Video and polysomnography
All children underwent overnight PSG in a sleep unit. Heart rate, electrocardiographic signals, electroencephalographic (EEG) and electrooculographic activities, thoracic and abdominal respiratory movements, respiratory rate, arterial oxygen saturation (SaO2) and airflow were continuously measured throughout the sleep period. Along with the PSG recording, a video was recorded with a Sony® Ipela SNC-ER550 rotating camera coupled to an infrared device to allow recording in low light conditions without disturbing the child’s sleep. The video was synchronized to PSG. The camera was placed on the wall opposite the child’s bed, at 2.5 meters and positioned to focus on the child’s head and chest. Sheets and blankets were avoided as often as possible, and if needed for the child to fall asleep, removed as soon as he slept. A sound recording, assessed in decibels, was also made at the same time.
Interpretation of the PSG data, blinded to the clinical examination and SHS, was carried out according to the AASM criteria of the AASM [12, 13]. Patients were classified in four OSAHS severity groups: absence (OAHI < 1.5 per hour), mild (OAHI 1.5-5), moderate (OAHI 5-10) and severe (OAHI ≥ 10). The ODI was calculated as the number of decreases in SaO2 ≥ 3% [20].
The PSG videos were evaluated throughout the second sleep cycle identified by EEG signals as the period between the end of the first rapid eye movement (REM) sleep and the end of the second one. The analysis was blinded to OAHI, ODI and SHS values. The videos were interpreted by an otorhinolaryngologist investigator according to a slightly modified version of Sivan’s scoring system. The following parameters were considered: presence of inspiratory noise, type of inspiratory noise, movements during sleep, number of waking episodes, number of apneas, chest retractions, and mouth breathing (Table 1). The movements of more than four body parts (four limbs, head, and chest) were considered as a whole-body movement. Any opening of the eyes or a shift to a sitting position were counted as a waking episode. All events characterized by breathing interruption with persistent respiratory effort for more than two respiratory cycles were counted as apneas. Diaphragmatic paradoxical movements were analyzed together with the chest retractions. For the third item (movements during sleep), we set the threshold for numerous movements at 2 rather than 3 in Sivan’s original scoring system. Likewise, for the fifth item (number of apneas), we set the threshold of numerous apneas at 2 rather than 3.
The score obtained for the second sleep cycle of each child was then brought back first to a 30-minute duration, leading to a 30-minute risk score (RS30), then to a 10-minute duration, leading to a 10-minute risk score (RS10). The number of apneas, waking episodes and movements was averaged for a 30-minute period using the following formula: (number of apneas + number of waking episodes + number of movements) over 30 minutes = (number of apneas + number of waking episodes + number of movements) over the entire second sleep cycle*30/ length of the second sleep cycle in minutes. The same calculations were performed for a 10-minute period. For the other variables (inspiratory noise, intercostal retractions, mouth respiration), only their intermittent or continuous nature was noted, hence their value was unchanged irrespective of the duration.