ADVERTISEMENT

Issue Navigator

Volume 15 No. 04
Earn CME
Accepted Papers





Scientific Investigations

Validation of the Nox-T3 Portable Monitor for Diagnosis of Obstructive Sleep Apnea in Patients With Chronic Obstructive Pulmonary Disease

Yuan Chang, MD1; Liyue Xu, MD2,3; Fang Han, MD2; Brendan T. Keenan, MS3; Elizabeth Kneeland-Szanto, MBA3; Rongbao Zhang, MD2; Wei Zhang, MD4; Yongbo Yu, BS4; Yuhua Zuo, BS2; Allan I. Pack, MBChB, PhD3; Samuel T. Kuna, MD3,5
1Department of Respiratory Medicine, Peking University International Hospital, Beijing, China; 2Department of Respiratory Medicine, Peking University People's Hospital, Beijing, China; 3Department of Medicine and Center for Sleep and Circadian Neurobiology, University of Pennsylvania, Philadelphia, Pennsylvania; 4PKU-UPenn Sleep Center, Peking University International Hospital, Beijing, China; 5Department of Medicine, Corporal Michael J. Crescenz Veterans Affairs Medical Center, Philadelphia, Pennsylvania

ABSTRACT

Study Objectives:

Clinical practice guidelines recommend polysomnography (PSG) for diagnosis of obstructive sleep apnea (OSA) in patients with major comorbidities. We evaluated home sleep apnea testing (HSAT) using a type 3 portable monitor (PM, Nox-T3, Nox Medical, Reykjavik, Iceland) to diagnose OSA in adults with chronic obstructive pulmonary disease (COPD).

Methods:

Ninety adults with COPD (89.0% males, mean ± standard deviation age 66.5 ± 7.8 years, body mass index 27.5 ± 5.8 kg/m2, forced expiratory volume in the first second/forced vital capacity [FEV1/ FVC] 53.5 ± 12.4%, FEV1 54.0 ± 18.4% predicted) underwent unattended HSAT followed by an in-laboratory PSG with simultaneous PM recording.

Results:

Scoring hypopneas with a ≥ 4% oxygen desaturation, the apnea-hypopnea index (AHI) was 16.7 ± 20.6 events/h on HSAT, 20.0 ± 23.3 events/h on in-laboratory PM, and 21.2 ± 26.2 events/h on PSG (P < .0001). Bland-Altman analysis of AHI on HSAT versus PSG showed a mean difference (95% confidence interval) of −5.08 (−7.73, −2.42) events/h (P = .0003) and limits of agreement (± 2 standard deviations) of −30.00 to 19.85 events/h; HSAT underestimated AHI to a greater extent for more severe values (rho = −.529, P < .0001). Using an AHI ≥ 5 events/h to diagnose OSA, HSAT had 95% sensitivity, 78% specificity, 88% positive predictive value, and 89% negative predictive value compared to PSG. Mean oxygen saturation was 93.2 ± 3.7% on PSG, 91.0 ± 3.3% on in-laboratory PM, and 90.8 ± 4.0% on HSAT (P < .0001). Percentage time oxygen saturation ≤ 88% was 17.9 ± 26.4% on HSAT, 17.4 ± 25.5% on in-laboratory PM, and 10.0 ± 21.1% on PSG (P < .0001).

Conclusions:

The Nox-T3 PM can be used to diagnose OSA in patients with COPD but, most likely due to differences among pulse oximeters, a greater number of patients with COPD and without OSA qualified for nocturnal oxygen treatment using this PM than PSG.

Citation:

Chang Y, Xu L, Han F, Keenan BT, Kneeland-Szanto E, Zhang R, Zhang W, Yu Y, Zuo Y, Pack AI, Kuna ST. Validation of the Nox-T3 portable monitor for diagnosis of obstructive sleep apnea in patients with chronic obstructive pulmonary disease. J Clin Sleep Med. 2019;15(4):587–596.


BRIEF SUMMARY

Current Knowledge/Study Rationale: Current clinical practice guidelines recommend in-laboratory polysomnography to diagnose obstructive sleep apnea in patients with major comorbidities due to lack of evidence supporting the use of home sleep apnea tests. This study evaluated the performance of a portable monitor (Nox-T3) used for home sleep apnea testing to diagnose obstructive sleep apnea in patients with chronic obstructive pulmonary disease (COPD).

Study Impact: Our results showed good agreement between home testing with in-laboratory polysomnography to diagnose obstructive sleep apnea in patients with COPD, potentially improving their access to care. However, the lower oxygen saturation levels recorded on this portable monitor than on the in-laboratory polysomnography, likely due to differences among pulse oximeters, would result in more patients with COPD being started on oxygen treatment based on this portable monitor.

