ObjectiveTo establish a screening model for obstructive sleep apnea hypopnea syndrome (OSAHS) through data analysis, and explore the risk factors of OSAHS. MethodsA total of 558 patients who underwent polysomnography in the Sleep Monitoring Room of Zigong Fourth People’s Hospital were recruited in the study. Among them there were 163 cases in a snore group and 395 cases in an OSAHS group. Risk factors of OSAHS were screened by both univariate analysis and multivariate analysis, then the model was established by means of binary logistic regression analysis. Finally, the screening model was evaluated by receiver operating characteristic (ROC) curve of the combined predictive factor. ResultsThe screening model of OSAHS was established as: X=–10.286+0.280×body mass index+1.057×snoring degree+1.124×sex+0.085×Epworth score+0.036×age. In this equation, sex value was 1 for men and 0 for women. If the value of X is higher than 1.123, it is likely that OSAHS would occur, and the probability (P)=ex/(1+ex). The sensitivity of the screening model was 77.70%, the specificity was 85.89%, the area under the ROC curve was 0.890, and the 95% confidence interval ranged from 0.862 to 0.918. ConclusionThis study demonstrates that a screening model based on the snoring degree, Epworth score, and body measurement data is a valuable tool to predict and screen OSAHS in patients with snoring, and the screening model could be useful in clinical diagnosis of OSAHS.
Objective To prospectively verify the accuracy and reliability of the diagnostic model of obstructive sleep apnea (OSA), including the probability model and disease severity model, and to explore a simple and cost-effective method for screening of OSA. Methods A total of 996 patients who underwent polysomnography in Zigong Fourth People’s Hospital(590 cases) and West China Hospital of Sichuan University(406 cases) were consecutively and prospectively included as the research subjects. Firstly, the OSA diagnostic model was used for the diagnostic test; then polysomnography was performed; Finally, taking polysomnography as the gold standard, the sensitivity, specificity, accuracy, positive predictive value, negative predictive value, positive likelihood ratio, negative likelihood ratio and area under the ROC curve of OSA diagnostic model were calculated, and the reliability analysis of the model’s results was carried out. Results The sensitivity, specificity and accuracy of the OSA diagnostic model were 76.38%(595/779), 83.41%(181/217) and 77.91%(776/996) respectively, the positive predictive value is 94.29%, negative predictive value is 45.49%, positive likelihood ratio is 4.604, negative likelihood ratio is 0.283; and the area under the ROC curve was 0.866. The reliability analysis of OSA diagnostic model showed that there was no significant difference in the bias comparison of AHI; the intra-class correlation coefficient(ICC) between AHI in the OSA diagnostic model and AHI in polysomnography was 0.659, with a relatively strong consistency degree; the intra-class correlation coefficient between the lowest SpO2 in the OSA diagnostic model and the lowest SpO2 in polysomnography was 0.563, with a moderate consistency degree. Conclusions The OSA diagnostic model can better predict the probability of illness and assess the severity of the disease, which is helpful for the early detection, diagnosis and treatment of OSA. The OSA diagnostic model is suitable for popularization and application in primary hospitals and when polysomnography is not available in time.