Value of actigraphy and oximetry for diagnosing sleep apnea-hypopnea syndrome

Abstract

To explore the value of actigraphy and oximetry for diagnosing sleep apnea-hypopnea syndrome (SAHS). Suspected SAHS subjects were enrolled from sleep center of Peking University People's Hospital between December 2013 and September 2014. Light-out and light-on were simultaneously recorded for polysomnography (PSG), actigraphy and oximetry. Actigraphy was used to correct total sleep time (TST) for oximetry. Oxygen desaturation index (ODI4), namely the times of pulse oxygen saturation (SpO2) drop≥4% per hour, was detected by oximetry. ODIA was used instead of ODI4 with TST corrected by actigraphy. Receiver operating characteristic (ROC) curve was used for evaluating the value of ODI4 and ODIA for diagnosing SAHS and cut-off value was calculated. Paired t-test or signed rank t-test was used for data acquired using actigraphy or oximetry compared with data acquired by PSG. Sensitivity and specificity were calculated using fourfold table. Disease severity of SAHS was classified by apnea-hypopnea index (AHI) detected by PSG. If AHI<5/h, SAHS was not considered. 5-<15/h was classified as mild, 15-<30/h was classified as moderate and AHI≥30/h was classified as severe. Among 213 SAHS subjects, 38 of them were normal, 34 of them were mild, 51 were moderate and 90 were severe. Compared with PSG, actigraphy can correctly estimate TST in non-SAHS (P=0.408) and mild SAHS groups (P=0.949); while in moderate and severe SAHS groups, TSTs detected by actigraphy were shorter than TSTs acquired by PSG ((405±51) vs (419±40) min and (399 ± 62) vs (422 ± 60) min) (both P<0.05). Area under ROC curve (AUC) of ODI4 for diagnosing SAHS was 0.956 using a cut-off value of 5; the AUC of ODIA for diagnosing SAHS was 0.951 with a cut-off value of 5 as well. The sensitivity of ODI4 for mild, moderate and severe SAHS was 80.6%, 66.7% and 58.9% respectively, all with a specificity of 100%. After adjusting TST with actigraphy, the sensitivity of ODIA for mild, moderate and severe SAHS was 84.0%, 73.8 % and 68.9%, with specificity of 94.7%, 100% and 99.2% respectively. Actigraphy can correctly estimate TST in non-SAHS and mild SAHS groups. ODI4 can be used for diagnosing SAHS. After correcting TST by actigraphy, the sensitivity of ODIA is higher.