Performance of questionnaires aimed at detecting sleep disorders in patients attending a hypertension center.

INTRODUCTION: Hypertension (HT) and obstructive sleep apnea-hypopnea syndrome (OSA) are prevalent and frequently related pathologies. The predictive value of screening questionnaires in cardiovascular settings is not totally clear and could be useful to select candidates for the sleep test.

MATERIALS AND METHODS: We performed a prospective study in adults with suspected HT referred to a day-care hospital. Epworth Sleepiness Scale (ESS), Berlin, and STOP-BANG (SBQ) questionnaire were checked against data from home-based respiratory polygraphy (RP). We calculated sensitivity (S) and specificity (Sp), positive and negative predictive values (PPV and NPV), and the area under the ROC curve (AUC-ROC) for each questionnaire and their combinations.

RESULTS: We analyzed 382 patients; 234 men (61.3%) and 148 women, mean age: 54.5 ± 13.7 years, body mass index (BMI): 33.1 ± 7.8 kg/m2 , ESS: 7.4 ± 4.7. Seventy-eight percent had an apnea-hypopnea index (AHI) >5 events per hour (eV/h) and 58% presented >15 eV/h (mean: 17.5 ± 9.3 eV/h). With regard to clinically significant OSA (AHI > 15 eV/h), 5 SBQ components showed S: 100% (CI: 97-100) and Sp: 98% (CI: 95-99) with a NPV of 100, a PPV of 97.8, and an AUC-ROC of 0.90 (p < 0.0001). ESS > 10 + high-risk Berlin did not perform as effectively.

CONCLUSION: In a population of HT patients, SBQ > 5 performed better at identifying patients with >15 eV/hour with a high discrimination power.