This retrospective observational study aims to characterize the prevalence and clinical features of obstructive sleep apnea (OSA) phenotypes and develop a rater-independent algorithm for automated OSA phenotyping, improving diagnosis and personalized treatment.
Obstructive sleep apnea (OSA) is characterized by repeated upper airway blockages during sleep, but it presents with a range of phenotypic variations, each with potentially distinct clinical implications. Current clinical definitions are not always precise, making it difficult to clearly classify patients with overlapping features. This phenotypic overlap poses challenges for understanding the true prevalence of "pure" versus "mixed" OSA phenotypes and their respective clinical implications. To comprehensively characterize the prevalence and clinical features of distinct OSA phenotypes in a large, diverse patient population, this retrospective study analyzes polysomnography (PSG) data from two publicly available National Sleep Research Resource datasets. The findings are then compared to datasets from the University Hospital Basel and University Children's Hospital Basel to assess generalizability. Furthermore, the study employs computer-aided analysis of PSG data to develop rater-independent algorithms for objective and automated OSA phenotyping. These advancements aim to improve understanding of OSA heterogeneity, facilitating more precise diagnoses and personalized treatment strategies.
Study Type
OBSERVATIONAL
Enrollment
1,055
University Children's Hospital Basel
Basel, Canton of Basel-City, Switzerland
University Hospital Basel
Basel, Canton of Basel-City, Switzerland
Phenotype characterization
To characterize the prevalence and clinical features of distinct obstructive sleep apnea (OSA) phenotypes in a large, diverse patient population, polysomnography (PSG) recordings from two publicly available National Sleep Research Resource data collections are analyzed and compared to datasets from the University Hospital Basel (USB) and University Children's Hospital Basel (UKBB) to assess generalizability.
Time frame: 2025
Computer-aided phenotyping
To evaluate the potential of computer-aided pattern recognition in improving the accuracy and efficiency of OSA phenotype identification from PSG data, an algorithm is developed that objectively and automatically identifies OSA phenotypes from PSG recordings.
Time frame: 2025
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.