a return each time to the holding pressure, until flow limitation was observed11 (Figure 1).

Passive and active Pcrit measurements.

This figure shows how passive (measured with acute, transient, intermittent drops in nasal pressure) and active (measured with stepwise, decremental, steady state decreases in nasal pressure) Pcrit measurements were determined during natural sleep and during dexmedetomidine sedation. Pcrit = critical airway closing pressure.

jcsm.13.5.721a.jpg

jcsm.13.5.721a.jpg
Figure 1

Passive and active Pcrit measurements.

(more ...)

Active Pcrit

For this technique, the nasal pressure was by 1 cm H2O steps every 5 breaths until flow limitation was observed (Figure 1). The Pcrit difference was defined as the difference between the active and passive measurements and was considered a measure of the strength of dynamic neuromuscular responses to airway obstruction.12

Pcrit was measured during stage 2 natural sleep. DEX-induced sedation was performed on a separate day at the time of imaging of the upper airway.

Sedation

Prior to cine MRI, sedation was induced with sevoflurane and/or nitrous oxide (N2O) in oxygen. No premedications were administered. An intravenous (IV) catheter was inserted, IV atropine (10 μg/kg) was then administered, and inhalational agents (eg, sevoflurane and N2O) were discontinued. Next, all patients received an IV loading dose of DEX (2 μg/kg) over 10 minutes followed by a DEX infusion at 2 μg/kg/h. After confirmation of adequate sedation, patients were positioned supine with the cervical spine in the neutral position on the MRI table; all patients continued to breathe spontaneously.

All Pcrit measurements were carried out immediately after the conclusion of the loading dose of DEX, and the depth of sedation was assessed by the University of Michigan Sedation Scale (UMSS).13 This is a validated, reliable 5-point scale devised to facilitate rapid assessment and documentation of anesthetic depth of sedation in patients receiving sedative agents for diagnostic or therapeutic procedures. It is scored as follows: 0 denotes an awake/alert state, 1 denotes minimal sedation where the patient is tired and sleepy, with appropriate response to verbal conversation and/or sounds, 2 refers to a state of moderate sedation where the patient is somnolent or sleeping and is easily aroused with light tactile stimulation, 3 is used for a state of deep sedation where the patient is in deep sleep and arousable only with significant physical stimulation, and 4 is used when patients are unarousable.

All patients had an end-tidal sevoflurane concentration of < 0.1% before measuring the Pcrit. During Pcrit measurement, all patients were monitored with continuous electrocardiography, pulse oximetry, and intermittent automated blood pressure. Upon completion of Pcrit measurements, patients were transferred to the MRI suite to undergo a dynamic cine MRI.

Statistical Analysis

Continuous variables were summarized using median values and interquartile ranges (IQR) or mean ± standard deviation (SD), while categorical variables were summarized as n (%). Spearman correlation coefficients were used to estimate the degree of association between AHI and Pcrit outcomes. The three primary outcomes were active Pcrit, passive Pcrit, and the difference (active − passive). Sensitivity analysis was performed excluding the potential outliers identified using diagnostic plots and whether the externally studentized residual magnitudes exceeded 2. Passive and active Pcrit under normal sleep and under DEX were compared through paired two-sample t test.

Each outcome was compared between natural and DEX using a linear mixed effects model that included sleep status (natural versus DEX) as a covariate with adjustment for potential confounders. Patient-specific random intercepts were included to account for between-subject heterogeneity and within-subject correlation. Relevant confounders, included in the model as covariates, were sex, age, body mass index, and AHI. Missing data from Pcrit measurements were accounted for in the mixed model, assuming that the data are missing at random.14,15 This assumption allows for inclusion of all available data for analysis. Backward elimination was then performed to retain only statistically significant covariates (P < .1) and sleep status. The estimated least-squares means, 95% confidence interval, and P value from each model are reported for the effect of sleep status. All analyses were implemented using SAS 9.3 (SAS Institute, Cary, North Carolina, United States).

RESULTS

Subjects

Fifty children and young adults (66% male, 80% Caucasian) ranged in age from 2.4 to 21.9 years with a median age of 11.4 (IQR = 7.0–13.9) and had a median body mass index of 23.0 kg/m2 (IQR = 18.4–29.1). All subjects had an AHI ≥ 4 events/h (Table 1).

