LAUSR

data from other reports, caution should be paid to the category of OSA severity. Second, the authors clarified that male gender, greater age and a higher BMI all significantly contributed to differentiate “treat” group of OSA. The authors also recognized that these three variables were included in items of Stop-Bang Scale as binary information and I feel that statistical model of simultaneous inclusion of BMI, age, gender and Stop-Bang Scale score as independent variables should be paid with caution to avoid multicollinearity. Finally, the authors recognized that almost a quarter of Stop-Bang Scale scores were missing in their study. I also reported on this point that two items out of eight in StopBang Scale were subjective response during sleep, and more than one-third of missing values were recognized [2]. Although I agree that there is an advantage of Stop-Bang Scale scores against Epworth Sleepiness Scale score for screening OSA [3], validation study for predicting the severity of OSA should be confirmed according to the standard classification in the severity of OSA. In addition, a metaanalysis is also recommended to know the effectiveness of Stop-Bang Scale score for predicting the severity of OSA [4, 5].