INTRODUCTION

Current clinical practice guidelines recommend that polysomnography (PSG) should be used to diagnose obstructive sleep apnea (OSA) in patients with conditions such as chronic obstructive pulmonary disease (COPD) and chronic heart failure. This recommendation is based on the lack of evidence supporting the use of home sleep apnea testing (HSAT) to diagnose OSA in patients with comorbid conditions.1 Almost all previous trials validating HSAT to diagnose OSA excluded patients with cardiorespiratory disorders. The reason for their exclusion is not apparent, particularly with regard to type 3 portable monitors (PM) that can differentiate obstructive from central apneas and detect Cheyne-Stokes respiration using the nasal pressure and rib cage and abdominal movement signals. Exclusion of patients with comorbid conditions in previous studies may have been out of concern that patients with COPD and other cardiorespiratory disorders might not be capable of performing the self-administered HSAT. For example, in a previous study evaluating HSAT in patients with COPD, a high percentage of the PM recordings were unsuccessful and those that could be interpreted were not in close agreement with the PSG results.2

Flenley3 coined the term “overlap” syndrome to describe patients with both OSA and COPD. Compared to patients with either condition alone, patients with overlap syndrome are at greater risk of adverse clinical outcomes.46 Marin et al. found that patients with overlap syndrome who were not on continuous positive airway pressure (CPAP) treatment had a higher mortality and were more likely to suffer a severe COPD exacerbation leading to hospitalization compared to patients with overlap syndrome treated with CPAP and a COPD-only group.7 The prevalence and adverse consequences of overlap syndrome make it particularly important that these patients have readily available access to diagnostic testing and, therefore, treatment of OSA. However, the higher cost and longer wait times of in-laboratory PSG can limit access to care. The ability to use lower cost and more accessible HSAT to diagnose COPD could remove this barrier. The purpose of the current study was to evaluate HSAT using a type 3 PM (Nox-T3, Nox Medical Inc., Reykjavik, Iceland) to diagnose OSA in adults with COPD.

METHODS

Protocol

A total of 91 patients with COPD referred for evaluation of OSA volunteered to participate. The diagnosis of COPD was confirmed on pulmonary function testing (forced expiratory volume in the first second/forced vital capacity (FEV1/FVC) < 70% of predicted, FEV1 < 80% of predicted, and persistence of airflow obstruction following acute administration of an inhaled bronchodilator) obtained within the previous 2 years; one person was withdrawn because his pulmonary function test results did not support the diagnosis of COPD. The remaining participants were between 18 and 80 years old and had no previous sleep testing or treatment for OSA. Individuals were considered ineligible (excluded) for the following reasons: prior diagnosis of central sleep apnea/Cheyne-Stokes respiration, obesity hypoventilation syndrome, narcolepsy, rapid eye movement sleep behavior disorder, heart failure; shift work schedule, regular “jet lag” or irregular work schedules by history over 3 months; a clinically unstable medical condition as defined by a change in medications in the previous month; or a COPD exacerbation or new medical diagnosis in the previous 2 months (eg, myocardial infarction, active infection, thyroid disease, depression or psychosis, cirrhosis, surgery, or cancer). Individuals prescribed supplemental oxygen therapy were also excluded because this treatment could prevent the oxygen desaturation associated with hypopneas. The Institutional Review Boards at Peking University People's Hospital (2015PHB1878) and Crescenz VA Medical Center (ID#: 01603, Prom#: 0034) approved the project. Written informed consent was obtained from all participants.

All participants were asked to initially perform an overnight HSAT using the Nox-T3 PM followed, within 1 week, by an in-laboratory PSG (Alice6, Philips Respironics, Inc, Murrysville, Pennsylvania, United States at Peking University People's Hospital; Sandman, Natus Medical Inc., Pleasanton, CA at Crescenz VA Medical Center) with a simultaneous Nox-T3 PM recording. The order of home and in-laboratory testing was fixed in order to assess the ability of individuals with no previous experience with sleep testing to successfully perform the HSAT. Participants were instructed to sleep in whatever positions were comfortable for them and to take their regular medications.

Portable Monitor and PSG Recordings

The following signals were recorded during the PM recordings: nasal pressure as a surrogate measure of airflow, rib cage and abdominal movement by inductance plethysmography, snoring, body position, activity, and heart rate and oxygen saturation by pulse oximetry. For the HSAT, the participants, with or without family members, came to the sleep center to receive instructions on how to apply the sensors and perform the recording. During the session, a trained sleep technologist demonstrated how to apply the sensors and the participant or family member was then asked to apply the sensors. After the technician confirmed proper placement, the sensors were removed and the participant or family member reapplied the sensors at home just prior to bedtime. The morning after the HSAT, the participant completed a poststudy questionnaire to report events during the recording. During the in-laboratory testing, the sleep technologist applied the PM sensors and initiated the PM recording. Separate sensors were used for the simultaneous PM and PSG recordings.

The ability of participants to perform HSAT was assessed as the percentage of individuals with a successful initial HSAT and the quality of the HSAT. A successful HSAT required at least 3 hours of recording containing the oxygen saturation and at least one of the respiratory signals (nasal pressure, rib cage movement, abdominal movement). If the initial HSAT was unsuccessful, the participant took a PM home after the PSG and performed another HSAT. If the second attempt was unsuccessful, the HSAT was not repeated. The quality of the HSAT was assessed by the automated analysis by the Noxturnal software program (Nox Medical) of the percentage of analysis time that HSAT signals were present and able to be used in scoring.