PSG results for all participants.

jcsm.13.5.721.t01.jpg

table icon
Table 1

PSG results for all participants.

(more ...)

Pcrit

The mean ± SD of the time between PSG and Pcrit measurement under DEX sedation was 18 ± 22 days. At the time of Pcrit measurement, one child had a UMSS score of 2 (moderate sedation), whereas the other 49 subjects had a UMSS score of 3 (deep sedation).

Pcrit measurements were obtained during DEX sedation in 47 patients. Passive Pcrit measurements were obtained in 47 patients and active Pcrit measurements in 42 patients. Eighteen patients had their Pcrit measured during natural sleep; subjects without measurable Pcrit during natural sleep had significantly shorter total sleep time compared to those who did have measurable Pcrit (Table 2).

Demographic and PSG characteristics of subjects with and without Pcrit measurement during natural sleep.

jcsm.13.5.721.t02.jpg

table icon
Table 2

Demographic and PSG characteristics of subjects with and without Pcrit measurement during natural sleep.

(more ...)

Passive and Active Pcrit Under Normal and DEX-Induced Sleep

There were no significant correlations between the Pcrit outcomes and age or body mass index when children were sedated with DEX and under natural sleep.

Under natural sleep, passive Pcrit (−2.59 ± 4.75 cm H2O) was significantly higher than active Pcrit (−5.97 ± 7.64 cm H2O) (P = .015). Under DEX-induced sleep, passive Pcrit (−2.11 ± 5.86 cm H2O) was also significantly higher than active Pcrit (−3.29 ± 6.39 cm H2O) (P = .02). This indicated that patients were able to compensate for the airway mechanical load by lowering the active Pcrit during DEX sedation.

Comparison Between Pcrit Under Normal and DEX-Induced Sleep

There were no significant differences between Pcrit measurements (passive, active, and difference between passive and active) obtained during natural sleep and those obtained during DEX sedation based on the linear mixed-effect model (Table 3).

Pcrit measurements between DEX-induced sedation and natural sleep, adjusted for confounders.

jcsm.13.5.721.t03.jpg

table icon
Table 3

Pcrit measurements between DEX-induced sedation and natural sleep, adjusted for confounders.

(more ...)

As a sensitivity analysis, comparisons were also performed, restricted only to those subjects who had completed both natural and DEX sleep. Conclusions were similar, as there were no significant differences between Pcrit measurements.

Association Between AHI and Pcrit

There was a positive correlation between AHI and passive Pcrit (r = 0.53, P = .0001) as well as between AHI and active Pcrit (r = 0.55, P = .0002) under DEX sedation (Figure 2). There was, however, no significant correlation between the Pcrit difference and AHI when children were sedated with DEX (r = 0.15, P = .34) (Figure 2).

Association between Pcrit under DEX sedation and OSA severity (all patients).

The association between Pcrit under DEX sedation and OSA severity as defined by the AHI for all patients. AHI = apnea-hypopnea index, DEX = dexmedetomidine, OSA = obstructive sleep apnea, Pcrit = critical airway closing pressure.

jcsm.13.5.721b.jpg

jcsm.13.5.721b.jpg
Figure 2

Association between Pcrit under DEX sedation and OSA severity (all patients).

(more ...)

Correlations excluding identified outliers showed weaker but still significant results for both passive and active Pcrit. There was still a positive correlation between AHI and passive Pcrit (r = 0.48, P = .0008) as well as between AHI and active Pcrit (r = 0.48, P = .0016) under DEX sedation. There was no significant correlation between the difference between passive and active Pcrit and AHI when children were sedated with DEX (r = 0.11, P = .52) neither (Figure 3).

Association between Pcrit under DEX sedation and OSA severity (excluding outliers).

The association between Pcrit under DEX sedation and OSA severity as defined by the AHI after excluding the outliers. AHI = apnea-hypopnea index, DEX = dexmedetomidine, OSA = obstructive sleep apnea, Pcrit = critical airway closing pressure.

jcsm.13.5.721c.jpg

jcsm.13.5.721c.jpg
Figure 3

Association between Pcrit under DEX sedation and OSA severity (excluding outliers).

