0602 Determinants and Clinical Consequences of Treatment Emergent Central Sleep Apnea

Abstract

Introduction: Treatment for Obstructive Sleep Apnea (OSA) with continuous positive airway pressure (CPAP) may precipitate central sleep apnea (CSA), also known as treatment emergent CSA (TECSA). However, the factors contributing to TECSA are not clear. The purpose of this study was to examine the key determinants and impact of Treatment Emergent CSA. Methods: The study was a retrospective analysis of consecutive Veteran patients (n=463), age >18 year, who were evaluated for sleep disordered breathing at the Detroit VA Medical Sleep Disorders Center in 2011, with either a baseline polysomnogram followed by PAP titration or with a split-night polysomnogram (n=304). TECSA was defined as the presence OSA (AHI>5/hour) without CSA at baseline, but with superimposed CSA (CAI≥5/hour) during PAP titration. Control group patients had OSA alone without TECSA or OSA with CSA on the diagnostic portion of the study. Comorbidities such as heart failure (CHF), coronary artery disease, chronic obstructive pulmonary disease and atrial fibrillation were recorded as well as the use of prescription narcotics and sedatives. Descriptive analyses and multiple logistic regression analyses were performed; presence/absence of TECSA was the dependent variable, whereas demographics, split-night/full-night PAP study, AHI, nadir oxygen saturation, comorbid factors and drug use or morphine equivalent dose were the independent variables. Results: Ninety-six percent of Veterans were men (n=443). Mean standard deviation for key variables were: BMI 33.7 ± 6.7 kg/m2, baseline AHI 47.7 ± 34.6/hour, baseline nadir oxygen saturation 83.5 ± 7.5%, CAI on PAP 5.1 ± 10.7/hour OSA. TECSA was demonstrated in 131 out of 463 patients (28.3%). Multiple regression models demonstrated that only AHI severity (odds ratio, OR 1.01; 95% C.I. 1.007 to 1.021) and and the presence of CHF (OR 2.94; 95% C.I. 1.19 to 7.31, p=0.02) were significant predictors for the presence of TECSA, after adjusting for the occurrence of other comorbidities and prescription opioid/sedative use or the morphine equivalent dose. Conclusion: The severity of OSA at baseline and presence of CHF predicted TECSA, independent of other covariates. The impact of TECSA on long-term clinical outcomes needs to be studied further.