Sleep apnea diagnosis in children using software-generated apnea-hypopnea index (AHI) derived from data recorded with a single photoplethysmogram sensor (PPG) Results from the Childhood Adenotonsillectomy Study (CHAT) based on cardiopulmonary coupling analysis

Objective Sleep quality is vital for healthy development in children. Sleep disorders are prevalent and negatively affect sleep quality. Early identification and appropriate intervention can improve children's health and quality of life. The current reference standard, polysomnography (PSG) has limitations regarding availability, cost, and access and may not replicate normal sleep patterns in the home. Simple, accurate sleep tests, available for repeated testing should be beneficial in management of sleep disorders. Method Secondary analysis of PSG data from the prospective multicenter Childhood Adenotonsillectomy Trial (CHAT) to evaluate FDA-cleared cloud-based software (Software-as-a-Medical-Device), which is based on analysis of photoplethysmogram data (PPG; plethysmogram-signal (PLETH) and oxygen saturation data (SpO(2))), to automatically generate a novel apnea-hypopnea index (sAHI). sAHI is compared to manually scored AHI from PSG. Results Significant correlation is observed comparing the software-generated sAHI and manually derived AHI from the in-laboratory PSG-studies (Pearson correlation = 0.954, p < 0.0001) and receiver operating characteristics (ROC) demonstrate strong agreement in all OSA categories (mild, moderate, severe) 91.4%[CI(95%)89.5, 93.4]; 96.7%[CI(95%)95.4, 97.9]; 98.6%[CI(95%)97.8, 99.4], sensitivities 95.4%[CI(95%)93.2, 97.0]; 86.5%[CI(95%)80.3, 91.3]; 88.4%[CI(95%)78.4, 94.9] and specificities 84.4%[CI(95%)79.7, 88.4]; 99.2%[CI(95%)98.2, 99.7]; 99.6%[CI(95%)98.8, 99.9], respectively. Conclusion sAHI is comparable to manual scoring of AHI from in-laboratory PSG studies and effective to rule-in and rule-out obstructive sleep apnea (OSA) in all disease categories, providing safe and convenient approach for diagnosis and management of OSA in children. The data is recorded with a single-sensor, making the method suitable for multi-night testing in the child's home at considerably lower cost. This technology provides a simple tool to adhere to guidelines for diagnosis and management of OSA in children.