Obstructive sleep apnea is the most common type of sleep breathing disorder. Therefore, the purpose of our research is to construct and verify an objective and easy-to-use nomogram that can accurately predict a patient's risk of obstructive sleep apnea. In this study, we retrospectively collected the data of patients undergoing polysomnography at the Sleep Medicine Center of the First Affiliated Hospital of Guangzhou Medical University. Participants were randomly assigned to a training cohort (50%) and a validation cohort (50%). Logistic regression and Lasso regression models were used to reduce data dimensions, select factors and construct the nomogram. C-index, calibration curve, decision curve analysis and clinical impact curve analysis were used to evaluate the identification, calibration and clinical effectiveness of the nomogram. Nomograph validation was performed in the validation cohort. The study included 1035 people in the training cohort and 1078 people in the validation cohort. Logistic and Lasso regression analysis identified age, gender, diastolic blood pressure, body mass index, neck circumference and Epworth Sleepiness Scale as the predictive factors included in the nomogram. The training cohort (C-index = 0.741) and validation cohort (C-index = 0.745) had better identification and calibration effects. The areas under the curve of the nomogram and STOP-Bang were 0.741 (0.713-0.767) and 0.728 (0.700-0.755), respectively. Decision curve analysis and clinical impact curve analysis showed that the nomogram is clinically useful. We have established a concise and practical nomogram that will help doctors better determine the priority of patients referred to the sleep centre.
Keywords: STOP-Bang; nomogram; obstructive sleep apnea; predictive factors.
© 2022 European Sleep Research Society.