SNORES Classification: Multicenter Prospective Validation for Obstructive Sleep Apnea.

Abstract

To validate a clinically feasible, multidimensional classification system-SNORES (Severity of apnea-hypopnea index, Nasal airflow limitation, Oxygen desaturation burden, Respiratory effort, Awake Endoscopic anatomy, and Sleep position dependency)-for obstructive sleep apnea (OSA), and to evaluate its ability to predict therapy escalation, CPAP adherence, and surgical intervention beyond the apnea-hypopnea index (AHI) alone. In this prospective multicenter cohort study, 1560 adults with newly diagnosed OSA confirmed by in-laboratory polysomnography were enrolled. Each SNORES domain was scored from 0 to 3 using predefined objective criteria derived from routine diagnostic testing. Primary outcomes were therapy escalation, CPAP adherence, and surgical intervention. Predictive performance was compared with AHI alone. Higher SNORES scores were independently associated with therapy escalation (hazard ratio [HR] 3.10, 95% CI 2.62-3.67; p < 0.001) and surgical intervention (HR 2.71, 95% CI 2.21-3.33; p < 0.001). The SNORES framework demonstrated superior discrimination for therapy escalation (area under the curve [AUC] 0.82, 95% CI 0.78-0.86) compared with AHI alone (AUC 0.52, 95% CI 0.46-0.57; ΔAUC +0.30, p < 0.001). Removal of individual domains-particularly oxygen desaturation or respiratory effort-meaningfully attenuated predictive performance (ΔAUC -0.09 to -0.11). SNORES is a feasible, reproducible, multidimensional framework that captures clinically relevant heterogeneity in OSA and guides individualized therapy more effectively than traditional AHI-only models.