Are the Epworth Sleepiness Scale and Stop-Bang model effective at predicting the severity of obstructive sleep apnoea (OSA); in particular OSA requiring treatment?

摘要

Abstract Obstructive sleep apnoea (OSA) is a condition characterised by repetitive upper airway collapse during sleep. The condition carries a range of health sequelae that can prove fatal in cases with co-existing risk factors for the condition, such as obesity and hypertension. Utilisation of a high-performance screening tool for OSA is thus important. A retrospective audit using the ESS and Stop-Bang scores, alongside Apnoea–Hypopnea Index values, for patients who underwent polysomnography over 1?year. Multinomial logistic regression was used to compare the predictive abilities of ESS, SBM, and body mass index (BMI) for the patient outcome groups, “None” (No OSA), “Notreat” (OSA not requiring treatment) and “treat” (OSA requiring treatment). The influences of age, gender and BMI on outcome group were also assessed. 126 bariatric and 66 non-bariatric patients were included. Multinomial logistic regression failed to demonstrate predictive ability of ESS. A higher Stop-Bang score significantly increases the risk being in the “treat” group. In addition, male gender, greater age and a higher BMI each individually increase the risk of OSA requiring treatment. Stop-Bang failed to demonstrate predictive significance when age and gender were controlled for. ESS is not an appropriate screening tool for OSA. Stop-Bang, however, remains a useful screening tool, with the ability to detect patient with OSA in need of treatment. Further study may benefit the development and implementation of a concise and more specific screening tool that considers high evidence-based risk factors for OSA, including male gender, greater age and raised BMI.