Multinight Prevalence, Variability, and Diagnostic Misclassification of Obstructive Sleep Apnea

Area of Science:

• Sleep Medicine
• Cardiorespiratory Physiology
• Biomedical Engineering

Background:

• Obstructive sleep apnea (OSA) severity exhibits significant night-to-night variability.
• This variability has critical implications for accurate diagnosis and effective management of OSA.
• Previous diagnostic approaches often rely on single-night polysomnography, potentially missing crucial variations.

Purpose of the Study:

• To determine the global prevalence of OSA using multinight in-home sleep recordings.
• To evaluate the impact of night-to-night OSA severity fluctuations on diagnostic accuracy.
• To assess the reliability of single-night apnea-hypopnea index (AHI) values for OSA classification.

Main Methods:

• Analysis of over 11.6 million nights of data from 67,278 individuals (aged 18-90).
• Utilized a validated, novel under-mattress sleep analyzer for in-home nightly monitoring.
• Defined OSA as a nightly mean apnea-hypopnea index (AHI) > 15 events/h.

Main Results:

• Global OSA prevalence was determined to be 22.6% (95% CI, 20.9-24.3%).
• Single-night AHI testing resulted in misdiagnosis likelihoods ranging from 20% to 50% for individuals with OSA.
• Misdiagnosis rates decreased with extended monitoring, achieving high accuracy (F1-score 0.94) after 14 nights.

Conclusions:

• Multinight in-home monitoring reveals a global OSA prevalence of approximately 20%.
• Single-night diagnostic studies may misclassify up to 20% of individuals with OSA.
• Findings underscore the necessity of considering night-to-night variability in OSA diagnosis and treatment strategies.