Department of Otolaryngology-Head and Neck Surgery, Seoul St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea.
Department of Otolaryngology-Head and Neck Surgery, Bucheon St. Mary's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea.
PLoS One. 2022 May 19;17(5):e0268585. doi: 10.1371/journal.pone.0268585. eCollection 2022.
To assess the diagnostic utility of smartphone-based measurement in detecting moderate to severe obstructive sleep apnea syndrome (OSAS).
Six databases were thoroughly reviewed. Random-effect models were used to estimate the summary sensitivity, specificity, negative predictive value, positive predictive value, diagnostic odds ratio, summary receiver operating characteristic curve and measured the areas under the curve. To assess the accuracy and precision, pooled mean difference and standard deviation of apnea hypopnea index (AHI) between smartphone and polysomnography (95% limits of agreement) across studies were calculated using the random-effects model. Study methodological quality was evaluated using the QUADAS-2 tool.
Eleven studies were analyzed. The smartphone diagnostic odds ratio for moderate-to-severe OSAS (apnea/hypopnea index > 15) was 57.3873 (95% confidence interval [CI]: [34.7462; 94.7815]). The area under the summary receiver operating characteristic curve was 0.917. The sensitivity, specificity, negative predictive value, and positive predictive value were 0.9064 [0.8789; 0.9282], 0.8801 [0.8227; 0.9207], 0.9049 [0.8556; 0.9386], and 0.8844 [0.8234; 0.9263], respectively. We performed subgroup analysis based on the various OSAS detection methods (motion, sound, oximetry, and combinations thereof). Although the diagnostic odds ratios, specificities, and negative predictive values varied significantly (all p < 0.05), all methods afforded good sensitivity (> 80%). The sensitivities and positive predictive values were similar for the various methods (both p > 0.05). The mean difference with standard deviation in the AHI between smartphone and polysomnography was -0.6845 ± 1.611 events/h [-3.8426; 2.4735].
Smartphone could be used to screen the moderate-to-severe OSAS. The mean difference between smartphones and polysomnography AHI measurements was small, though limits of agreement was wide. Therefore, clinicians should be cautious when making clinical decisions based on these devices.
评估基于智能手机的测量在检测中重度阻塞性睡眠呼吸暂停综合征（OSAS）中的诊断效用。
彻底审查了 6 个数据库。使用随机效应模型来估计汇总敏感性、特异性、阴性预测值、阳性预测值、诊断比值比、汇总受试者工作特征曲线，并测量曲线下面积。使用随机效应模型计算智能手机和多导睡眠图（95%一致性界限）之间的呼吸暂停低通气指数（AHI）的平均差值和标准差，以评估准确性和精密度。使用 QUADAS-2 工具评估研究的方法学质量。
分析了 11 项研究。智能手机对中重度 OSAS（呼吸暂停/低通气指数>15）的诊断比值比为 57.3873（95%置信区间[CI]：[34.7462；94.7815]）。汇总受试者工作特征曲线下面积为 0.917。敏感性、特异性、阴性预测值和阳性预测值分别为 0.9064[0.8789；0.9282]、0.8801[0.8227；0.9207]、0.9049[0.8556；0.9386]和 0.8844[0.8234；0.9263]。我们根据各种 OSAS 检测方法（运动、声音、血氧饱和度和组合）进行了亚组分析。虽然诊断比值比、特异性和阴性预测值差异显著（均 p<0.05），但所有方法的敏感性均>80%。各种方法的敏感性和阳性预测值相似（均 p>0.05）。智能手机和多导睡眠图 AHI 之间的平均差值和标准差为-0.6845±1.611 次/小时[-3.8426；2.4735]。
智能手机可用于筛查中重度 OSAS。智能手机和多导睡眠图 AHI 测量之间的平均差值较小，但一致性界限较宽。因此，临床医生在基于这些设备做出临床决策时应谨慎。
