This was a planned post hoc analysis of a prospective cohort study conducted at 5 pain clinics. Patient characteristics and daily morphine milligram equivalent (MME) dose were recorded. All consented patients underwent home overnight oximetry (PULSOX-300i, Konica Minolta Sensing, Inc, Osaka, Japan) and in-laboratory polysomnography. The predictive performance of ODI 4% from oximetry was assessed against AHI from polysomnography.
Among 332 consented patients, 181 with polysomnography and overnight oximetry data were analyzed. The mean age and body mass index of 181 patients were 52 ± 13 years and 29 ± 6 kg/m2, respectively, with 40% men. The area under the receiver operating curve for ODI to predict moderate-to-severe sleep apnea (AHI ≥15 events/h) and severe sleep apnea (AHI ≥30 events/h) was 0.82 (95% confidence interval [CI], 0.75-0.88) and 0.87 (95% CI, 0.80-0.94). ODI ≥5 events/h had a sensitivity of 85% (95% CI, 74-92) and specificity of 57% (95% CI, 52-61) to predict moderate-to-severe sleep apnea. ODI ≥15 events/h had a sensitivity of 71% (95% CI, 55-83) and specificity of 88% (95% CI, 84-91) to predict severe sleep apnea.
Overnight home oximetry has a high predictive performance in predicting moderate-to-severe and severe sleep apnea in patients on opioids for chronic pain. It is a useful additional tool for health care providers for the screening of sleep apnea in this high-risk group.
Copyright © 2021 International Anesthesia Research Society.