Performance of STOP-BANG questionnaire as sleep apnoea screening tool and development of a novel pre-selection model in patients with atrial fibrillation - insights from the Virtual-SAFARI study

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Sleep-disordered breathing (SDB) is highly prevalent in patients with atrial fibrillation (AF) and its screening is recommended in this population. The STOP-BANG questionnaire is widely used as SDB screening tool, but the subjective assessment of some of its parameters might limit its validity to detect SDB in AF patients. Purpose We assessed the performance of the STOP-BANG questionnaire to detect SDB with the use of a mobile health-based portable home sleep test given as reference in a cohort of AF patients. Moreover, we explored options to improve pre-selection for SDB screening in this cohort. Methods Consecutive AF patients who were referred for AF catheter ablation in two AF outpatient clinics and without previous history of SDB and/or SDB screening, were included prospectively. Patients characteristics together with STOP-BANG questionnaire were assessed at baseline visitation. STOP-BANG scores of 3-4 and 5-8 were interpreted as intermediate and high risk for SDB, respectively. All patients were digitally referred to a virtual SDB management pathway, independent of STOP-BANG score. Patients received a portable home sleep test to detect SDB, which patients used for one night at home after instructions. Sleep recordings were accessed by sleep physicians via a secured cloud and sleep reports were discussed with patients and referring physicians. Apnoea-hypopnoea-indexes (AHI) of >=15 and of >= 30 were interpreted as moderate-to-severe and severe SDB, respectively. Results of the STOP-BANG and the home sleep test were compared and performance of the STOP-BANG questionnaire was assessed. Uni- and multivariable logistic regression analyses based on patient characteristics were used to construct a nomogram to improve pre-selection for SDB screening in this cohort. Results A total of 156 patients (median age 65 years, 63.8% male) were included. According to results from the home sleep test, 53% of patients were diagnosed with moderate-to-severe SDB, including 16% with severe SDB. STOP-BANG questionnaire performed poorly with an area under the receiver operating characteristic curve (AUROC) of 0.661 and 0.684 in predicting moderate-to-severe and severe SDB, respectively. On multivariable logistic analysis, independent predictors for moderate-to-severe SDB were body-mass-index (BMI), age and apnoeas. A nomogram was built on this variables (Fig. 1), which performed good with an AUROC of 0.728 in predicting moderate-to-severe SDB. Conclusion In a cohort of AF patients scheduled for AF catheter ablation, SDB was highly prevalent. STOP-BANG questionnaire had poor value for the prediction of moderate-to-severe and severe SDB compared to a systematical SDB screening via home sleep test. Using STOP-BANG questionnaire as a pre-selection tool might lead to frequently undetected and untreated SDB. Whether the nomogram based on our analyses improves pre-selection for SDB screening in AF patients will be validated in a future cohort.