PSG tests were performed according to the recommendations of the American Academy of Sleep Medicine.8 The following signals were recorded: electroencephalogram (F3M2, F4M1, C3M2, C4M1, O1M2, O2M1), bilateral electrooculogram, chin muscle electromyogram, oronasal thermistor, nasal pressure, rib cage and abdominal movement, electrocardiogram (lead 1), snoring, body position, bilateral anterior tibialis electromyograms, and heart rate and oxygen saturation by pulse oximetry.

Portable Monitor and PSG Scoring

Scoring of the PM recordings and PSG were performed by experienced scorers. The scorers were blinded to whether the PM recording was performed at home or in-laboratory and to the participant's PSG results when scoring a HSAT.

Analysis start time and stop time on the PM recordings was determined based on the participant's responses on the post-study questionnaire and the activity signal on the recording. The PM recordings were initially scored automatically using Noxturnal software and then manually edited. The PSG tests were scored manually with the aid of computer software. Using American Academy of Sleep Medicine 2012 scoring criteria,9 apneas were defined as ≥ 90% reduction in airflow from baseline for at least 10 seconds. Obstructive apneas were defined as an apnea associated with respiratory effort and central apneas were defined as an apnea during which respiratory effort was absent. Mixed apneas were defined as an apnea during which respiratory effort was initially absent but appeared in the latter part of the event. Two separately manually edited scorings were performed using different definitions for hypopnea on PM tests and PSG tests: (1) a ≥ 30% reduction in a respiratory signal for ≥ 10 seconds associated with a ≥ 4% reduction in oxygen saturation, and (2) a ≥ 30% reduction in a respiratory signal for at least 10 seconds associated with an oxygen desaturation event of 3% or greater on PM recordings, and an oxygen desaturation event of ≥ 3% and/or an arousal on PSG. When the PM nasal pressure signal was absent or not able to be scored, respiratory events were scored using the flow signal derived from the respiratory inductance plethysmography signals.10 The average number of apneas and hypopneas per hour (apnea-hypopnea index, AHI), and oxygen desaturation events per hour (oxygen desaturation index, ODI) were calculated based on total analysis time on the Nox-T3 recordings and total sleep time on PSG. The number of apneas and hypopneas per hour of recording on HSAT is commonly referred to as the respiratory event index (REI). However, the complexity of using AHI and REI throughout the manuscript, especially in the tables, would have been unwieldy. Therefore, the term AHI is used throughout the manuscript for both technologies.

Statistical Analysis

Continuous variables are summarized using mean and standard deviation (SD) and categorical variables using count and percentage. Comparisons of respiratory parameters across the three monitoring methods (HSAT, in-laboratory PM and in-laboratory PSG) were compared using a repeated-measure analysis of variance, accounting for multiple observations per participant. Statistical significance in these comparisons was defined using domain-specific Bonferroni corrected values of P (equal to .05 divided by the number of measures within each domain) of (1) P < .0071 for respiratory indices (n = 7 measures), (2) P < .0083 for number of respiratory events (n = 6 measures), and (3) P < .0083 for oxygen saturation measures (n = 6 measures); a value of P < .05 was considered suggestive evidence of an association. Pairwise contrasts among methods were performed if there was at least suggestive evidence (P < .05) to reject global null hypothesis of no difference between monitoring methods. To assess the level of agreement between the monitoring methods, we used paired tests and methods described by Bland and Altman.11,12 Specifically, for a given metric (eg, AHI), we first calculated the participant-specific difference for each pair of methods and tested whether this was significantly different from zero using paired t tests. Next, for each pair of monitoring methods, we examined the relationship between the participant-specific difference and the participant-specific average value using the two techniques. This relationship was evaluated graphically and statistically for bias, including examining the average participant-specific difference and associated limits of agreement (equal to the mean difference ± 2 SD) and testing for significant correlation between the participant-specific difference and mean (eg, whether differences between techniques are larger/smaller for higher/lower average values). Paired comparisons of proportions meeting specific AHI thresholds were evaluated using exact P values from the McNemar test. Primary agreement analyses compared the ≥ 4% AHI on PSG to that obtained from HSAT and in-laboratory PM, separately, with PSG results considered the gold standard. Secondary analyses repeated this comparison using the ≥ 3% rule. Using similar methods, we also examined the agreement between manual and automated AHI scoring within the HSAT and in-laboratory monitors, separately, with manual scoring considered the gold standard.

To understand the ability of home-based testing to accurately diagnose OSA when compared to PSG in patients with COPD, we examined the diagnostic characteristics of home-and laboratory-based PM testing. In particular, we calculated the sensitivity, specificity, positive predictive value, and negative predictive value at AHI thresholds of ≥ 5, ≥ 10, ≥ 15 and ≥ 30 events/h for each PM method, using the results of the in-laboratory PSG as the gold standard. In addition, we calculated percent agreement and kappa coefficients among the three methods for both a 4-level AHI grouping (< 5, 5 to < 15, 15 to < 30, and ≥ 30 events/h) and these binary cutoff points. Finally, to understand the effect of PM testing on treatment for COPD in the absence of OSA, we examined the proportion of participants with an AHI < 5 events/h who met criteria for prescribing home nocturnal supplemental oxygen therapy, defined as saturation of oxygen (SpO2) ≤ 88% for at least 5 minutes of sleep in an individual who demonstrates an arterial oxygen saturation ≥ 89% while awake.13

