This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Heitman, S. J. Articles by Flemons, W. W. Search for Related Content PubMed PubMed Citation Articles by Heitman, S. J. Articles by Flemons, W. W.

Validation of Nasal Pressure for the Identification of Apneas/Hypopneas during Sleep

Departments of Medicine and Sociology, Foothills Hospital and University of Calgary, Calgary, Alberta, Canada

Correspondence and requests for reprints should be addressed to W. Ward Flemons, Foothills Hospital/University of Calgary, 1403 29th Street NW, Calgary, Alberta, T2N 2T9 Canada. E-mail: flemons{at}ucalgary.ca

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
The reference standard for identifying apneas and hypopneas is a pneumotachograph, but using this can disrupt sleep. Nasal airflow estimation by measuring nasal pressure via nasal prongs is better tolerated by patients. However, nasal pressure has not been validated, using an event-by-event analysis, for detecting apneas/hypopneas during sleep. Eleven patients undergoing polysomnography wore a nasal mask capable of measuring nasal airflow (via pneumotachograph) and nasal pressure simultaneously. Each study was screened for respiratory disturbances, and from these 550 were randomly selected and blindly scored as an apnea/hypopnea or no event each using the pneumotachograph, nasal pressure, square root nasal pressure, and respiratory inductance sum signals independently. Agreement was measured using Cohen's kappa statistic. Intermeasurement agreements between the pneumotachograph and nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum were 0.76, 0.73, and 0.50, respectively. Inter- and intrarater agreements were, respectively, 0.68 and 0.60 for the pneumotachograph, 0.66 and 0.82 for nasal pressure, 0.61 and 0.78 for square root nasal pressure, and 0.47 and 0.76 for respiratory inductance plethysmography sum. These results indicate that nasal pressure has excellent agreement compared with a pneumotachograph and very good inter-/intrarater agreement. Square root transformation of the nasal pressure signal does not improve these levels of agreement, indicating that it is unnecessary in routine clinical practice for scoring apneas/hypopneas.

Key Words: sleep apnea syndromes • diagnostic techniques and procedures • evaluation studies

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
The reference standard for identifying and quantifying reductions in airflow during sleep is a pneumotachograph (1). Direct measurement of airflow using a pneumotachograph is seldom used in sleep laboratories because it requires patients to wear a tight-fitting facemask, which is uncomfortable and can disrupt sleep. Therefore, a number of surrogate measures have been developed to identify apneas and hypopneas during sleep. Respiratory inductance plethysmography has been used as a surrogate of tidal volume (2, 3), and thermistors (4) and nasal pressure (2, 5, 6) are often used as surrogates of airflow. Fluctuations in nasal pressure during inspiration and expiration reflect changes in inspiratory and expiratory airflow. Nasal prongs are well tolerated by patients, and the nasal pressure signal has excellent dynamic response (7). These factors make nasal pressure monitoring a promising new technique for detecting breathing events during sleep. However, nasal pressure has not been adequately validated to recommend it as a primary signal for detecting apneas/hypopneas (1).

During wakefulness, nasal pressure has been shown to measure minute ventilation with moderate accuracy compared with body plethysmography (2). Nasal pressure has better negative predictive value but poorer positive predictive value than respiratory inductance plethysmography for identifying hypopneas in awake subjects (2). Thurnheer and colleagues have recently shown reasonably close agreement between the apnea–hypopnea index calculated from a pneumotachograph and that from nasal pressure in 20 sleeping subjects (8). Their study also suggested excellent agreement between these two methods in the apnea/hypopnea scores when analyzed epoch-by-epoch (using 2.7-minute epochs) (8).

The identification and classification of breathing events such as apneas and hypopneas depends on human judgment. Therefore, proper validation of a new measurement technique requires demonstration of good agreement with a reference standard, as well as good inter- and intrarater agreement. This is done best using an event-by-event analysis and Cohen's statistic (1).

