0576 Validation of the Contact-free Sleep Monitoring Device for Diagnosis of Obstructive Sleep Apnea in Chinese Adults. (12th April 2019)

• Record Type: Journal Article
• Authors: Dong, Xiaosong; Zhi, Hui; Chen, Jianan; Han, Fang; Zhao, Rui
• Is Part Of: Sleep. Volume 42(2019)Supplement 1
• Page Start: A229
• Page End: A229
• Publication Date: 2019-04-12
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsz067.574
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Polysomnography (PSG) is the gold standard for sleep monitoring, but has several disadvantages, including intensive resource consumption, cost, and limited accuracy of interrater interpretation. Most portable home sleep monitoring devices require body contact. One of the newest devices, a contact-free device, is developed to detect sleep stage as well as respiratory events. Methods: 244 Chinese adult participations (mean ± standard deviation age 47.30 ± 13.77 years, 73.0% males, body mass index 26.53 ± 4.21 kg/m 2 ) who presented with snoring were recruited between May 2017 and November 2017. They underwent an overnight monitoring using the Contact-free sleep monitoring (CFMS) device as well as polysomnography (PSG) in the sleep lab. The CFSM recordings were analyzed automatically, and PSG was manually scored based on AASM scoring manual. Then we evaluated the agreement of CFSM and calculated the optimum diagnostic efficiency of the CFSM. Results: (1) Total sleep time of the CFSM was longer than PSG significantly (447.80 ± 78.96 min and 399.25 ± 73.06 min respectively, p<0.0001), the apnea-hypopnea index (AHI) was 27.40 ± 21.40 events/h on the CFSM and 29.47 ± 26.38 events/h on PSG (p=0.0011); (2) There was a significant correlation between the AHI of those two tests (Pearson's coefficient=0.94, p<0.0001). Bland-Altman analysis of AHI on PSG versus CFSM showed a mean difference of 2.1 events/h, limits of agreement (equal to ± 2 standard deviations) was -17.1 to 21.3 events/h. (3) Based on the golden standard that AHI ≥ 5 events/h on PSG, the optimum diagnostic cut-off value of the CFSM was AHI > 10.6 events/h with 85.4% sensitivity, 100.0% specificity, 100.0% positive predictive value and 50.8% negative predictive value compared to PSG. Conclusion: When controlling for night-to-night variability and changes of sleeping environment, the results confirmed close agreement between the CFSM and PSG. If the clinical diagnosis allows the measurement error of AHI is no less than 2.1 events/h, the Contact-free Sleep Monitoring Device could be an alternative tests used for diagnosing obstructive sleep apnea in Chinese adults. Support (If Any)