Detecting OSA early: the role of tiredness in screened patients

Detecting OSA early might contribute to a decreased public health economic burden. Especially when limited resources are available, low specificity might lead to unnecessary full- night recordings in healthy patients. The current study aimed to investigate predictive contribution of each item of the STOP-BANG questionnaire and pulse oximetry in OSA. Additionally, the role of older age was assessed. The study included 3128 participants admitted to Split Sleep Medicine Center, referred for polygraphy or polysomnography and assessed with STOP-BANG and Pittsburgh Sleep Quality Index (PSQI). Full-night polysomnography was conducted in-laboratory and full-night polygraphy was performed unattended. Following sleep assessment, 467 participants had no OSA and 2661 had mild to severe OSA. When only STOP-BANG variables were included in a model predicting confirmed OSA diagnosis, the largest amount of variability was explained in respondents 65 year old respondents, lower predictive value was achieved in both models (R2=19, 9% and R2=23, 5% ; p