São João, RicardoCardoso, AndreiaDomingues, Tiago DiasFradinho, MartaSilva, VâniaFeliciano, Amélia2023-01-102023-01-102022São João, R., Cardoso, A., Domingues, T. D., Fradinho, M., Silva, V. & Feliciano, A. (2022). A retrospective study on obstructive sleep apnea. In R. Bispo, L. Henriques-Rodrigues, R. Alpizar-Jara & M. Carvalho (Eds), Recent developments in statistics and data science: Vol. 398. SPE 2021. Springer Proceedings in Mathematics & Statistics (pp.281-292). Springer. doi:10.1007/978-3-031-12766-3_19978-3-031-12765-6978-3-031-12766-3http://hdl.handle.net/10400.15/4243Obstructive sleep apnea (OSA) is a sleep-related breathing disorder with worldwide increasing prevalence. Polysomnography is the traditional gold standard for the diagnosis of OSA, but the fact that it is a complex, time-consuming, and expensive test contributes to the underdiagnosis of this pathology. For this reason, one usually opts for the simpler, less labor-intensive, and cheaper cardiorespiratory sleep test for the diagnosis of this syndrome. The manual analysis of these tests, which usually involves two or more qualified observers, is one of the aspects that most contributes to the amount of time spent in the analysis and, consequently, to diagnostic delay. Automatic analysis emerges as a faster alternative to the manual analysis. Based on a sample of 2559 patients monitored by the Pulmonology Department—Sleep Unit of the Hospital da Luz Setúbal during the period 2011–2019, this research concludes that there is no agreement between the manual and automatic readings of two popular OSA classification indexes.engAutomatic readingAssociation measuresConcordance measuresManual readingOSAbook part10.1007/978-3-031-12766-3_19