LAUSR

OBJECTIVE The aim of this study was to assess the predictive accuracy of the STOP-Bang questionnaire in relation to obstructive sleep apnea (OSA) detected by nocturnal oximetry, as well as postoperative outcomes, in a population undergoing cardiac surgery. DESIGN A prospective observational cohort study. SETTING The specialist cardiothoracic center at the Royal Papworth Hospital, Cambridge University Health Partners, United Kingdom. PARTICIPANTS All adult patients, undergoing elective coronary artery bypass grafting with or without cardiac valve surgery between March 2013 and July 2014 were included. The authors excluded patients participating in other interventional studies, those who had a tracheostomy before surgery, and those who required emergency surgery or were due to be admitted on the day of surgery. INTERVENTIONS None. MEASUREMENTS AND RESULTS Cardiac surgical patients were screened for the risk of OSA with the use of STOP-Bang questionnaire. The presence of OSA prior to surgery was assessed with overnight oximetry. The predictive performance of the STOP-Bang questionnaire was assessed by calculating sensitivity, specificity, positive predictive value, negative predictive value, and area under the curve (AUC)-receiver operating characteristic curve (ROC). Multiple-logistic regression models were used to assess for associations between the STOP-Bang scores and postoperative outcomes. The STOP-Bang questionnaire discriminated poorly between mild OSA (AUC-ROC 0.57 [95% confidence interval (CI) 0.47-0.67]) and moderate/severe OSA (AUC-ROC 0.82 (95% CI 0.69-0.95)]. Accuracy was increased by modifying the cut-off value to 6 or greater, with sensitivity and specificity of 75% and 77%, respectively. A STOP-Bang score indicating the possibility of OSA was not significantly associated with prolonged intensive care unit lengths of stay (hazard ratio [HR] 1.1; 95% CI 0.99-1.19; p = 0.08) or postoperative complications (odds ratio [OR] 1.0; 95% CI 0.59-1.72; p = 0.98). CONCLUSIONS In the study population, a STOP-Bang questionnaire score of 3 or greater had limited predictive value for identifying cardiac surgical patients at high risk of OSA. STOP-Bang scores were not significantly associated with worse postoperative outcomes. A STOP-Bang score of 6 or greater could help identify patients in need of a sleep study to confirm the presence of OSA as such patients may be at increased risk of postoperative complications.