Development of an Objective Tool for Predicting Obstructive Sleep Apnea among Adults: PAN Apnea Index

Obstructive sleep apnea (OSA) is an important health problem, which is commonly under-diagnosed especially in workplace settings. We tried to obtain a model with more objective variables due to the greater reliability in occupational settings.

A total of 374 suspected patients with OSA who underwent their first polysomnography (PSG) at Baharloo Sleep Clinic in Tehran, Iran, were enrolled in the study. Before PSG, all patients completed a questionnaire including demographic characteristics. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured for all participants. Furthermore, a blood sample was collected for measuring fasting blood sugar (FBS) and hemoglobin A1c (HbA1c). All the patients underwent full PSG. Respiratory Disturbance Index (RDI) was calculated and recorded for all patients. Different multiple adjusted logistic regression models were constructed to find the best model for prediction of OSA.

A total of 271 (72.5%) participants were men. The mean age and body mass index (BMI) were 48.58 ± 13.04 years and 30.4 ± 5.0 kg/m2 , respectively. The prevalence of RDI ≥ 15 was 78.87% (n = 295). Using regression analysis, several models were obtained, where the best one yielded sensitivity and specificity of 77.29% and 67.09%, respectively. Area under the curve (AUC) of this model was 82%. The variables of this model included SBP, age, neck circumference-height ratio (NHR), FBS, BMI, and gender (PAN apnea index) with a cutoff point ≥ 8 for high-risk individuals.

In this study, we considered only objective parameters to predict OSA which enhances reliability for diagnosis especially in occupational settings.