Prevalence and predictive factors of sleep breathing disorders (SBD) in severe COPD patients

Background : Severe COPD opposes challenges to the clinician for appropriate and rational prescription of ventilatory support at home. Ventilatory and blood gases (ABGs) abnormalities may worsen during sleep while remaining "esoteric" to the clinician. Aims and objectives : Evaluate the prevalence and typology of SBD in severe COPD patients. Identify predictive factors of SBD in severe COPD patients in order to design a predictive score. Methods : Prospective study. Consecutive patients admitted to ICU or the pulmonology department within a period of 9 months. Clinical and functional data were collected. Epworth and Pittsburgh scales were measured. Respiratory polygraphy was performed at steady state. Univariate then multivariate analyses were performed to identify factors independently associated with SBD in severe COPD patients. Results : 51 patients were included. They were 62.3±9.6yr aged ; BMI, 33.8±9.9Kg/m 2 ; mMRCII, 52.9% and pCO 2 , 43.6±6.5mmHg. The prevalence of SBD was 49% (64% overlap syndrome, 36% nocturnal hypoventilation-related desaturation). 18(72) patients demonstrated SBD while their diurnal pCO2 remained below 55mmHg. Multivariate analysis identified snoring, nocturia, mood disorders and excessive daytime sleepiness as only factors independently associated with SBD. The combination of these four symptoms in a so called SBD-C score demonstrated very interesting discriminative properties (AUC=0.92). Conclusion : This study identified a high prevalence of SBD while ABGs abnormalities remains below thresholds of prescription of home ventilatory support. A predictive novel score, the SBD-C may be a useful tool in an integrative diagnostic strategy of SBD in severe COPD patients.