RESULTS

Sample Characteristics

The 90 adults with COPD had a mean age of 66.5 ± 7.8 years, BMI of 27.5 ± 5.8 kg/m2 and 89% were male (Table 1). The mean FEV1/FVC was 53.5 ± 12.4% and FEV1 was 54.0 ± 18.4% of predicted. Chinese participants (n = 59) had lower BMI than the American participants (n = 31, P = .009), as well as longer total sleep time (P = .0008), a greater percentage of stage N1 sleep (P < .0001), lower percentage of stage N2 sleep (P < .0001), and lower FEV1 (P = .022) and FVC (P = .003) values, but no difference in percentage of predicted FEV1, FVC, or FEV1/FVC ratio.

Participant characteristics overall and by site.

jcsm.15.4.587.t01.jpg

table icon
Table 1

Participant characteristics overall and by site.

(more ...)

Success Rate

The initial HSAT was unsuccessful in 5 of the 90 participants (5.6%). Following repeat testing in 3 participants, HSAT was successfully completed in 88 of 90 patients (97.8%).

Comparison of Respiratory Parameters Across Techniques

Figure 1 illustrates the percentage of participants with no OSA and mild, moderate, and severe OSA based on the AHI on PSG, in-laboratory PM recording and HSAT. As compared to PSG, a similar proportion of individuals received a diagnosis of OSA (defined as AHI ≥ 5 events/h) using both HSAT (P = .344) and in-laboratory PM (P = .453). However, a lower proportion of participants met criteria for moderate or severe OSA (AHI ≥ 15 events/h) based on HSAT when compared to results from PSG (P = .012) or in-laboratory PM (P = .031). When compared to total sleep time on PSG, total analysis time was approximately 100 minutes greater on both the in-laboratory PM (P < .0001) and HSAT (P < .0001) recordings, as expected given the lack of sleep staging on PM (Table S1 in the supplemental material). Suggestive differences across techniques were also seen for number of total apneas (P = .0167) and obstructive apneas (P = .0220) with the in-laboratory PM measuring more total apneas than PSG (P = .016) and HSAT (P = .011) and more obstructive apneas than PSG (P = .016) and HSAT (P = .017). Although there were no differences among the different types of recordings in the number of 3% and 4% hypopneas, we saw suggestive differences in their indices (P = .0427 and .0285, respectively). Compared to PSG, HSAT estimated a lower hypopnea index at 3% (P = .014) and 4% (P = .016), whereas in-laboratory PM resulted in a lower hypopnea index 4% (P = .030). Relatedly, we observed significant differences in AHI 4% (P < .0001) and AHI 3% (P < .0001) across testing methods, with HSAT resulting in lower AHI estimates compared to both PSG and in-laboratory PM (all P ≤ .003). The average supine time was not significantly different among the recording methods (P = .3265).

OSA groupings.

Graph illustrating the percentage of patients falling into clinical OSA groupings of none (AHI < 5 events/h), mild (AHI 5 to < 15 events/h), moderate (AHI 15 to < 30 events/h), and severe (AHI ≥ 30 events/h) based on PSG, Nox-T3lab, and Nox-T3home. Scoring of hypopneas in all three recordings required a ≥ 4% oxygen desaturation event. AHI = apnea-hypopnea index, Nox-T3home = home testing using the Nox-T3 device, Nox-T3lab = in-laboratory portable monitor recording using the Nox-T3 device, OSA = obstructive sleep apnea, PSG = polysomnography.

jcsm.15.4.587a.jpg

jcsm.15.4.587a.jpg
Figure 1

OSA groupings.

(more ...)

Agreement Between Monitoring Methods

Bland-Altman and identity plots in Figure 2 compare manually edited AHI 4% on PSG with that on HSAT and in-laboratory portable recording. The Bland-Altman analysis of AHI 4% on HSAT versus PSG showed a mean difference of −5.08 events/h (95% confidence interval [CI]: −7.73, −2.42; P = .0003), with limits of agreement ranging from −30.00 to 19.85 events/h. In contrast, the mean difference in the Bland-Altman plot of AHI on in-laboratory PM recording versus PSG was only −1.77 events/h (95% CI: −3.17, −0.36; P = .0142) with narrower limits of agreement of −14.86 to 11.33 events/h. In both cases, there was evidence for a significant negative correlation between the difference and mean, suggesting that at higher AHI values the PM results in larger underestimates of the PSG AHI. The squared correlation coefficient (R2) for AHI was .791 comparing the PSG versus HSAT and .949 when comparing the PSG versus in-laboratory PM recording, suggesting a high amount of shared variability. Similar results were observed when scoring of hypopneas on PM required a ≥ 3% oxygen desaturation event and hypopneas on PSG required a ≥ 3% oxygen desaturation event and/or an arousal (Figure S2 in the supplemental material). The closer relationship between PSG and simultaneous in-laboratory PM recording than between PSG and HSAT supports the importance of differences in environment and night-to-night variability on sleep test results.