(more ...)

There were no significant correlation between active Pcrit (r = 0.33, P = .23), passive Pcrit (r = 0.34, P = .20), Pcrit difference (r = 0.43, P = .15), and AHI when children were under natural sleep.

Adverse Events

All patients were hemodynamically stable during Pcrit measurement and were discharged on the same day as their upper airway evaluation, with the exception of one patient. The patient who stayed overnight had an elevated body mass index of 49 kg/m2, and was admitted due to increased oxygen requirements while in the recovery room. None of the patients had significant complications when queried on the day after the Pcrit examination.

DISCUSSION

The study demonstrated that upper airway reflexes in children sedated with DEX are active and contribute to maintaining airway patency. This novel observation provides a strong rationale for using DEX in children with OSA as the sedative of choice because it is less likely to compromise airway patency compared to other anesthetic agents. Furthermore, the evaluation of the upper airway by endoscopy or imagining during DEX induced sleep is likely to be more analogous to the evaluation of airway dynamics during normal sleep.

Additionally, we demonstrated a positive association between Pcrit under DEX and the AHI derived from the overnight PSG. The positive association between OSA severity and Pcrit measurement during DEX sedation suggests that the response of the upper airway tone under DEX sedation parallels that seen during natural sleep.

Currently available sedative and anesthetic agents reduce upper airway dilator muscle activity to varying degrees.1618 These agents induce unconsciousness by altering neurotransmission at multiple areas in the cerebral cortex, brainstem, and thalamus. The molecular targets for DEX are the central α2-adrenergic receptors. The sedative effects of DEX are mediated via stimulation of the α2A-receptor subtype19 in the locus coeruleus.20 Studies of electroencephalogram activity during DEX sedation in healthy adults and children report that DEX sedation results in a state that closely resembles stage 2 sleep.21,22

In previous studies, we performed imaging of the upper airway with MRI using a dose of DEX (1 μg/kg/h) and a higher dose of DEX (3 μg/kg/h). We found that airway dimensions were smaller at the higher dose of DEX. However, the airway changes in airway dimension were small enough not to cause clinically significant airway obstruction.23 In children with OSA, increasing DEX dosing was not associated with any significant changes in airway morphology.24 Collectively, these studies reinforce the observation that DEX provides a relatively safe sedative to children with OSA.

Although a large percentage of children preserve airway reflexes under normal sleep and when sedated with DEX, close to one-third of the children were unable to do so under both conditions. Multiple mechanisms for the failure to preserve upper airway reflexes have been proposed such as decreased pharyngeal mechanoreceptor input secondary to chronic upper airway obstruction, mucosal inflammation,25 and or decreased ventilatory responses to hypercapnia and hypoxemia.26,27

The study has a number of limitations. Based on the knowledge of the effect of DEX on sleep architecture, it is unlikely that the Pcrit measurements were obtained during rapid eye movement sleep where the majority of the obstructive events take place. Therefore, Pcrit during DEX induced sleep might be an underestimate from that during rapid eye movement sleep. We also examined airway collapsibility using only a single dose of DEX, which allowed for an adequate level of sedation during Pcrit measurement. Evaluation at lower doses of DEX may have resulted in a lesser degree of airway collapse than was observed in our study. However, it is likely that awakening from sedation would have occurred more frequently, precluding an accurate and complete measurement of Pcrit. Our findings regarding the difference in Pcrit between natural and DEX sleep may be limited by the number of patients who were able to tolerate the Pcrit protocol during natural sleep. Although there is the potential for bias in these findings, sensitivity analyses indicate that inability to perform Pcrit during natural sleep was not specific to patients with certain characteristics, aside from total sleep time.

Although both passive and active Pcrit measurements under DEX correlated with AHI, they did not significantly correlate with AHI under natural sleep. This lack of correlation is likely due to the small sample size as many children with DS had a low arousal threshold when stimulated during sleep, leading to a high failure rate in obtaining the measurement.

