Background: Continuous positive airway treatment (CPAP) is the recommended treatment for patients with obstructive sleep apnea (OSA). Outcome measures often focus on clinical and/or self-rated variables related to the medical condition. However, a brief validated instrument focusing on the whole life situation (i.e., ethos) suitable for clinical practice is missing. The aim of this study was to investigate factorial structure, categorical functioning of the response scale, and differential item functioning across sub-populations of the Ethos Brief Index (EBI) among patients with obstructive sleep apnea (OSA) before and after initiation of continuous positive airway pressure (CPAP).

Methods: A prospective design, including 193 patients with OSA (68% men, 59.66 years, SD 11.51) from two CPAP clinics, was used. Clinical assessment and overnight respiratory polygraphy were used to diagnose patients. Questionnaires administered before and after 6 months of CPAP treatment included EBI, Epworth Sleepiness Scale (ESS), Hospital Anxiety and Depression Scale, and global perceived health (initial item in SF-36). The validity and reliability of the EBI were investigated using Rasch and confirmatory factor analysis models. Measurement invariance, unidimensionality, and differential item functioning across gender groups, Apnea-Hypopnea Index, and ESS groups were assessed.

Results: The reliability of the EBI was confirmed using composite reliability and Cronbach’s alpha. The results supported unidimensionality of the EBI in confirmatory factor analysis and the Rasch model. No differential item functioning was found. A latent profile analysis yielded two profiles of patients with low (n = 42) and high (n = 151) ethos. Patients in the low ethos group were younger and had higher depression scores, lower perceived health, and higher body mass index.

Conclusions: The EBI is a valid tool with robust psychometric properties suitable for use among patients with OSA before and after treatment with CPAP is initiated. Future studies should focus on its predictive validity.