AHI 4% measured by Nox-T3home and Nox-T3lab compared to PSG.

Bland-Altman (A, C) and identity (B, D) plots of manually edited AHI 4% on PSG compared to manually edited Nox-T3home(A, B) and Nox-T3lab(C, D). Scoring of hypopneas in all three recordings required a ≥ 4% oxygen desaturation event. AHI = apnea-hypopnea index, Nox-T3home = home testing using the Nox-T3 device, Nox-T3lab = in-laboratory portable monitor recording using the Nox-T3 device, PSG = polysomnography.

jcsm.15.4.587b.jpg

jcsm.15.4.587b.jpg
Figure 2

AHI 4% measured by Nox-T3home and Nox-T3lab compared to PSG.

(more ...)

Table 2 compares the diagnostic characteristics for different cutoffs of manually edited AHI from the HSAT and in-laboratory PM recording compared to gold-standard PSG when hypopneas in all recordings required ≥ 4% oxygen de-saturation. Using a threshold of AHI ≥ 5 events/h, the HSAT had 95% sensitivity, 78% specificity, 88% positive predictive value, 89% negative predictive value, and there was 88.5% agreement between the two methods with a high kappa coefficient of 0.746 (Table 3). Similar results for HSAT were observed at AHI cutoffs of ≥ 10 and ≥ 15 events/h, with specificity increasing to 98% with decrease in sensitivity (74%) at an AHI cutoff ≥ 15 events/h. Thus, despite differences in the estimated AHI, the HSAT shows good performance in diagnosing OSA when compared to corresponding criteria from gold-standard PSG. Simultaneous in-laboratory PM testing showed similar or higher predictive value. Moreover, even when dividing AHI 4% into four clinically meaningful groups (< 5, 5–15, 15–30, and ≥ 30), we continued to observe moderate to high reliability, with kappa coefficients of 0.567 for PSG versus HSAT, 0.808 for PSG versus in-laboratory PM and 0.610 for HSAT versus in-laboratory PM, and percent agreements ranging from 67.8% to 86.1% (Table 3). Similar results were observed based on a scoring rule of a ≥ 3% oxygen de-saturation event and/or an arousal (Table 3 and Table S2 in the supplemental material).

Values for different cutoffs of manually edited AHI 4% measured by Nox-T3home and Nox-T3lab versus PSG.

jcsm.15.4.587.t02.jpg

table icon
Table 2

Values for different cutoffs of manually edited AHI 4% measured by Nox-T3home and Nox-T3lab versus PSG.

(more ...)

Comparison of categorized data.

jcsm.15.4.587.t03.jpg

table icon
Table 3

Comparison of categorized data.

(more ...)

AHI on automatic scoring of the PM recording had good agreement with manually edited scoring (Figure 3). On the HSAT, there was no significant mean difference in manual or automatic AHI values (−0.63 [95% CI: −0.22, 1.47]; P = .1432) and associated limits of agreement ranged from −7.31 to 8.57 events/h. For the in-laboratory PM, there was a slightly larger difference between the automatic and manual scoring (mean difference −1.42 [0.60, 2.25]; P = .0009); although statistically significant, this difference is not clinically meaningful. The limits of agreement ranged from −6.20 to 9.05 events/h, which is almost the same as that observed for HSAT. For both monitoring methods, there was no significant correlation between the difference and mean of manual and autoscoring, suggesting any differences in AHI are consistent across disease severity. Supporting these results, the R2 between automatic and manually scored AHI in identity plots was .963 for HSAT and .974 for in-laboratory PM, reflecting near-perfect correlation.

Automatically scored AHI 4% compared to manually edited AHI 4% measured by Nox-T3home and Nox-T3lab.

Bland-Altman (A, C) and identity (B, D) plots comparing automatically scored AHI 4% to manually edited AHI 4% on Nox-T3home(A, B) and Nox-T3lab(C, D). Scoring of hypopneas in all three recordings required a ≥ 4% oxygen desaturation event. AHI = apnea-hypopnea index, Nox-T3home = home testing using the Nox-T3 device, Nox-T3lab = in-laboratory portable monitor recording using the Nox-T3 device.

jcsm.15.4.587c.jpg

jcsm.15.4.587c.jpg
Figure 3

Automatically scored AHI 4% compared to manually edited AHI 4% measured by Nox-T3home and Nox-T3lab.

(more ...)

Comparison of Oxygen Saturation Across Techniques

Statistically significant differences in oxygen desaturation severity measures were observed among the three recording methods (Table S1), including mean SpO2 (P < .0001) and percent time SpO2 < 90% (P < .0001) and SpO2 ≤ 88% (P < .0001). Compared to PSG, mean oxygen saturation was significantly lower on both the in-laboratory PM (P < .0001) and HSAT recording (P < .0001) (Figure 4). This difference was observed at both sites (data not shown). Percentage times with SpO2 < 90% and ≤ 88% were greater on both in-laboratory PM (P < .0001 and P = .0001) and HSAT (P < .0001 for both comparisons) compared to PSG. When examining ODI between methods, ODI ≥ 4% was lower on HSAT compared to the PSG (P = .006). The ODI ≥ 3% and ODI ≥ 4% were both lower on HSAT when compared to in-laboratory PM (P = .001 and P = .011) (Figure S1 in the supplemental material). Bland-Altman analysis of ODI ≥ 4% on in-laboratory PM versus PSG showed a mean difference (95% CI) of −0.24 (−2.45, 1.97) events/h (P = .829) and limits of agreement (± 2 SD) of −20.89 to 20.40 events/h.

