Utility of Level 3 sleep studies in diagnosing OSA in a high risk population with limited resource settings

Abstract Background Obstructive sleep apnea (OSA) is a common and underdiagnosed disorder, particularly in low- and middle-income countries where access to laboratory polysomnography (PSG) remains limited. Portable Level 3 sleep studies provide a pragmatic alternative for diagnosis in resource-constrained settings. Objectives To describe the diagnostic yield and severity distribution of OSA detected using Level 3 sleep studies in a tertiary-care referral population and to evaluate associations between apnea–hypopnea index (AHI), oxygenation parameters, respiratory indices, and anthropometric variables. Methods This retrospective observational study included consecutive Level-3 sleep studies performed between January 2025 and December 2025 in a tertiary-care referral population. Data were analysed using Spearman correlation, multivariable linear regression, and independent samples t-tests. Results Of 377 studies, 359 were eligible for analysis. OSA was diagnosed in 83.3% of patients (95% CI: 79.2–86.8), with moderate-to-severe disease present in 53.5% (95% CI: 48.3–58.6). AHI showed strong correlation with RDI (ρ = 0.984) and ODI (ρ = 0.863), and a moderate inverse correlation with minimum SpO₂ (ρ = − 0.509) (all p < 0.001). In multivariable linear regression ODI was the strongest independent predictor of AHI (β = 0.786, p < 0.001), while BMI and minimum SpO₂ were not independently associated. Males showed higher OSA prevalence and mean AHI values compared to females (85.4% vs. 75.9%, χ 2 = 4.11, p = 0.043) in our cohort. Conclusion In this tertiary referral cohort, Level-3 sleep studies demonstrated a high diagnostic yield of clinically significant OSA. These findings support Level-3 testing as a scalable diagnostic strategy for high-risk populations in resource-limited settings.