Correlation of STOP-bang obstructive sleep apnoea screening tool to apnoea-hypopnea index in a general cardiology population

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Introduction The link between obstructive sleep apnoea (OSA) and cardiovascular disease is well known. Prevalence can be very high in this population, sharing many pathophysiological drivers in addition to being a risk factor. More concerning however is the increasing number of patients with undiagnosed OSA, who are at higher risk of long-term complications such as coronary artery disease, hypertension, atrial fibrillation, heart failure or stroke. There are several OSA screening tools available to clinicians, with variable sensitivities and specificities depending on patient demographics. Predictive efficacy of the STOP-Bang screening tool in a general cardiology population with mixed cardiovascular disease has not been established. Proposal To identify the correlation between the STOP-Bang assessment score (1-8) and real-world Apnoea-Hypopnoea Index (AHI) in a general cardiology clinic population. Methods 250 successive patients in a general cardiology clinic were asked to complete a STOP-Bang assessment. Of those, 50 patients with a variety of STOP-Bang scores were randomly selected to undergo sleep studies. A STOP-Bang value of ≥ 5 was taken as high-risk for moderate-severe OSA. Statistical linear regression analysis was performed in IBM SPSS version 26. Results Of n = 50 patients, mean STOP-Bang score =3.7 (SD +/-2.17) with mean AHI =15.8, (SD +/- 14.3). See Table 1 for categories 1-8. N = 44 (88%) had AHI ≥ 5 (At least mild OSA). N = 18 (36%) had an AHI ≥ 15 (Moderate-severe OSA). N = 29 (58%) had symptoms such as sleepiness, loud snoring or witnessed apnoea. Correlation between STOP-Bang score and AHI was high, with an R value of 0.704, R2 = 0.496 (p < 0.01). A STOP-Bang score increase of 1 corresponded with a mean AHI rise of 4.648. When taking a STOP-Bang value of ≥ 5 as "positive", we demonstrated a sensitivity of 83.33% (95% C.I. 58.58-96.42) and specificity of 93.94% (95% C.I. 79.77-99.26%) for moderate-severe OSA (AHI ≥ 15). Conclusion We discovered a very high number of patients with undiagnosed OSA of at least mild severity. We found a strong correlation between STOP-Bang score and real-world AHI in a general cardiology population with mixed cardiovascular disease with 5 as an acceptable screening score for moderate-severe OSA. Table 1 STOP-Bang Value 1 2 3 4 5 6 7 8 Patients (N=) 12 5 7 9 5 5 5 2 AHI (Mean) 6.67 6.62 9.97 11.95 24.44 22.20 39.54 35.25 Standard Deviation (±) 3.29 2.74 3.45 4.63 12.87 5.69 27.2 6.72 Summary of mean apnoea-hypopnoea index and standard deviation for each STOP-Bang patient category. Abstract Figure 1