ABSTRACT
RATIONALE: Owing to resource limitations, the testing of patients for obstructive sleep apnea (OSA) is often delayed. There is a need to accurately triage and expedite testing in those with a high pretest probability of OSA. Acoustic pharyngometry is a simple, noninvasive technique used to assess the upper airway cross-sectional area (UA-XSA), which is known to be reduced in those with OSA. OBJECTIVES: To determine the discriminative ability and predictive value of UA-XSA measurements by acoustic pharyngometry for OSA. METHODS: We conducted a cross-sectional study with a clinical cohort of consecutive adults with suspected OSA who had undergone both polysomnography and acoustic pharyngometry. OSA was defined as an apnea-hypopnea index greater than or equal to 5. Multivariable logistic regression analyses and receiver operating characteristic curves were used. MEASUREMENTS AND MAIN RESULTS: The cohort included 576 subjects, 87% of whom had OSA and 64% of whom were men. The subjects' median body mass index (BMI) was 30.3 kg/m2, and their median age was 57 years. The median UA-XSA at FRC when sitting was significantly smaller in those with OSA compared with those without OSA (3.3 cm2 [interquartile range, 2.7-3.8] vs. 3.7 cm2 [interquartile range, of 2.9-4.2]). When the analysis was controlled for age, sex, BMI, and comorbidities, the odds of OSA increased for every 1-cm2 decrease in the mean UA-XSA FRC when sitting (odds ratio, 1.62; 95% confidence interval, 1.23-2.13). The mean UA-XSA provided fair discrimination for OSA (area under the curve, 0.60). A cutoff value of 3.75 cm2, the point with the best sum of sensitivity and specificity, had sensitivity of 73% and specificity of 46%. The magnitude of the incremental discriminative value of UA-XSA over clinical variables (age, sex, BMI, and comorbidities) was small and nonsignificant (P = 0.5). CONCLUSIONS: The mean UA-XSA at FRC when sitting or supine provided no further significant advantage over clinical variables for the discernment of OSA.