The relationship between nasal pressure and actual airflow is nonlinear. This may result in an underestimation of airflow and reduce event detection (false-negatives). If patients experience nasal obstruction or start mouth breathing, there may be false-positive identification of hypopneas. The square root of a nasal pressure signal provides a more linear estimate of actual airflow than nasal pressure alone (7, 8). Using an experimental apparatus with a flow generator to produce inspiratory waveforms, it has been shown that the square root of nasal pressure is more accurate than the untransformed nasal pressure signal for quantifying the reduction in flow amplitude during hypopneas and the degree of inspiratory flow limitation (9). Whether this transformation improves its inter- and intrarater agreement for identifying apneas and hypopneas during sleep in the clinical setting is not known.

Standard laboratory pressure transducers are expensive and require large amplifiers, which limits their use in both routine laboratory and ambulatory settings. Less expensive and smaller transducers are now available but have not been compared with the laboratory standards.

The objectives of the present study were the following: (1) to determine the intermeasurement agreement between the pneumotachograph and nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum for identifying apneas/hypopneas during sleep; (2) to compare the inter- and intrarater agreement of nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum with that of the pneumotachograph for identifying apneas/hypopneas during sleep; and (3) to compare a less expensive pressure transducer with a standard laboratory pressure transducer.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Eleven patients underwent overnight diagnostic polysomnography conducted in a standard fashion (10). Subjects wore a nasal mask designed to measure nasal airflow and nasal pressure simultaneously (Figure 1) . Nasal airflow was measured with a pneumotachograph (Fleisch, Lausanne, Switzerland) connected to a differential pressure transducer (Validyne, Northridge, CA) and nasal pressure was measured via nasal cannulae connected to a differential pressure transducer (Validyne). Nasal pressure was referenced to mask pressure. Square root nasal pressure was transformed from the nasal pressure signal, and respiratory inductance plethysmography sum was calculated through summation of chest and abdominal inductance bands (Respitrace; Ambulatory Monitoring, Ardsley, NY). The inductance bands were uncalibrated. To avoid heat and moisture build-up under the mask, cool air was circulated through the mask every 30 minutes for 5-minute cycles via an external port. Pneumotachograph and nasal pressure data were collected only during periods when this airflow was off. All signals were collected using Sandman NT software (Mallinckrodt Inc., St. Louis, MO). Sleep staging was performed according to standard criteria (11).

View larger version (131K): [in this window] [in a new window]   Figure 1. Experimental apparatus. The nasal mask was fitted to the subject's face. Nasal airflow was measured with a pneumotachograph (B) and estimated using nasal prongs (D) connected to a pressure transducer. Nasal pressure was measured with respect to mask pressure (E). Air was circulated through an external port (C) for 5-minute time intervals every 30 minutes to avoid heat and moisture build-up. A switching device was used to turn the airflow on and off (A). Data were collected only when the external airflow was switched off.

Protocol
Initial collection of events.
One sleep scorer screened each study in 2-minute epochs for potential breathing events, defined as discernable decreases ( 10%) from baseline in any of the pneumotachograph, nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum signals lasting not less than 5 seconds. The O₂ saturation and electroencephalogram (EEG) and electromyograph (EMG) signals were not used during the initial event collection.

Classification of events.
Breathing events were randomly selected using computer randomization and blindly scored by two scorers. Apneas and hypopneas were classified together and termed an event. An event was defined as periods of 10 seconds or more with either (1) a clear (> 50%) reduction in a pneumotachograph, nasal pressure, square root nasal pressure, or respiratory inductance plethysmography sum signal; or (2) a clear (but < 50%) reduction in a pneumotachograph, nasal pressure, square root nasal pressure, or respiratory inductance plethysmographys sum signal in association with either a decrease of more than 3% in O₂ saturation or an arousal (1). Arousals were identified using standard criteria (12). The software permitted each possible event to be independently and blindly scored as an event or no event using each of the pneumotachograph, nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum signals separately. This was accomplished by removing all signals from the computer screen except the signal in question and the EEG, EMG, and O₂ saturation signals such that arousals and O₂ desaturations could be identified. The order in which the events were scored was random. Only the 2-minute epoch containing the event in question was viewed on the screen during each rescoring run. Scorer 1 repeated the analysis twice. Scorers tabulated their results on separate data sheets and were thus blinded from each other's classification.