Average SpO2 measured by Nox-T3home and Nox-T3lab compared to PSG.

Bland-Altman (A, C) and identity (B, D) plots of average SpO2 on PSG compared to Nox-T3home(A, B) and Nox-T3lab(C, D). Nox-T3home = home testing using the Nox-T3 device, Nox-T3lab = in-laboratory portable monitor recording using the Nox-T3 device. PSG = polysomnography, SpO2 = oxygen saturation on pulse oximetry.

jcsm.15.4.587d.jpg

jcsm.15.4.587d.jpg
Figure 4

Average SpO2 measured by Nox-T3home and Nox-T3lab compared to PSG.

(more ...)

Based on Medicare guidelines, adults with COPD with an awake SpO2 ≥ 89% who do not have OSA or are on CPAP treatment for OSA qualify for home supplemental oxygen therapy if their oxygen saturation is ≤ 88% for at least 5 minutes during sleep.13 Using this guideline, 5 of the 34 participants without OSA (14.7%) (AHI < 5 events/h) on PSG achieved the Medicare guidelines for nocturnal oxygen therapy (Table 4). In contrast, 51.7% of participants without OSA on in-laboratory PM and 60.7% of participants without OSA on HSAT met these criteria for home oxygen therapy.

Participants without OSA who qualified for nocturnal oxygen therapy based on Medicare guidelines.13

jcsm.15.4.587.t04.jpg

table icon
Table 4

Participants without OSA who qualified for nocturnal oxygen therapy based on Medicare guidelines.13

(more ...)

DISCUSSION

The results demonstrate that HSAT using a type 3 PM shows good performance in diagnosing OSA in adults with COPD when compared to PSG. When comparing HSAT with PSG, we found high positive and negative predictive probabilities and high sensitivity and specificity. These results are similar to those of our previous study that validated the use of HSAT to diagnose OSA in Chinese adults without major comorbidities.14 Comparing simultaneous PM recordings and PSG in the patients with COPD, limits of agreement on Bland-Altman analysis ranged from −14.86 to 11.33, with a relatively small bias (−1.77 events/h). Perhaps reflecting differences in environment on HSAT versus PSG and the known night-to-night variability of AHI even using PSG,15 wider limits of agreement (−30.00 to 19.85 events/h) and a larger estimated difference in AHI (−5.08 events/h) were observed between HSAT and PSG. Nevertheless, the kappa coefficient of AHI 4% between HSAT and in-laboratory PM was 0.610, which was similar to the agreement in respiratory disturbance index (K = 0.62) in a previous study evaluating night-to-night PM testing performed in a sleep laboratory.16 The failure rate of 5.6% on initial HSAT in the participants with COPD (decreased to 2.23% on repeat testing) is similar to that in our previous study in adults without major comorbidities.14 The high diagnostic performance of HSAT when compared to PSG and the low failure rate demonstrate the ability to use the Nox-T3 PM for home testing to diagnose OSA in patients with COPD.

Indeed, one would predict that patients with COPD would be ideal candidates for HSAT. The relatively low baseline oxygen saturation in patients with COPD makes it more likely that the AHI on a type 3 PM recording, which lacks the ability to detect arousals, will be similar to the AHI on PSG when hypopneas on PSG are scored based on an associated oxygen desaturation event and/or arousal. Give their relatively low baseline oxygen saturation, patients with COPD are closer to the steep portion of oxyhemoglobin dissociation curve and more likely to have an oxygen desaturation associated with a reduction in airflow. This reduces the importance of not being able to detect arousals on the type 3 PM recording. Despite the similarities in AHI between PM and PSG recording, a greater number of apneas were identified on the in-laboratory PM than PSG. This may have been due to scoring events during wakefulness in the PM recordings and the use of oronasal thermistor and nasal pressure to detect airflow during PSG, but just nasal pressure during PM testing.

