Screening for obstructive sleep apnea in an Afro-Caribbean Population: Diagnostic performance of common questionnaires compared with respiratory polygraphy — The MASOS Study A prospective, cross-sectional, descriptive, and analytical study

Obstructive sleep apnea hypopnea syndrome (OSAHS) is a major public health issue. Its prevalence is likely underestimated in many parts of the world, and experts project a substantial future increase. In settings with limited access to diagnostic tools, clinical screening questionnaires may improve the efficiency of patient management. This study aimed to evaluate the diagnostic performance of the main OSAHS screening questionnaires in an Afro-Caribbean population. We conducted a cross-sectional, descriptive, and analytical study, in a high-risk population of patients referred to the Sleep and Home Respiratory Assistance Unit of the Centre hospitalier universitaire de Martinique, using home respiratory polygraphy as the reference standard. Among the 134 included patients, 60 (37.3%) were men. The mean age was 57 ± 14.9 years, the mean BMI was 30.5 ± 7.5 kg/m², and the median apnea–hypopnea index (AHI) was 11.5 [6.3–25.4]. The most frequent comorbidities were hypertension (45%), allergies (29%), asthma (25%), and diabetes (19%). For detecting OSAHS (AHI ≥ 5), the area under the curve (AUC) was above 0.7 for the NoSAS, No-Apnea, STOP-BANG (moderate risk) and GOAL questionnaires, while the Berlin Questionnaire had an AUC of 0.65. The poorest performance was observed for the Epworth Sleepiness Scale (AUC = 0.55). For moderate (AHI ≥ 15) and severe OSAHS (AHI ≥ 30), No-Apnea and NoSAS achieved AUCs above 0.70, whereas GOAL and STOP-BANG showed 0.67 and 0.63, respectively. Threshold optimization using the Youden index identified higher cut-offs than standard values for some scores: No-Apnea = 6 (vs. 4), NoSAS = 10 (vs. 8), and GOAL = 3 (vs. 2). A model combining age, BMI, and neck circumference achieved an AUC of 0.81. In an Afro-Caribbean population, NoSAS and No-Apnea were the most effective screening questionnaires in this referred high-risk population for OSAHS. Based on a multivariable model combining age, BMI, and neck circumference, we developed the MASOS score (Age ≥ 70 years + 2; BMI ≥ 30 kg/m² +1; neck circumference ≥ 39 cm + 1; threshold ≥ 2), which showed a higher discriminative performance (AUC = 0.79) but still requires validation.