Comparison of standard laboratory and ambulatory pressure transducers.
Nasal pressure measured with a standard laboratory transducer (Validyne) was compared with that measured with an ambulatory device (Ultima Airflow Pressure Sensor; Braebon Medical Corp., Carp, ON, Canada). The standard laboratory and ambulatory transducers were calibrated to ± 3 cm H₂O, and DC amplification was used for both. Using a respirator (Harvard Apparatus, South Natick, MA), tidal volumes of 100, 300, 500, and 700 ml were delivered at 15, 20, and 25 breaths per minute, and pressures generated at each of these settings were measured using both devices. Data were sampled at 256 Hz.

Statistics
Intermeasurement agreement was assessed by comparing a scorer's classification of each event using nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum with their classifications using a pneumotachograph. Intermeasurement agreement was determined for both Scorer 1 and Scorer 2. Inter-rater agreement was determined by comparing Scorer 1's event classification using each device with Scorer 2's respective classifications. Intrarater agreement was determined by comparing Scorer 1's first classification of each event using each device to his second classification. Agreement for event scoring was calculated using Cohen's statistic (13), where a of 1.0 represents complete agreement and a of 0 indicates agreement no better than chance. A value of 0.6 or more was considered good agreement (14). Agreement between the standard laboratory and ambulatory transducers was calculated using the method described by Bland and Altman (15).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Eight male and three female subjects aged 32–72 were studied. The mean respiratory disturbance index (RDI) for the group was 48.6 (range 9.9–143). A total of 1,278 breathing events were identified; 550 of the events identified (50 per patient) were randomly selected for analysis. All 550 events were analyzed using a pneumotachograph, nasal pressure, and respiratory inductance plethysmography sum signals. During two of the studies, the square root transformation could not be performed due to technical difficulties; therefore, only 450 events were analyzed using square root nasal pressure.

Intermeasurement Agreement between a Pneumotachograph and Nasal Pressure, Square Root Nasal Pressure, and Respiratory Inductance Plethysmography Sum
Both nasal pressure and square root nasal pressure had very good intermeasurement agreement compared with the pneumotachograph for both scorers (Table 1) . The intermeasurement agreement of respiratory inductance plethysmography sum was worse ( < 0.6) for both scorers and the 95% confidence intervals did not overlap with those of nasal pressure (Table 1). The square root transformation of nasal pressure did not improve the intermeasurement agreement with the pneumotachograph compared with the untransformed signal (Table 1).

Nasal pressure, square root nasal pressure, and respiratory inductance plethysmography sum all had higher intrarater agreement than the pneumotachograph. However, the intrarater agreement of the pneumotachograph was still good, with a value of 0.6.

Overall, the pneumotachograph, nasal pressure, and square root nasal pressure had higher reliability than respiratory inductance plethysmography sum. Although respiratory inductance plethysmography sum had good intrarater agreement, it had only moderate inter-rater agreement. Results are summarized in Table 1.