Despite the good performance of the Nox-T3 HSAT in diagnosing OSA in adults with COPD, we found significantly lower mean oxygen saturation and greater percentage time of oxygen saturation < 90% and ≤ 88% on the HSAT and in-laboratory PM recordings compared to PSG. Our previous study validating Nox-T3 HSAT in participants without major comorbidities also showed lower oxygen saturation results on HSAT and in-laboratory PM compared to PSG.14 However, given the higher oxygen saturation levels in those participants, the differences were not clinically significant. Based on Medicare guidelines, approximately 15% of our COPD participants without OSA on PSG would have qualified for home oxygen therapy. This number greatly increased when based on Nox-T3 PM testing, with 52% of participants without OSA qualifying for home oxygen on in-laboratory PM and 61% on HSAT. These results are most likely due to the different pulse oximeters in the recording instruments. The Alice 6 system uses a Masimo pulse oximeter (Masimo Corp., Irvine, California, United States), Sandman uses a Nellcor pulse oximeter (Medtronic, Minneapolis, Minnesota, United States), and the Nox-T3 uses a Nonin pulse oximeter (Nonin Medical, Plymouth, Minnesota, United States). To our knowledge, there is no restriction as to which pulse oximeter is used on overnight testing to determine the need for supplemental oxygen therapy during sleep. However, it is well known that significant differences in signal processing exist across pulse oximeters.1720 Hannhart et al.21 compared SpO2 from six pulse oximeters in patients with hypoxia and COPD against arterial oxygen saturation (SaO2) measured in simultaneously withdrawn arterial blood. The bias (mean SpO2 – SaO2 difference) ranged from 1.2% to 4%, and the lower the SaO2, the greater the difference with SpO2. Of note, the equipment differences in our study did not appear to affect the ODI 4% and ODI 3%, which were in close agreement on simultaneous PM and PSG recordings (Table S1 and Figure S1).

There are a limited number of studies evaluating portable sleep monitor testing to diagnose OSA in patients with cardiorespiratory disorders. Recently, Aurora et al.22 examined portable sleep monitoring to diagnose sleep apnea in patients hospitalized for heart failure and concluded that sleep apnea can be accurately and efficiently identified by PM testing. Oliveira et al.2 evaluated the use of HSAT to diagnose OSA in 72 adults with COPD using the same study design as in our study, but with a different type 3 PM. The severity of COPD in their participants was similar to that in our participants. In that prior study, the high failure rates of 39% for HSAT and 22% for in-laboratory PM recordings did not support the routine use of HSAT for diagnosis of OSA in adults with COPD. In the 26 participants included in their analysis, the kappa coefficient of agreement of AHI 5–30 events/h (.35) and AHI ≥ 30 events/h (.56) were lower than the kappa coefficients for AHI ≥ 5 (.746) and AHI ≥ 30 (.642) in our study. Bland-Altman analysis of their PSG versus HSAT results revealed a mean difference of −0.06 events/h with limits of agreement of −30.7 to 29.5 events/h. We found similar limits of agreement between PSG and HSAT recordings, although our results suggested that HSAT underestimates AHI when compared to PSG, on average. Moreover, compared with the study by Oliveira et al., our study had a larger sample size and much lower failure rates of PM recording.

We found close agreement in the participants with COPD between automatic and manually edited scoring of AHI on the Nox-T3 studies. This agreement increases testing efficiency by reducing the amount of time needed to edit the automatically scored recording. Although current guidelines for HSAT strongly recommend manually edited scoring,9 the close agreement between automatic and manually edited scoring could potentially allow practitioners without sufficient resources for manual editing to rely on the automatic score for clinical management.

The study has some limitations. First, patients prescribed supplemental oxygen therapy were excluded in this study, which reduces the generalizability of our results. Second, the participants were recruited from patients with COPD referred to a sleep center, who likely had a high pretest probability of OSA. The results might differ in a community-based population of patients with COPD with a lower probability of OSA. Furthermore, type 3 PM devices do not evaluate CO2 level. In patients with COPD, CO2 retention may develop during sleep that would go undetected by HAST. Of particularly importance, it was not possible to determine which pulse oximeter provided the most accurate determination of overnight arterial blood oxygen saturation. Critically important clinical management decisions are made based on pulse oximetry recordings and home oxygen therapy is expensive. It is imperative that measurement of SpO2 be standardized across manufacturers.

In summary, this study validates and supports the use of the Nox-T3 monitor for HSAT to diagnose OSA in patients with COPD. Moderate agreement of AHI was present between PSG and HSAT, and even closer agreement was observed on the simultaneous in-laboratory PSG and PM recording. However, the lower oxygen saturation levels on Nox-T3 PM testing compared to PSG, likely due to differences among pulse oximeters, resulted in a greater proportion of patients with COPD without OSA achieving the criteria for being prescribed nocturnal oxygen treatment. Thus, PM may not be appropriate for determining the need for home oxygen therapy. Until pulse oximetry is standardized across manufacturers, clinicians need to be aware of the operating characteristics of the pulse oximeter they are using for diagnosis and patient management.

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. Work for this study was performed at Peking University People's Hospital, Crescenz VAMC, and University of Pennsylvania. Dr. Pack is The John L. Miclot Professor of Medicine at the University of Pennsylvania; funds for this endowment were provided by the Philips Respironics Foundation. The other authors report no conflicts of interest. STK was supported by NIH HL094307. FH was supported by research grants from Beijing Municipal Science & Technology Commission No. Z161100002616012. LX was supported by Young Elite Scientists Sponsorship Program of China Association for Science and Technology.

ABBREVIATIONS

AHI

apnea-hypopnea index

BMI

body mass index

CI

confidence interval

COPD

chronic obstructive pulmonary disease

CPAP

continuous positive airway pressure

HSAT

home sleep apnea testing

FEV1

forced expiratory volume in the first second

FVC

forced vital capacity

ODI

oxygen desaturation index

OSA

obstructive sleep apnea

PM

portable monitor

PPV

positive predictive value

PSG

polysomnography

REI

respiratory event index

SD

standard deviation

SpO2

oxygen saturation on pulse oximetry

ACKNOWLEDGMENTS

Author contributions: STK, AIP, and FH contributed to the study design; CY, STK, LX, WZ, YY, YZ, and RZ contributed to the data collection; STK, LX, and BTK contributed to the data analysis; EK contributed to coordinating this project; STK and LX are the guarantors of the manuscript and take responsibility for the integrity of the data; STK, LX, BTK, YC, FH, and AIP contributed to the writing and editing of the manuscript.

