Does an undetected obstructive sleep apnea influence the natural course and success of cardiac rehabilitation after cardiac surgery?

ABSTRACT

BACKGROUND: Obstructive Sleep Apnea (OSA) is common in patients with cardiovascular diseases (CVD) and can negatively impact the course of CVD. However, scarce data are available for patients before or after cardiac surgery (CS) in the context of OSA. AIM: This study investigated the impact of an undetected OSA on the results of cardiac rehabilitation (CR) in patients after CS. DESIGN: Observational study over a period of 3 months following CS. SETTING: The setting of this study was inpatient CR after CS or cardiac interventions. POPULATION CS cohort referred to a CR program to an inpatient rehabilitation clinic in Switzerland. METHODS: In this prospective observational study 256 patients were screened for OSA at the entry of CR via a level III screening device (ApneaLink AirTM, Resmed, San Diego, CA, USA). They were stratified into two groups apnea hypopnea index (AHI) >15 or <15/h. A comprehensive assessment was performed at entry and end of CR including six-minute walk test (6-MWT), Functional Independence Measure (FIM), Hospital Anxiety and Depression Scale, MacNewHeart (MNH), STOP-Bang Questionnaire (SBQ) and Epworth Sleepiness Scale (ESS).All patients participated in a comprehensive CR program with a mean duration of approximately 3 weeks. Another OSA screening was performed at the end of the PR program and after 3 months in order to observe the clinical course of OSA. RESULTS: An AHI>15/h was found in 133 patients (59%) at baseline, 54% after 3 weeks and 43% after 3 months. The AHI>15/h-group was older, had a higher BMI, more frequent hypertension and coronary artery disease, and higher ESS and SBQ scores compared to the AHI<15/h-group. The results of the STOP-Bang and ESS questionnaires showed a statistically significant but weak positive correlation with AHI. However, in both groups ESS did not improve from baseline to the end of CR. A multivariable logistic regression model confirmed age and ESS as independent positive predictors of OSA. No differences were found between both groups according to the results of the 6-MWT, HADS and MNH. CONCLUSIONS: OSA had a high prevalence in a large CS cohort referred to CR. However, moderate-severe OSA, though symptomatic, had no significant influence on the outcome during CR and on the parameters representing success. CLINICAL REHABILITATION IMPACT In our study the improvements during CR after CS were not influenced by the presence of significant OSA which is remarkable since the presence of OSA is thought to be associated with increased rates of cardiovascular adverse events after cardiovascular intervention or CS. This is true at least for the success of CR after CS. The long-term consequences of untreated OSA in CVD remain unclear and are still the subject of current research.