Comparison of Standard Laboratory and Ambulatory Pressure Transducers
Figure 2 shows inspiratory and expiratory flow data measured by the standard laboratory and ambulatory transducers simultaneously. The differences in the pressures measured by the two transducers against the means of the two pressure measurements at 100-, 300-, 500-, and 700-ml tidal volumes at 20 breaths per minute are found in Figure 3 . The mean differences and limits of agreement ± the 95% confidence intervals at each flow setting are shown in Table 2 . The mean pressures generated at the 100-, 300-, 500-, and 700-ml tidal volume settings ranged between approximately ± 0.15, ± 0.5, ± 1.25, and ± 3 cm H₂O, respectively (Figure 3). The mean differences as percentages of these ranges in pressures were 18, 5.8, 2.0, and 1, respectively (Table 2, Figure 3). Except for the 100-ml tidal volume setting, where the pressures were the lowest, the differences in pressure are well contained within the upper and lower limits of agreement (Figure 3). The differences in pressure are greatest near 0 cm H₂O and beyond approximately ± 1.5 cm H₂O. The increased scatter of pressure differences over these pressure ranges illustrates this finding. Excluding those values near zero and beyond approximately ± 1.5 cm H₂O, the differences in pressure are small in relation to the mean pressure differences, reflecting a high level of agreement between the standard laboratory and ambulatory transducers (Figure 3).

View larger version (6K): [in this window] [in a new window]   Figure 2. Inspiratory and expiratory flow data measured by the standard laboratory (1) and ambulatory (2) transducers. Flow is shown at 15 (A), 20 (B), and 25 (C) breaths per minute at 500-ml tidal volumes.

View larger version (25K): [in this window] [in a new window]   Figure 3. Bland and Altman plots of difference in pressure against mean pressure for the standard laboratory and ambulatory transducers. The plots are for 20 breaths per minute at 100-ml (A), 300-ml (B), 500-ml (C), and 700-ml tidal volumes (D). The middle line in each plot is the mean difference, the upper line is the mean difference + 2 SD, and the lower line is the mean difference - 2 SD. Note the different scales on the y axis.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The process of validating a surrogate measurement to be used for breathing event recognition involves demonstrating that it has good agreement with a pneumotachograph and that it has good reliability (both inter- and intrarater) (1). This is done best through an event-by-event analysis using Cohen's statistic, which adjusts for the expected agreement due to chance (1, 13). We determined that nasal pressure had very good agreement with a pneumotachograph (average = 0.76). In addition, nasal pressure demonstrated very good inter-rater ( = 0.66) and intrarater agreement ( = 0.82).

Others have attempted to validate nasal pressure by comparing it with a pneumotachograph during sleep (8, 16), but not on an event-by-event basis. Nasal pressure has also been compared with thermistors and respiratory inductance plethysmography (2, 5, 17). Although these studies have often reported high levels of agreement, most have tended to use correlation to measure agreement. Correlation coefficients measure the strength of a relation between two variables, not the agreement between them. Variables that are well correlated do not necessarily agree. Correlation coefficients are not recommended for this purpose because changes in the scale of measurement, although having no effect on correlation, do affect agreement. In addition, if the range over which measurements are taken is wide, the correlation will be greater than if the range of measurements is narrow (15).

The American Academy of Sleep Medicine (AASM) Task Force suggested that differentiation of apneas from hypopneas was not necessary in clinical practice because both event types share a common pathophysiology and clinical consequences (1). However, separating apneas and hypopneas may be useful for research studies designed to compare their health consequences or to elucidate potential differences in pathogenic mechanisms. Because our study was designed to evaluate event scoring in clinical practice, we considered apneas and hypopneas a single entity and scored them visually rather than using an automated computer algorithm.

The AASM criteria for event recognition are based on reductions in peak signal amplitude. However, other signal characteristics such as the shape of the inspiratory curve, inspiratory time, and rib cage and abdomen asynchrony in the case of respiratory inductance plethysmography can be useful. We focused solely on changes in signal amplitude in the presence or absence of O₂ desaturations and arousals because it is common to all the signals used and is recommended by the AASM Task Force (1).