REFERENCES

1 

Kapur VK, Auckley DH, Chowdhuri S, et al. Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(3):479–504. [PubMed Central][PubMed]

2 

Oliveira MG, Nery LE, Santos-Silva R, et al. Is portable monitoring accurate in the diagnosis of obstructive sleep apnea syndrome in chronic pulmonary obstructive disease? Sleep Med. 2012;13(8):1033–1038. [PubMed]

3 

Flenley DC. Sleep in chronic obstructive lung disease. Clin Chest Med. 1985;6(4):651–661. [PubMed]

4 

Shteinberg M, Weiler-Ravel D, Adir Y. [The overlap syndrome: obstructive sleep apnea and chronic obstructive pulmonary disease]. Harefuah. 2009;148(5):333–336. 348[PubMed]

5 

Weitzenblum E, Krieger J, Apprill M, et al. Daytime pulmonary hypertension in patients with obstructive sleep apnea syndrome. Am Rev Respir Dis. 1988;138(2):345–349. [PubMed]

6 

Bradley TD, Rutherford R, Grossman RF, et al. Role of daytime hypoxemia in the pathogenesis of right heart failure in the obstructive sleep apnea syndrome. Am Rev Respir Dis. 1985;131(6):835–839. [PubMed]

7 

Marin JM, Soriano JB, Carrizo SJ, Boldova A, Celli BR. Outcomes in patients with chronic obstructive pulmonary disease and obstructive sleep apnea: the overlap syndrome. Am J Respir Crit Care Med. 2010;182(3):325–331. [PubMed]

8 

Kushida CA, Littner MR, Morgenthaler T, et al. Practice parameters for the indications for polysomnography and related procedures: an update for 2005. Sleep. 2005;28(4):499–521. [PubMed]

9 

Berry RB, Brooks R, Gamaldo CE, et al; for the American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Version 2.0. Darien, IL: American Academy of Sleep Medicine; 2012.

10 

Eberhard A, Calabrese P, Baconnier P, Benchetrit G. Comparison between the respiratory inductance plethysmography signal derivative and the airflow signal. Adv Exp Med Biol. 2001;499:489–494. [PubMed]

11 

Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1(8476):307–310. [PubMed]

12 

Bland JM, Altman DG. Comparing methods of measurement: why plotting difference against standard method is misleading. Lancet. 1995;346(8982):1085–1087. [PubMed]

13 

Centers for Medicare & Medicaid Services website. MLN Booklet: Home Oxygen Therapy [PDF]. https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/MLN-Publications-Items/ICN908804.html. Published October 2017. Accessed March 19, 2019.

14 

Xu L, Han F, Keenan BT, et al. Validation of the Nox-T3 portable monitor for diagnosis of obstructive sleep apnea in Chinese adults. J Clin Sleep Med. 2017;13(5):675–683. [PubMed Central][PubMed]

15 

Chediak AD, Acevedo-Crespo JC, Seiden DJ, Kim HH, Kiel MH. Nightly variability in the indices of sleep-disordered breathing in men being evaluated for impotence with consecutive night polysomnograms. Sleep. 1996;19(7):589–592. [PubMed]

16 

Lord S, Sawyer B, O'Connell D, et al. Night-to-night variability of disturbed breathing during sleep in an elderly community sample. Sleep. 1991;14(3):252–258. [PubMed]

17 

Zafar S, Ayappa I, Norman RG, Krieger AC, Walsleben JA, Rapoport DM. Choice of oximeter affects apnea-hypopnea index. Chest. 2005;127(1):80–88. [PubMed]

18 

Farre R, Montserrat JM, Ballester E, Hernandez L, Rotger M, Navajas D. Importance of the pulse oximeter averaging time when measuring oxygen desaturation in sleep apnea. Sleep. 1998;21(4):386–390. [PubMed]

19 

Davila DG, Richards KC, Marshall BL, et al. Oximeter's acquisition parameter influences the profile of respiratory disturbances. Sleep. 2003;26(1):91–95. [PubMed]

20 

Davila DG, Richards KC, Marshall BL, et al. Oximeter performance: the influence of acquisition parameters. Chest. 2002;122(5):1654–1660. [PubMed]

21 

Hannhart B, Michalski H, Delorme N, Chapparo G, Polu JM. Reliability of six pulse oximeters in chronic obstructive pulmonary disease. Chest. 1991;99(4):842–846. [PubMed]

22 

Aurora RN, Patil SP, Punjabi NM. Portable sleep monitoring for diagnosing sleep apnea in hospitalized patients with heart failure. Chest. 2018;154(1):91–98. [PubMed]


Supplemental Material

Supplemental Material

(.pdf | 737 KB)