Usefulness of Home Overnight Pulse Oximetry in Patients with Suspected Sleep-Disordered Breathing

Abstract

Background and Objective. To determine the diagnostic yield of nocturnal oximetry versus polygraphy for the diagnosis and classification of sleep apnea hypopnea syndrome (SAHS). Methods. Prospective study conducted in a university hospital. Subjects with a clinical suspicion of SAHS were included. All of them underwent home polygraphy and oximetry on the same night. A correlation was made between the apnea-hypopnea index (AHI) and the oximetry variables. The variable with the highest diagnostic value was calculated using the area under the curve (AUC), and the best cut-off point for discriminating between patients with SAHS and severe SAHS was identified. Results. One hundred and four subjects were included; 73 were men (70%); mean age was 52 ± 10.1 years; body mass index was 30 ± 4.1, and AHI = 29 ± 23.2/h. A correlation was observed between the AHI and oximetry variables, particularly ODI3 (r = 0.850; P < 0.001 ) and ODI4 (r = 0.912; P < 0.001 ). For an AHI ≥ 10/h, the ODI3 had an AUC = 0.941 (95% confidence interval (CI) = 0.899–0.982) and the ODI4, an AUC = 0.984 (95% CI = 0.964–1), with the ODI4 having the best cut-off point (5.4/h). Similarly, for an AHI ≥ 30/h, the ODI4 had an AUC = 0.922 (95% CI = 0.859–0.986), with the best cut-off point being 10.5/h. Conclusion. Nocturnal oximetry is useful for diagnosing and evaluating the severity of SAHS. The ODI4 variable was most closely correlated with AHI for both diagnosis.

Diagnostic accuracy of oximetry for obstructive sleep apnea: a study on older adults in a home setting

“…Other studies have identified similar values for specificity, but reported the sensitivities of 42.0% ( 14) and 80.0% (15) for detecting mild apnea using ODI (compared to PSG type I). For moderate apnea, we found the same sensitivity as that for mild apnea (66.7%); however, specificity was slightly lower (85.3%) and in an intermediate range compared to the reports in the literature (69.1 to 95.1%) (14,15,20,26,29). For severe apnea, ODI sensitivity was 95.5% and specificity was 88.9%, similar to previously reported studies (14,15,20,29).…”

“…For moderate apnea, we found the same sensitivity as that for mild apnea (66.7%); however, specificity was slightly lower (85.3%) and in an intermediate range compared to the reports in the literature (69.1 to 95.1%) (14,15,20,26,29). For severe apnea, ODI sensitivity was 95.5% and specificity was 88.9%, similar to previously reported studies (14,15,20,29). Evaluation of the degree of correlation to explain the variability of ODI in relation to that of AHI revealed an r of 0.93 and R 2 of 86.17% (po0.05).…”

“…The use of oximetry has been extensively investigated in the last decade as an auxiliary method in OSA diagnosis by expanding access to diagnosis, thereby improving the quality of life of more vulnerable populations (24,25). The results of this study, as well as the existing literature (7)(8)(9)(10)(11)(12)(13)(14)(15)(16)19,21,(25)(26)(27)(28)(29)(30), indicate that, greater the degree of apnea, greater the chances of correct diagnosis based on ODI. A recent systematic review (7) demonstrates the methodological heterogeneity with respect to the following: criterion for ODI (3% and 4%), cut-off point for the classification of the degree of apnea, and the different statistical methods used.…”

“…One such study (27) evaluated the parameters of 2, 4, and 6% oxyhemoglobin desaturation from the baseline for event recording, in which the 2% criterion showed better results. Recent studies (14,15,20,29) have reported the sensitivity of ODI for OSA detection to range from 42.03 to 95.1%. In this sample, sensitivity ranged from 66.7 to 95.5% and specificity ranged from 85.3 to 91.8%.…”

Predictive Performance of Oximetry to Detect Sleep Apnea in Patients Taking Opioids

“…The best cutoff for ODI to predict moderate-to-severe and severe sleep apnea was chosen by comparing the predictive parameters of 3 commonly used ODI cutoffs (ODI ≥5, ODI ≥10, and ODI ≥15 events/h) reported in the literature. 24,25 Sensitivity of at least 70% was considered adequate to maximize the predictive performance. High sensitivity was prioritized over specificity, because in clinical setting, it will enable identification of most patients with sleep apnea.…”