Accuracy and acceptability of home sleep apnea testing with electroencephalography compared to in-lab polysomnography for the diagnosis of obstructive sleep apnea in children.
Darko Stefanovski, Mahalakshmi Somayaji, Michelle Ward, Jennifer Falvo, Mary Anne Cornaglia, Ignacio E Tapia, Yaelis Roman, Melissa Xanthopoulos, Christopher M Cielo
Author Information
Darko Stefanovski: School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA.
Mahalakshmi Somayaji: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
Michelle Ward: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
Jennifer Falvo: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
Mary Anne Cornaglia: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
Ignacio E Tapia: Division of Pediatric Pulmonary Medicine, University of Miami, Miami, FL.
Yaelis Roman: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
Melissa Xanthopoulos: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
Christopher M Cielo: Division of Pulmonary and Sleep Medicine, Children's Hospital of Philadelphia, Philadelphia, PA.
STUDY OBJECTIVES: In-laboratory polysomnography is recommended for the evaluation of obstructive sleep apnea (OSA) in children, but availability is limited. We hypothesized that home sleep apnea testing including electroencephalogram (HSAT) could accurately detect OSA in children and be an alternative to polysomnography. METHODS: Children clinically referred for polysomnography underwent testing with the HSAT device twice: once in their home as well as concurrently with in-lab polysomnography (portable lab testing [PLT]). HSAT and PLT were compared to reference polysomnography for OSA diagnosis using an obstructive apnea-hypopnea index (OAHI) greater than 2 events/hour. OAHI cutoffs of 1 and 5 events/hour were explored. The diagnostic accuracy was further analyzed using the area under the receiver operating characteristics curve (AUC ROC). RESULTS: All 15 participants, median (range) age 8.1 (2.9-17.1) years completed in-lab testing and HSAT. OSA was identified by polysomnography in 7 (47%) participants. HSAT correctly identified OSA status from polysomnography in 14 (93%). OAHI was similar between polysomnography [1.7 (0, 26) events/hour] and PLT [1.6 (0.3, 24.4)] and HSAT [1.8 (0.3, 23)], p=0.98. HSAT OAHI showed strong correlation with polysomnography OAHI (Spearman's r = 0.8, p=0.0001). AUC ROC referenced with polysomnography were excellent using OAHI threshold values of 1,2, and 5 compared with PLT (AUC ROC=0.96, 0.96, 1) and HSAT (AUC ROC= 0.79, 0.95, 0.98 respectively). CONCLUSIONS: HSAT with EEG was accurate compared to polysomnography for the diagnosis of pediatric OSA. Electroencephalography may improve the diagnostic accuracy of HSAT in children, particularly for mild OSA and younger children.
