Children with a Clinical Diagnosis of Sleep Apnea May Not Meet Sleep Study Criteria

Despite abundant evidence to show that sleep-disordered breathing (SDB) exists in childhood, the full extent of the effects of the disease on children is not well characterized. Most recent studies have linked the relationship between some measure of SDB and daytime behavior and cognition. However, the details of relationship demand further study.

The accepted standard to classify the quality of sleep is the polysomnography (PSG) or sleep study, where brain waves (EEG), eye movements (EOG), muscle activity (EMG), heartbeat (EKG), blood oxygen levels, and respiration are all measured during sleep. Since obstructive sleep apnea in children was first described nearly 30 years ago, there has not been agreement on the definition of normal versus abnormal polysomnographic criteria. It is not clear to the physician how disordered a child's sleep and breathing must be for him/her to suffer negative clinical consequences. It is also not clear which PSG variables are most predictive of improvement in outcome following treatment.

Despite the lack of standards in interpreting sleep study results, there have been few studies in which large numbers of children are tested before and after intervention for SDB, with the treatment effects associated PSG variables. All agree that an improved understanding of these relationships will be necessary before deciding on proper methods of diagnosis and treatment of suspected childhood sleep apnea.

A team of researchers have undertaken a longitudinal study in which a large number of children scheduled for adenotonsillectomy (AT) and a smaller number of controls (children with no clinical symptoms of SDB scheduled for non-airway procedures) are investigated by collecting PSG, cognitive and behavioral data at two distinct time points, just before surgery and one year later. In this study, patients' bedside diagnoses, pre-surgical PSG results, and parents' answers to the validated Pediatric Sleep Questionnaire were compared.

The authors of "Polysomnography in Children Scheduled for Adenotonsillectomy" are Robert A. Weatherly MD, Deborah L.Ruzicka, R.N PhD, Deanna J. Marriott, and Ronald D. Chervin, MD, all from the University of Michigan Health System, Ann Arbor, MI. Their findings are to be presented at the American Academy of Otolaryngology--Head and Neck Surgery Foundation Annual Meeting and OTO EXPO, September 22-25, 2002, at the San Diego Convention Center, San Diego, CA.

Methodology: Thirty-four children, aged 5.0-12.9 years, scheduled to undergo adenotonsillectomy for some degree of airway obstruction were enrolled in the study. Prior to their adenotonsillectomy, each child underwent a nocturnal polysomnography (sleep study), a Multiple Sleep Latency Test (MSLT), and comprehensive behavioral and cognitive investigations. The polysomnograms were interpreted and scored by a single registered polysomnographic technician unaware of the patient's clinical diagnosis. Parents also completed a Pediatric Sleep Questionnaire (PSQ), a validated 74-item list of questions about their child's sleeping habits and behaviors. The sleep-related breathing disorder "subscale" of the PSQ focuses on 16 questions that are most closely related to the presence or absence of SDB.

For the purposes of analysis, the clinical diagnosis (provided by the child's surgeon) was compared to PSG results as the "gold standard" assessment of SDB. Three different PSG thresholds were compared to each other for each patient; the PSG was considered abnormal if it met criteria based on the following different criteria: Criterion I: The child had more than one obstructive apnea (absent airflow despite respiratory efforts for two breaths or longer) per hour of sleep; Criterion II: The child had more than five apneas or hypopneas (reduction in airflow, chest excursion, or abdominal excursion for two breaths or longer followed by a desaturation of 4 percent or more, an arousal, or an awakening) per hour of sleep; or Criterion III: The child was found to have more than one apnea, hypopnea, or respiratory-related arousal per hour of sleep.

Parental responses to the PSQ were then analyzed to help determine which parent observations could help in the clinical determination that SDB is not present, in those cases where the PSG did not meet Criterion I for being abnormal.

Results: The children who make up this report all had some evidence of nocturnal airway obstruction based on clinical history as determined by an otolaryngologist. The mean age of the participants was 8.1+/-1.8 (S.D.) years; the age at enrollment ranged from 5.3 to 12.5 years. Eighteen patients were girls, 16 boys. The mean number of obstructive apneas and hypopneas per hour of sleep was 3.5+/-8.3 (SD), the mean number of apneas, hypopneas, and respiratory-related arousals was 10+/-13.6 (SD.), and the mean minimum oxygen saturation was 90.9+/- 7.0 (SD) percent. Depending on the PSG criteria used, the presence of PSG-confirmed SDB in the group of children with clinical signs and symptoms of SDB ranged from 53 percent for the first criteria to 88 percent for the third criteria. The percentage was highest when the PSG criteria used were least restrictive, or when more subtle respiratory events were counted.

The data revealed that the clinical diagnosis of SDB and the polysomnographic diagnosis of SDB is quite variable depending on the actual cutoff or threshold used. If the measure consists only of apneas and hypopneas, the agreement rate was 53% for Criterion I and 56% for Criterion II, which compares closely with several existing studies where such agreement ranged from 37 to 56 percent. When the diagnosis of Upper Airway Resistance Syndrome (UARS) was considered in our subjects (Criterion III), the rate of agreement increased to 88%. It has long been felt that UARS is an important consideration in the realm of pediatric SDB, a point that is supported by the researchers' findings.

Conclusions: Agreement in the area of the diagnosis of clinically significant sleep-disordered breathing is lacking, in part because there is no standard to assess the significance of sleep study data. This study highlights the need to develop a generally agreeable interpretation of SDB in children to make treatment decisions regarding groups of children in different categories, to continue to compare the effectiveness of different therapeutic modalities, and to look at long-term outcomes. If this standard is to include polysomnographic data, as recommended by the American Academy of Pediatrics, then the definition of an abnormal PSG will have to be chosen carefully so as to maximize the sensitivity and specificity of such a test, realizing that one cutoff point will probably never serve this purpose.

The researchers suggest that a combination of clinical and laboratory measures will be needed to allow for the identification of different manifestations of the disease in different types of patients. The costs of various tests will also have to be included in developing such recommendations. Sound medical judgement will require such knowledge and analysis.