Can the analysis of built-in software of CPAP devices replace polygraphy in children?

Polysomnography (PSG) is the gold standard for the scoring of residual respiratory events during continuous positive airway pressure (CPAP). Studies comparing PSG scoring with automatic scoring by the built-in software of CPAP devices have reported acceptable agreements except for the hypopnea index (HI) in adult patients, but no study has yet been conducted in children. The aim of the present study was to compare the automatic scoring by CPAP device and manual scoring using the software tracings of the CPAP device integrating pulse oximetry (SpO₂) with in-lab polygraphy (PG).

Consecutive clinically stable children treated with constant CPAP (ResMed) for at least one month and scheduled for a nocturnal PG were recruited. A pulse oximeter was connected to the CPAP device. The PG apnea-hypopnea index (AHI_PG), scored according to modified AASM guidelines, was compared with the automatic AHI reported by the CPAP device (AHI_{A CPAP}) and the manual scoring of the AHI on the CPAP software (AHI_{M CPAP}).

Fifteen children (1.5-18.6 years) were included. Mean residual AHI_PG was 0.9 ± 1.2/hour (0.0-4.6/hour) vs. AHI_{A CPAP} of 3.6 ± 3.6/hour (0.5-14.7/hour) (p < 0.001), and AHI_{M CPAP} of 1.2 ± 1.6/hour (0.0-5.1/hour) (p = 0.01). Correlation between AHI_PG and AHI_{A CPAP} was good (r = 0.667; p = 0.007), and improved when considering AHI_{M CPAP} (r = 0.933; p < 0.001). Strong correlations were also observed between the PG apnea index (AI) and HI, and the manually scored AI and HI on CPAP, respectively.

Manual scoring of respiratory events on the built-in software tracings of CPAP devices integrating SpO₂ signal may be helpful. These results have to be confirmed in patients with higher AHI.