Additionally, all study participants had DS, which limits the generalizability of our findings. Given that children with DS frequently have persistent OSA after treatment, success in this population is even more likely in populations with fewer risk factors for persistent OSA. Last, we did not document the differences in neuromuscular activity between the passive and active state using electromyography, but rather relied on Pcrit to indicate differences in pharyngeal airway muscle dilator tone. Future studies would benefit from measurement of this parameter.

CONCLUSIONS

Our study demonstrated that patients with OSA after T&A can compensate for airway obstruction under DEX-induced sleep. Moreover, the close association between Pcrit and AHI suggests that airway reflexes under DEX sedation parallel those seen during natural sleep. We suggest that DEX is a reasonable anesthetic option to sedate children and young adults with DS and OSA, especially those presenting for upper airway evaluation.

DISCLOSURE STATEMENT

This was not an industry supported study. This study was supported by National Institutes of Health grant (RO1HL105206-01). The project described was supported by Grant 8 UL1 TR000077 from the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health. The authors have indicated no financial conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

DEX

dexmedetomidine

DS

Down syndrome

DYMOSA

Dynamic Computational Modeling of Obstructive Sleep Apnea in Down Syndrome

IQR

interquartile ranges

LS-means

least squares means

MRI

magnetic resonance imaging

N2O

nitrous oxide

NREM

non-rapid eye movement

OSA

obstructive sleep apnea

Pcrit

critical closing airway pressure

PSG

polysomnography

SD

standard deviation

T&A

adenotonsillectomy

UMSS

University of Michigan Sedation Scale

ACKNOWLEDGMENTS

The authors thank the staff who helped with this project: Denise Wetzel, MD and Aliza Cohen for editing of the manuscript; Katie Fields, MHA, Clinical Research Coordinator; Jennifer Kondik, CRNA, Department of Anesthesia, Cincinnati Children's Hospital Medical Center, University of Cincinnati, Cincinnati, Ohio.

REFERENCES

1 

Malhotra A, White DP. Obstructive sleep apnoea. Lancet. 2002;360(9328):237–245. [PubMed]

2 

Marcus CL, Moore RH, Rosen CL, et al. A randomized trial of adenotonsillectomy for childhood sleep apnea. N Engl J Med. 2013;368(25):2366–2376. [PubMed Central][PubMed]

3 

Amin R, Anthony L, Somers V, et al. Growth velocity predicts recurrence of sleep-disordered breathing 1 year after adenotonsillectomy. Am J Respir Crit Care Med. 2008;177(6):654–659. [PubMed Central][PubMed]

4 

Maeda K, Tsuiki S, Nakata S, Suzuki K, Itoh E, Inoue Y. Craniofacial contribution to residual obstructive sleep apnea after adenotonsillectomy in children: a preliminary study. J Clin Sleep Med. 2014;10(9):973–977. [PubMed Central][PubMed]

5 

Donnelly LF. Obstructive sleep apnea in pediatric patients: evaluation with cine MR sleep studies. Radiology. 2005;236(3):768–778. [PubMed]

6 

Truong MT, Woo VG, Koltai PJ. Sleep endoscopy as a diagnostic tool in pediatric obstructive sleep apnea. Int J Pediatr Otorhinolaryngol. 2012;76(5):722–727. [PubMed]

7 

Shott SR, Donnelly LF. Cine magnetic resonance imaging: evaluation of persistent airway obstruction after tonsil and adenoidectomy in children with Down syndrome. Laryngoscope. 2004;114(10):1724–1729. [PubMed]

8 

Donnelly LF, Shott SR, LaRose CR, Chini BA, Amin RS. Causes of persistent obstructive sleep apnea despite previous tonsillectomy and adenoidectomy in children with Down syndrome as depicted on static and dynamic cine MRI. AJR Am J Roentgenol. 2004;183(1):175–181. [PubMed]

9 

Mahmoud M, Mason KP. Dexmedetomidine: review, update, and future considerations of paediatric perioperative and periprocedural applications and limitations. Br J Anaesth. 2015;115(2):171–182. [PubMed]

10 

Smith PL, Wise RA, Gold AR, Schwartz AR, Permutt S. Upper airway pressure-flow relationships in obstructive sleep apnea. J Appl Physiol. 1988;64(2):789–795. [PubMed]

