Comparison of diagnostic reliability of out-of-center sleep tests for obstructive sleep apnea between adults and children

ObjectivesSleep studies for diagnosing obstructive sleep apnea (OSA) in children are laborious, expensive, inconvenient, and often not readily available. Out-of-center sleep test (OCST) devices have been studied for diagnosing OSA in adults, but few OCST studies have been done in children. The purpose of this study was to clarify the diagnostic reliability of OCST devices for children.MethodsOCSTs using pulse oximetry and in-laboratory polysomnography (PSG) were performed separately in 686 adults and 119 children. For each apnea-hypopnea index (AHI) measured with PSG, accuracy, sensitivity, specificity, positive/negative likelihood ratio (PLR/NLR), and positive/negative predictive value (PPV/NPV) were calculated for several cutoff values of 3% oxygen desaturation index (ODI) measured with OCST and analyzed.ResultsFor definitive diagnosis in adults, the specificity, PLR, and PPV with a cutoff value of OCST-ODI 20/h were 98.3%, 29.26, and 97.4%, respectively, to detect PSG-AHI ≥20/h. Corresponding values with a cutoff value of OCST-ODI 15/h were 99%, 46.19, and 99.6% to detect an AHI ≥5/h. For exclusive diagnosis (screening) in adults, sensitivity, NLR, and NPV with a cutoff value of OCST-ODI 5/h were 96.4%, 0.068, and 91.9% to detect PSG-AHI <20/h and 84.1%, 0.21, and 45.9% to detect PSG-AHI <5/h. or definitive diagnosis in children, the corresponding values with a cutoff value of OCST-ODI 25/h were 98.6%, 16.0, and 90.9% to detect PSG-AHI ≥10/h and 98.1%, 8.281, and 90.9% for PSG-AHI ≥5/h. For exclusive diagnosis in children, with a cutoff of OCST-ODI 10/h, the corresponding values were 62.2%, 0.446, and 78.2% to detect PSG-AHI <10/h, 45.3%, 0.674, and 55.1% for PSG-AHI <5/h, and 34.0%, 0.908, and 10.3% for PSG-AHI <1/h. Statistical data of preschool children tended to be worse than those of school age children.ConclusionsIn adults, OCST is reliable for the definitive diagnosis of AHI ≥20/h or ≥5/h and the exclusive diagnosis of AHI <20/h. However, in children, OCST should not be used alone for the definitive diagnosis or exclusive diagnosis.