Measurement of sleep-related breathing disturbances in epidemiologic studies. Assessment of the validity and reproducibility of a portable monitoring device

The feasibility and reliability of measuring sleep-related breathing disorders with a portable monitor (PM) were assessed in a heterogeneous population, consisting of 31 patients recruited from a sleep laboratory and pulmonary disease clinic, 16 participants in a genetic-epidemiologic study of sleep apnea, and four volunteers with no specific sleep complaints. The validity of measurements made by the PM was assessed with comparisons of respiratory parameters made with the PM to those determined with in-hospital polysomnography (PSG) (25 studies). Reproducibility was assessed in 29 subjects who underwent in-home monitoring on two occasions. There was a high level of agreement between the number of respiratory events (apneas or hypopneas) per hour of estimated sleep (respiratory disturbance index, RDI) recorded with the PM and PSG and log-transformed (r = 0.96). Using a RDI of greater than or equal to 10 to define "abnormality," 20 of the 21 subjects who would have been classified as abnormal with PSG were classified similarly with use of the PM. A similar high level of agreement was demonstrated for the log-transformed RDI determined with replicate in-home studies (r = 0.94). No evidence of a "first-night effect" for the RDI was suggested in studies performed with the PM; ie, RDI was 18.4 +/- 27.7 and 17.4 +/- 25.7 (mean +/- SD) for first and second night studies, respectively (p = 0.21). A second compared with an initial study with the PM would have resulted in reclassification of abnormality based on an RDI of greater than or equal to 10 in one subject. These findings suggest that measurement of the RDI with in-home monitoring provides a valid and highly reproducible index for assessment of sleep-related respiratory disturbances for use in epidemiologic studies of general populations.