11 

Marcus CL, Katz ES, Lutz J, Black CA, Galster P, Carson KA. Upper airway dynamic responses in children with the obstructive sleep apnea syndrome. Pediatr Res. 2005;57(1):99–107. [PubMed]

12 

Patil SP, Schneider H, Marx JJ, Gladmon E, Schwartz AR, Smith PL. Neuromechanical control of upper airway patency during sleep. J Appl Physiol. 2007;102(2):547–556. [PubMed]

13 

Malviya S, Voepel-Lewis T, Tait AR, Merkel S, Tremper K, Naughton N. Depth of sedation in children undergoing computed tomography: validity and reliability of the University of Michigan Sedation Scale (UMSS). Br J Anaesth. 2002;88(2):241–245. [PubMed]

14 

Little RJA. Modeling the drop-out mechanism in repeated-measures studies. J Am Stat Assoc. 1995;90(431):1112–1121.

15 

Molenberghs G, Kenward M. Missing Data in Clinical Studies. West Sussex, England: John Wiley & Sons Ltd.; 2007.

16 

Eikermann M, Malhotra A, Fassbender P, et al. Differential effects of isoflurane and propofol on upper airway dilator muscle activity and breathing. Anesthesiology. 2008;108(5):897–906. [PubMed Central][PubMed]

17 

Eikermann M, Fassbender P, Zaremba S, et al. Pentobarbital dose-dependently increases respiratory genioglossus muscle activity while impairing diaphragmatic function in anesthetized rats. Anesthesiology. 2009;110(6):1327–1334. [PubMed Central][PubMed]

18 

Eikermann M, Eckert DJ, Chamberlin NL, et al. Effects of pentobarbital on upper airway patency during sleep. Eur Respir J. 2010;36(3):569–576. [PubMed]

19 

Lakhlani PP, MacMillan LB, Guo TZ, et al. Substitution of a mutant alpha2aadrenergic receptor via “hit and run” gene targeting reveals the role of this subtype in sedative, analgesic, and anesthetic-sparing responses in vivo. Proc Natl Acad Sci U S A. 1997;94(18):9950–9955. [PubMed Central][PubMed]

20 

Correa-Sales C, Rabin BC, Maze M. A hypnotic response to dexmedetomidine, an alpha 2 agonist, is mediated in the locus coeruleus in rats. Anesthesiology. 1992;76(6):948–952. [PubMed]

21 

Mason KP, O'Mahony E, Zurakowski D, Libenson MH. Effects of dexmedetomidine sedation on the EEG in children. Paediatr Anaesth. 2009;19(12):1175–1183. [PubMed]

22 

Huupponen E, Maksimow A, Lapinlampi P, et al. Electroencephalogram spindle activity during dexmedetomidine sedation and physiological sleep. Acta Anaesthesiol Scand. 2008;52(2):289–294. [PubMed]

23 

Mahmoud M, Radhakrishman R, Gunter J, et al. Effect of increasing depth of dexmedetomidine anesthesia on upper airway morphology in children. Paediatr Anaesth. 2010;20(6):506–155. [PubMed]

24 

Mahmoud M, Jung D, Salisbury S, et al. Effect of increasing depth of dexmedetomidine and propofol anesthesia on upper airway morphology in children and adolescents with obstructive sleep apnea. J Clin Anesth. 2013;25(7):529–541. [PubMed]

25 

Kimoff RJ, Sforza E, Champagne V, Ofiara L, Gendron D. Upper airway sensation in snoring and obstructive sleep apnea. Am J Respir Crit Care Med. 2001;164(2):250–255. [PubMed]

26 

Garcia-Rio F, Pino JM, Ramirez T, et al. Inspiratory neural drive response to hypoxia adequately estimates peripheral chemosensitivity in OSAHS patients. Eur Respir J. 2002;20(3):724–732. [PubMed]

27 

Gold AR, Schwartz AR, Wise RA, Smith PL. Pulmonary function and respiratory chemosensitivity in moderately obese patients with sleep apnea. Chest. 1993;103(5):1325–1329. [PubMed]