We attempted to select a broad range of event morphologies to be analyzed by studying subjects with a wide range of disease severity. The average RDI of subjects used in our study was 48.6. Although this would be generally considered high, the range of RDI was 9.9–143 and the median was 31.4. Three of the 11 patients had an RDI less than 20. During our initial collection of events, we included all events that had a discernable change ( 10%) in signal amplitude lasting a minimum of 5 seconds. Had we collected events of less than 5-second duration, this would have favored the no event classification by scorers; collecting only events lasting 10 seconds or longer would have favored true events. During the initial collection of events, all four signals were viewed simultaneously on the computer screen without using the EEG, EMG, or oxygen saturation channels so that our event selection was not biased. Between 40 and 45% of the events initially collected were classified as no event by the reference standard pneumotachograph. Therefore, it would appear that neither true nor false events were overselected. It was interesting that the pneumotachograph had the lowest intrarater agreement of the four signals tested. This may reflect the fact that event classification was not straightforward and that we had succeeded in collecting many events in the "gray zone."

Bias was also limited through our randomization and blinding process. We randomly selected 50 events per subject for analysis out of a total of 1,278 events for all subjects. The order in which these events were analyzed was also determined randomly. It has been our experience that breathing events frequently occur in clusters that are morphologically similar. Had we analyzed all the events from one subject in succession, it is possible that neighboring events might have influenced our event classification. Event scoring using each signal occurred independently and was completely blinded, as were the results obtained by each scorer.

One of the limitations of nasal pressure is false-positive detection of apneas/hypopneas due to nasal obstruction or mouth breathing. Reductions in airflow can occur through the nose, mouth, or both. It is possible that false-positive detection of apneas/hypopneas using the pneumotachograph or nasal pressure signals occurred in our study during transitions from nose or nose plus mouth breathing to periods of more predominant mouth breathing. Due to our choice of a nasal mask rather than a full facemask, we are unable to account for this potential problem. We elected to use a nasal mask because we felt that patients would be more likely to sleep, and a full facemask has the potential to alter airway mechanics by placing pressure on the lower jaw. Nevertheless, predominant mouth breathing appears to occur very rarely during sleep, and even when it occurs, there is often sufficient fluctuation in the nasal pressure signal to detect respiratory events (8, 18). Of the 550 events analyzed by Scorer 1, 334 (61%) were identified as true events using the pneumotachograph and 346 (63%) were identified as true events using respiratory inductance plethysmography sum. Although this does not eliminate mouth breathing as a potential confounder in our study, it is at least reassuring that the positive event detection rate using nasal flow was not significantly high compared with respiratory inductance plethysmography sum. Furthermore, during the initial event collection when all four signals were available, we did not identify lengthy periods of dampened nasal flow, which would have indicated possible mouth breathing.

Nasal pressure is known to have a nonlinear relationship with nasal airflow. The square root transformation creates a near-linear approximation of actual nasal airflow (7), which could make it more sensitive in identifying events in situations of low flow. However, we have shown in the present study that the square root transformation does not improve agreement with the pneumotachograph and does not improve the degree of inter- or intrarater agreement for the identification of apneas/hypopneas during sleep (Table 1). The square root of nasal pressure has been shown to be superior to untransformed nasal pressure for quantifying the reduction in flow amplitude during hypopneas (9), and it might also be important for measuring ventilation. Nevertheless, we have shown that the transformation does not improve the ability of nasal pressure to detect relative changes in breathing pattern and amplitude, which remains the crucial factor for clinical event identification. Our results are in keeping with those of Thurnheer and colleagues, who showed that despite the fact that the square root transformation of the nasal pressure signal provided a near linear relationship with airflow, it was not essential for estimating the apnea–hypopnea index (8).

There are a number of advantages to the use of nasal pressure over the pneumotachograph and other surrogate measures of flow. Nasal cannulae used to measure nasal pressure are well tolerated by patients, compared with a facemask. Thermistors have long time constants (19) and tend to underestimate changes in nasal flow. In contrast, nasal pressure has excellent dynamic response, which makes the interpretation of each different component within the breathing cycle more precise (7, 20). There is recent evidence to suggest that nasal pressure is capable of detecting airflow limitation (21–23). Identification of a plateau on the inspiratory waveform, which correlates with elevated upper airway resistance (23), may be useful in identifying patients with the upper airway resistance syndrome. The inter-rater agreement for identifying respiratory effort-related arousals is very good when compared with esophageal manometry (24).

