Eliot S. Katz and Carole L. Marcus

Introduction

Sleep-disordered breathing (SDB) is a common and serious cause of morbidity during childhood. This chapter is concerned with diagnosing the spectrum of obstructive SDB, ranging from the frank, intermittent occlusion seen in obstructive sleep apnea syndrome (OSAS), to persistent, primary snoring (PS). OSAS is characterized by recurrent episodes of partial or complete airway obstruction resulting in hypoxemia, hypercapnia, and/or respiratory arousal (Figure 28-1). The sleep fragmentation and gas exchange abnormalities observed with OSAS may produce serious cardiovascular and neurobehavioral impairment. The upper airway resistance syndrome (UARS) is characterized by brief, repetitive respiratory effort-related arousals (RERA) during sleep in the absence of overt apnea, hypopnea, or gas exchange abnormalities.¹ It has been linked to significant cognitive and behavioral sequelae in children including learning disabilities, attention deficit, hyperactivity, and aggressive behavior.² Obstructive hypoventilation (OH) features prolonged increased upper airway resistance accompanied by gas exchange abnormalities, but not frank apnea or hypopnea.³ Children with PS may have increased respiratory effort, but lack identifiable arousals, including respiratory-effort-related, EEG and subcortical arousals.⁴ Though PS has been traditionally defined as a benign condition, without polysomnographic abnormalities,⁵ recent evidence suggests that the increased respiratory effort in PS per se may be associated with untoward neurobehavioral consequences.^6–9

Habitual snoring has been reported in 3–12% of the general pediatric population, although only 1–3% will have OSAS.10–12 Early recognition of SDB is important insofar as treatment with adenotonsillectomy or continuous positive airway pressure is effective. Establishing strict criteria for OSAS diagnosis and severity is the basis for optimizing the surgical and medical management of this condition.¹³ The diagnosis and management of pediatric OSAS continues to evolve as more precise measures of flow limitation and sleep fragmentation are introduced. Both the intermittent hypoxemia and sleep fragmentation characteristic of OSAS pose a risk to the vulnerable developing brain. The increased recognition of subtle neurocognitive impairments in children with SDB has forced clinicians to rethink the threshold level of disease requiring intervention.

The optimal methodology and criteria for the diagnosis of OSAS in children has not been validated with outcomes data. Categorizing OSAS severity with threshold levels of the apnea–hypopnea index, gas exchange abnormalities, or sleep fragmentation have proven unsatisfactory, since they fail to account for the individual trait susceptibility to the neurocognitive, cardiovascular, and metabolic sequelae of OSAS. That is, the threshold amount of OSAS associated with adverse consequences varies widely among children.

History and Physical Examination

The high incidence of OSAS in children mandates that screening inquiries about sleep disturbances should be a routine part of the primary care interview¹⁴ (Table 28.1). Particular attention should be given to conditions known to exacerbate SDB such as craniofacial abnormalities, neuromuscular weakness, and genetic conditions. A parent typically provides the clinical history, with the patient oblivious to their condition, other than sometimes the complaint of excessive daytime sleepiness. Nightly snoring is observed in most children with OSAS. However, snoring may be absent in the setting of craniofacial abnormalities in infants.¹⁵ A parental report of a snoring child is an accurate predictor of polysomnographic snoring, but not of OSAS.¹⁶ In addition, a population-based study using home-based polysomnography found that loud snoring one to two times per week during the last month was absent in at least 25% of children with documented OSAS.¹⁷ Snoring is often accompanied by labored breathing, hyperextension of the neck, and witnessed apneic pauses. Subjective reports of excessive daytime sleepiness are less common in young children with OSAS, although they are often present in adolescents. A recent study reported that only 7.5% of children with polysomnographically proven OSAS had a history of EDS.¹⁸ Using an objective measure of sleepiness, such as the multiple sleep latency test, reveals that only 13% of children with OSAS have a sleep latency <10 minutes.¹⁸ However, the incidence of objective sleepiness is higher in obese children with OSAS.^19,20 Also, although not considered sleepy by adult standards, children with SDB may be relatively sleepier than normal children.²¹ Increasing BMI and apnea index (usually greater than 15–20 events/hour) have been independently correlated with shorter sleep latencies.¹⁸

The use of standardized screening questionnaires for OSAS has been disappointing. Brouillette et al. presented a questionnaire aimed at distinguishing children with OSAS from normal controls.²² However, subsequent application of this and other questionnaires to a population of snoring children demonstrated a wide-ranging positive predictive value, 48.3–76.9%, and negative predictive value, 26.9–93%.^22–25 Although children with OSAS are statistically more likely to have reported symptoms such as witnessed apnea, cyanosis, and/or labored breathing, no questionnaire has a sufficiently high positive predictive value and negative predictive value to be used as a primary diagnostic tool.^25,26 Thus, the clinical history alone is insufficient to diagnose OSAS amongst a population of snoring children. Questionnaires have also been developed that incorporate questions relating to the consequences of OSA, including hyperactivity and behavior. An affirmative response to at least one-third of these questions on the 22-question Pediatric Sleep Questionnaire had a sensitivity of 0.85 and a specificity of 0.87.27,28

Physical examination of children with OSAS is most often normal, with the exception of adenotonsillar hypertrophy or craniofacial abnormalities (Box 28-1). The majority of children with OSAS are of normal height and weight, although both obesity and failure to thrive may occur. Cardiovascular sequelae of SDB such as cor pulmonale and congestive heart failure are infrequently observed in current clinical practice, as heightened awareness has facilitated earlier diagnosis. Although blood pressure is statistically elevated in children with OSAS, the wide range of normal makes this measurement a poor screening tool.²⁹ The neurocognitive consequences of OSAS are non-specific, such as poor school performance,³⁰ aggressive behavior, and hyperactivity.³¹

Imaging and Laboratory Evaluation

The diagnosis of SDB is firmly established using polysomnography, and ancillary testing is rarely indicated. Further screening may be useful to facilitate the perioperative care and to exclude underlying conditions (Box 28-2). For example, concern regarding right ventricular dysfunction may necessitate an EKG or echocardiogram. Occasionally, chronic intermittent hypoxemia may induce polycythemia, while persistent hypercarbia can elevate serum bicarbonate. Routine pulmonary function testing is not indicated in children suspected of having OSAS unless restrictive or obstructive lung disease is suspected. A fluttering pattern in the flow volume loop has been described in adults with OSAS.³² Magnetic resonance imaging (MRI) of the upper airway in children with OSAS compared to normal controls reveals a statistically smaller upper airway luminal volume and elongation of the soft palate, as well as enlarged tonsils and adenoids.³³ However, there is considerable overlap between the groups, rendering MRI a poor screening tool. The measurement of the ratio of the width of the tonsil to the depth of the pharyngeal space on a lateral neck radiograph was reported to have a good sensitivity and specificity for distinguishing mild from moderate/severe OSAS in a small number of patients.³⁴ Adenoidal enlargement measured by cephalometry was present in over 80% of children with OSAS but also in 42% with primary snoring.³⁵ Though neck radiographs may be suggestive of adenoidal hypertrophy, direct visualization of the adenoids remains the diagnostic standard.³⁶ Dynamic fluoroscopy under sedation may demonstrate glossoptosis, particularly in children with macroglossia, micrognathia, or neuromuscular weakness.³⁷ Anatomical localization of the site of obstruction with cine MRI may alter the therapeutic approach in some children with residual OSAS following adenotonsillectomy or craniofacial anomalies.³⁸ Other promising but understudied tools for identifying OSAS in children include nasal rhinometry³⁹ and acoustic pharyngometry.⁴⁰