Chen Riken, Zhang Yuan, Ye Weilong, Chen Zhaojun, Liao Weifeng, Liao Huizhao, Sun Tingting, Li Huan, Cheng Junfen, Liu Wang, Yao Weimin, Yang Yitian
The Second Affiliated Hospital of Guangdong Medical University, Zhanjiang, China.
State Key Laboratory of Respiratory Disease, National Clinical Research Center for Respiratory Disease, Guangzhou Institute of Respiratory Health, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China.
Front Med (Lausanne). 2025 Jul 2;12:1581904. doi: 10.3389/fmed.2025.1581904. eCollection 2025.
To compare the predictive value of the STOP-Bang questionnaire, the Epworth Sleepiness Scale (ESS), and the Berlin questionnaire, while evaluating the combined application of these tools in a three-step screening strategy for obstructive sleep apnea (OSA).
From September 1, 2016, to October 31, 2020, at the Sleep Medicine Center of the First Affiliated Hospital of Guangzhou Medical University, 2,208 suspected OSA patients completed the ESS, STOP-Bang, and Berlin questionnaires and underwent polysomnography (PSG). The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and receiver operating characteristic (ROC) curves were calculated for each questionnaire, and the accuracy and predictive value of combining the STOP-Bang, ESS, and Berlin questionnaires for OSA screening were analyzed.
Among the individual scales, the Berlin questionnaire had the highest area under the curve (AUC), demonstrating the best diagnostic performance and the highest PPV. The ESS had the smallest AUC and the highest specificity, but the lowest sensitivity and NPV. The STOP-Bang had the highest sensitivity and NPV but the lowest specificity and PPV. When the scales were combined, the AUCs of all combinations were less than 0.7, indicating that the diagnostic performance of the combined scales slightly decreased compared to the individual scales. However, among the combined scales, the AUC of the three-scale combination was higher than that of the two-scale combinations. After combining the questionnaires, specificity and PPV increased, but sensitivity and NPV decreased. Among the two-questionnaire combinations, the sensitivity and NPV of the ESS and STOP-Bang combination were higher than those of the ESS and Berlin questionnaire combination, while specificity and PPV were lower. The combination of all three questionnaires resulted in the highest specificity and PPV, but the lowest sensitivity and NPV.
As the number of scales increases, sensitivity and NPV decrease, while specificity and PPV increase. Therefore, we recommend a three-step strategy, combining a STOP-Bang score of 3, an ESS score of 9, and the Berlin questionnaire to improve the specificity and PPV in screening for OSA.
比较STOP-Bang问卷、爱泼沃斯思睡量表（ESS）和柏林问卷的预测价值，同时评估这些工具在阻塞性睡眠呼吸暂停（OSA）三步筛查策略中的联合应用。
2016年9月1日至2020年10月31日，在广州医科大学附属第一医院睡眠医学中心，2208例疑似OSA患者完成了ESS、STOP-Bang和柏林问卷，并接受了多导睡眠图（PSG）检查。计算各问卷的敏感性、特异性、阳性预测值（PPV）、阴性预测值（NPV）和受试者工作特征（ROC）曲线，并分析联合使用STOP-Bang、ESS和柏林问卷进行OSA筛查的准确性和预测价值。
在各单项量表中，柏林问卷的曲线下面积（AUC）最大，诊断性能最佳，PPV最高。ESS的AUC最小，特异性最高，但敏感性和NPV最低。STOP-Bang的敏感性和NPV最高，但特异性和PPV最低。当各量表联合使用时，所有组合的AUC均小于0.7，表明联合量表的诊断性能与单项量表相比略有下降。然而，在联合量表中，三量表组合的AUC高于两量表组合。问卷联合后，特异性和PPV增加，但敏感性和NPV降低。在两问卷组合中，ESS与STOP-Bang组合的敏感性和NPV高于ESS与柏林问卷组合，而特异性和PPV较低。三个问卷联合使用时特异性和PPV最高，但敏感性和NPV最低。
随着量表数量的增加，敏感性和NPV降低，而特异性和PPV增加。因此，我们建议采用三步策略，即结合STOP-Bang评分为3、ESS评分为9和柏林问卷，以提高OSA筛查的特异性和PPV。