When properly calibrated, the sum of the respiratory inductance plethysmography abdominal and chest signals provides a measure of tidal volume. However, displacement of the inductance bands can alter signal amplitude, resulting in loss of one or both of the chest and abdominal channels. It can also be difficult to obtain good thoracoabdominal signals in obese patients, many of whom have obstructive sleep apnea. Nasal pressure does not have these limitations. It has been recently recommended that a relative reduction of 50% from baseline in the uncalibrated respiratory inductance plethysmography sum can be used to define a hypopnea (1). This is based on the assumption that a 50% reduction in tidal volume is approximately equivalent to a 50% reduction in inspiratory flow. This assumption may not be valid in all circumstances and could partially explain our findings that respiratory inductance plethysmography sum was inferior to nasal pressure when compared with the pneumotachograph. Another possible reason why respiratory inductance plethysmography sum had inferior inter-rater agreement and intermeasurement agreement with the pneumotachograph is that we used an uncalibrated sum signal. This was done because it is the method most often used in clinical practice. In addition, had mouth breathing been a major factor in our study (we believe that it was not), one would have expected reduced agreement between the pneumotachograph and respiratory inductance plethysmography sum. However, this factor should not have altered the expected degree of inter-rater agreement, which was the lowest using the uncalibrated respiratory inductance plethysmography sum signal.

Nasal pressure is measured by connecting cannulae to a differential pressure transducer. The most widely used transducers are expensive, which may limit the use of this promising signal in many diagnostic sleep laboratories and in the increasingly popular ambulatory setting. Therefore, we compared a less expensive pressure transducer with a standard laboratory transducer. The standard laboratory transducer used in the present study is approximately 10 times as expensive as the ambulatory device. The transducers were calibrated to ± 3 cm H₂O, which is more than the expected range of pressure changes at the nares during quiet breathing. The flow tracings in Figure 2 and the Bland and Altman plots in Figure 3 demonstrate that the two transducers agree very well over a range of physiologic nasal pressures. The Bland and Altman plots suggest that the differences in pressures between the two transducers are greatest near 0 cm H₂O and beyond ± 1.5 cm H₂O. The relatively large differences around 0 cm H₂O occurred because the absolute pressure measurements were very small, so that small oscillations in pressure show as large differences. Poor agreement near 0 cm H₂O may indicate that nasal pressure alone would be inadequate for differentiating between apneas and hypopneas, but this is not necessary in routine clinical practice. Poor agreement beyond ± 1.5 cm H₂O is also not likely to be important as this falls outside the range of physiologic pressure changes during normal breathing and certainly during periods of reduced breathing. The ambulatory transducer was not used during the sleep studies that constitute the present experiment. However, given the results in Figures 2 and 3, we believe that the two transducers could have been used interchangeably in our study.

Following an event-by-event analysis in sleeping subjects using an appropriate statistical method, we conclude that nasal pressure is a valid device for identifying apneas/hypopneas during sleep. In the present study, all sensors were used simultaneously to gather potential breathing events such that an event-by-event analysis on both true and nonevents could be performed. Given the high level of agreement between nasal pressure and its square root with the pneumotachometer, we propose that these signals would be equally useful for detecting apneas/hypopneas from an unscored sleep study. Although the square root transformation improves the linearity of the nasal pressure signal (7), it is not essential for the purpose of clinical event recognition. Nasal pressure is well tolerated by patients, and with the advent of commercially available, portable, and more affordable pressure transducers, it has the potential to be used widely.

	Acknowledgments

 
The authors thank Amalu Iyer, Michele Ostrowski, and John Laprairie for helping with patient recruitment and data collection. They are also grateful to Drs. W. A. Whitelaw and J. E. Remmers for their suggestions regarding the study. Finally, the authors thank Braebon Medical Corporation for donating two of the Ultima Airflow Pressure Sensors.

