Combined nocturnal pulse oximetry and questionnaire-based obstructive sleep apnea screening – A cohort study

Abstract

BackgroundScreening for obstructive sleep apnea (OSA) in both inpatient and outpatient settings to pursue diagnostic testing is becoming increasingly relevant, particularly given the estimates of 85–90% of patients with OSA remaining undiagnosed. Although many questionnaires are available for OSA screening, the STOP-BANG questionnaire is becoming increasingly used due to ease of use and positive performance characteristics. The utility of nocturnal oximetry, in conjunction with standard questionnaire-based strategies to enhance OSA screening in adults, has yet to be systematically examined. Research objectivesTo evaluate the utility of nocturnal oximetry measures combined with the standard STOP-BANG questionnaire as a screening strategy for OSA in the hospital setting and outpatient clinics. Study design and MethodsWe conducted a retrospective cohort study. We reviewed the electronic medical records of 130 patients who were referred to Sanford sleep center from both inpatient and outpatient settings over one year (August 1st, 2016 to August 1st, 2017). Nocturnal oximetry was conducted at home (in the outpatient group) and in the medical wards (in the inpatient group), and the following measures were obtained: Oxygen Desaturation Index (ODIPOx), mean SaO2POx and time spent below 88% SaO2 (T88Pox). Apnea-hypopnea index (AHI), mean SaO2PSG, and T88PSG from overnight polysomnography (PSG) and STOP-BANG score. ResultsBased upon likelihood ratio testing comparing discriminative ability, a model of (ODIPox + STOPBANG) was superior and more accurate than STOP-BANG alone in detecting mild OSA in the overall sample (AUC = 0.644 [0.549–0.739], p = 0.003) and inpatient sample (AUC = 0.710 [0.582–0.839], p = 0.001). This approach was also more accurate in detecting severe OSA in full sample (AUC = 0.839 [0.763–0.914], p < 0.0001), inpatient sample (AUC = 0.825 [0.711–0.939], p < 0.0001) and outpatient sample (AUC = 0.827 [0.699–0.955], p < 0.0001). The ODIPox alone was more accurate than STOP-BANG alone in detecting mild OSA in the overall sample (AUC = 0.620 [0.524–0.717], p = 0.014) and inpatient sample (AUC = 0.704 [0.574–0.835], p = 0.002) and severe OSA in full sample (AUC = 0.839 [0.764–0.915], p < 0.0001), inpatient sample (AUC = 0.827 [0.714–0.940], p < 0.0001) and outpatient sample (AUC = 0.861 [0.771–0.950], p < 0.0001). ConclusionThe use of nocturnal oximetry measures (ODIPOx) improved the accuracy of standard OSA screening with the STOP-BANG questionnaire as a screening tool in severe OSA in both inpatient and outpatient settings. Clinical implicationObstructive sleep apnea is a common sleep disorder that impacts many co-morbidities in different age groups. Enhancing affordable screening methods for OSA can facilitate early diagnosis and treatment and subsequently ameliorate morbidity and mortality related to sleep-disordered breathing.