Respiratory polygraphy in children: Feasibility in everyday practice in an ENT department and value of automatic detection of respiratory events.

OBJECTIVES: Using respiratory polygraphy (RP) in children for diagnosis of obstructive sleep apnea/hypopnea syndrome (OSAHS) can be challenging in terms of device acceptance and sensor displacement. Automatic analysis of respiratory events has never been evaluated in a pediatric population. The primary objective of this study was to determine the feasibility of pediatric RP in routine ENT department practice. The secondary objective was to evaluate the reliability of the automatic detection of obstructive and central respiratory events in children. METHODS: A single-center retrospective study included 50 children (32 boys, 18 girls; mean age 5.5±2.3years) undergoing overnight RP in an ENT department between January and August 2016 for suspected OSAHS. Manual detection of respiratory events was performed by one ENT specialist experienced in RP interpretation, and compared to automatic analysis. RESULTS: The device was well accepted in 98% of cases. Overall signal quality was>50% in 76% of cases, with average signal quality of 70.8% (86% in patients>3 yrs, 25% in patients<3 yrs, P=0.0013). There was no significant correlation between manual and automatic analyses, except for central apnea (Spearman coefficient 0.43; P=0.0015). One hundred percent of patients presented OSAHS according to automatic detection, compared to 32% according to manual detection (P<0.005). CONCLUSIONS: Pediatric RP is feasible in routine practice in an ENT department, with good acceptance and satisfactory signal quality in children older than 3years. Automatic analysis of respiratory events in children is unreliable, except for central apnea.