Predictive value of sleep apnea screenings in cardiac surgery patients

Abstract

Introduction: Obstructive sleep apnea (OSA) is a highly prevalent disorder associated with increased cardiovascular risks. We explored the predictive value of OSA screening instruments in cardiac disease patients awaiting cardiac surgery. Methods: In this prospective cohort, 107 participants awaiting cardiac surgery from Cleveland Clinic and Johns Hopkins underwent polysomnography after completing Epworth Sleepiness Scale (ESS), Sleep Apnea/Sleep Disorder Questionnaire (SA/SDQ), STOP, STOPBAG2 and Berlin questionnaires. Score comparisons between groups based on apnea-hypopnea index (AHI) ≥15 were performed. Logistic regression with receiver operating characteristic (ROC) analysis was used to investigate optimal threshold. Results: Prevalence of OSA (AHI ≥5) was 71.9% (77/107) and 51 (47.7%) had moderate-to-severe disease (AHI ≥15). Participants were primarily male (57%) and Caucasian (76.6%). Mean age was 67.3 ± 13.3 years and BMI was 26.5 ± 6.6. Of the five screening tools, STOPBAG2 with a cut-point of 0.381 provided 78% sensitivity and 38% specificity (AUC 0.66, 95%CI 0.55–0.77). SA/SDQ yielded a cut-point of 32 for all subjects (AUC: 0.62, 95%CI 0.51–0.73) with sensitivity and specificity of 60% and 62% respectively, while STOP score ≥2 provided sensitivity and specificity of 67% and 52% respectively (AUC: 0.61, 95%CI 0.51–0.72). Among STOP items, “observed apnea” had the strongest correlation with AHI ≥15 (OR 3.67, 95%CI 1.57–8.54, p = 0.003). The ESS and Berlin were not useful in identifying moderate-to-severe OSA. Conclusion: Common screening tools had suboptimal performance in cardiac surgery patients. STOPBAG2 was better at predicting the probability of moderate-to-severe OSA in patients undergoing cardiac surgery compared to ESS, SA/SDQ, STOP and Berlin questionnaires.