Pulse Oximetry Monitor Feasible for Early Screening of Obstructive Sleep Apnea (OSA)

Purpose We postulate that using a simple pulse oximetry monitor (POM) to detect the severity of OSA will help clinical staff confirm the need for early treatment. Hence, we compared the POM-derived oxygen desaturation index (ODI) (events/h) with the polysomnography (PSG)-derived apnea-hypopnea index (AHI) (events/h). Our study is intended to validate the SpO(2) measurements and related ODI4% and ODI3% (events/h) calculations from POM associated with AHI and ODI from PSG based on 2007 and 2012 criteria. Methods All 73 participants (mean age: 51.04 +/- 13.14 years old) underwent an overnight PSG test and wore wristwatch POMs (PULSOX 300i) to automatically collect POM oxygen saturation (SpO(2)) data. Pearson correlation and the Bland and Altman method were used to verify the correlation between POM and PSG. Results We found that the POM SpO(2) and the PSG(2007) and PSG(2012) scores were significantly highly correlated (total record time [TRT] and lowest SpO(2), R-2 = 0.815 and 0.817; ODI4%, R-2 = 0.912 and 0.863 and ODI3%, R-2 = 0.930 and 0.914). AHI was significantly correlated with ODI4% and ODI3%, but ODI3% was nonsignificantly higher (ODI4%, r = 0.955-0.929; ODI3%, r = 0.965-0.956). Both the ODI3% and the ODI4% were highly diagnostically sensitive and specific. The ODI3% score with the AHI 15 events/h cutoff was nonsignificantly higher (area under the curve [AUC] = 0.99, AHI 15 events/h; AUC = 0.95, AHI 5 events/h). Conclusion We conclude that the ODI3% score is a feasible early screening alternative for patients with moderate-to-severe OSA.