Abstract
To evaluate the efficiency of 5 screening questionnaires for obstructive sleep apnea (OSA), OSA frequency, and the association between OSA and COVID-19 severity in recent COVID-19 cases, and to compare the use of the oxygen desaturation index (ODI) as an alternative measure for the respiratory disturbance index (RDI). This open cohort study recruited patients with recent COVID-19 (within 30-180 days) diagnosed using reverse transcription polymerase chain reaction. Participants were screened for OSA using the following 5 sleep disorder questionnaires prior to undergoing type I polysomnography: the Sleep Apnea Clinical Score (SACS), Epworth Sleepiness Scale (ESS), STOP-Bang score, No-Apnea score, and Berlin questionnaire. Polysomnography revealed that 77.5% of the participants had OSA and that these patients exhibited higher COVID-19-related hospitalization (58%) than those exhibited by non-apneic patients. The Kappa coefficient showed reasonable agreement between RDI > 5/h and No-Apnea score, RDI > 15/h and Berlin questionnaire score, and Epworth Sleepiness Scale and STOP-Bang score, but only moderate agreement between RDI > 15/h and No-Apnea score. An OSA-positive No-Apnea score increased the specificity of the SACS to 100% when RDI > 5/h. The intraclass correlation coefficient showed 95.2% agreement between RDI > 5/h and ODI > 10/h. The sequential application of the No-Apnea score and SACS was the most efficient screening method for OSA, which had a moderately high incidence among the post-COVID-19 group. We demonstrated an association between OSA and COVID-19 related hospitalization and that ODI could be a simple method with good performance for diagnosing OSA in this population.