Supported by grants from the Alberta Lung Association. The ambulatory pressure transducers were donated by Braebon Inc., ON, Canada.

Received in original form May 18, 2001; accepted in final form April 9, 2002

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 

1. Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. Sleep 1999;22:667–689.[Medline]
2. Berg S, Haight JS, Yap V, Hoffstein V, Cole P. Comparison of direct and indirect measurements of respiratory airflow: implications for hypopneas. Sleep 1997;20:60–64.[Medline]
3. Cantineau JP, Escourrou P, Sartene R, Gaultier C, Goldman M. Accuracy of respiratory inductive plethysmography during wakefulness and sleep in patients with obstructive sleep apnea. Chest 1992;102:1145–1151.[Abstract/Free Full Text]
4. Farre R, Montserrat JM, Rotger M, Ballester E, Navajas D. Accuracy of thermistors and thermocouples as flow-measuring devices for detecting hypopnoeas. Eur Respir J 1998;11:179–182.[Abstract/Free Full Text]
5. Gugger M. Comparison of ResMed AutoSet (version 3.03) with polysomnography in the diagnosis of the sleep apnoea/hypopnoea syndrome. Eur Respir J 1997;10:587–591.[Abstract]
6. Series F, Marc I. Nasal pressure recording in the diagnosis of sleep apnoea hypopnoea syndrome [see comments]. Thorax 1999;54:506–510.[Abstract/Free Full Text]
7. Montserrat JM, Farre R, Ballester E, Felez MA, Pasto M, Navajas D. Evaluation of nasal prongs for estimating nasal flow. Am J Respir Crit Care Med 1997;155:211–215.[Abstract]
8. Thurnheer R, Xiaobin X, Bloch KE. Accuracy of nasal cannula pressure recordings for assessment of ventilation during sleep. Am J Respir Crit Care Med 2001;164:1914–1919.[Abstract/Free Full Text]
9. Farre R, Rigau J, Montserrat JM, Ballester E, Navajas D. Relevance of linearizing nasal prongs for assessing hypopneas and flow limitation during sleep. Am J Respir Crit Care Med 2001;163:494–497.[Abstract/Free Full Text]
10. Flemons WW, Whitelaw WA, Brant R, Remmers JE. Likelihood ratios for a sleep apnea clinical prediction rule. Am J Respir Crit Care Med 1994;150:1279–1285.[Abstract]
11. Rechtschaffen A, Kales A. Manual of standardized terminology, technique, and scoring system for sleep stages of human subjects. Los Angeles, CA: NIH; Publication 204 BIS/BRL 1968.
12. EEG arousals: scoring rules and examples: a preliminary report from the Sleep Disorders Atlas Task Force of the American Sleep Disorders Association. Sleep 1992;15:173–184.[Medline]
13. Altman DG. Practical statistics for medical research. London: Chapman and Hall; 1991.
14. Fleiss JL. The measurement of inter-rater agreement. In: Fleiss JL, editor. Statistical methods for rates and proportions. New York: Wiley; 1981. p. 212–225.
15. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310.[CrossRef][Medline]
16. Fleury B, Rakotonanahary D, Hausser-Hauw C, Lebeau B, Guilleminault C. A laboratory validation study of the diagnostic mode of the Autoset system for sleep-related respiratory disorders. Sleep 1996;19: 502–505.[Medline]
17. Bradley PA, Mortimore IL, Douglas NJ. Comparison of polysomnography with ResCare Autoset in the diagnosis of the sleep apnoea/hypopnoea syndrome. Thorax 1995;50:1201–1203.[Abstract/Free Full Text]
18. Norman RG, Ahmed MM, Walsleben JA, Rapoport DM. Detection of respiratory events during NPSG: nasal cannula/pressure sensor versus thermistor. Sleep 1997;20:1175–1184.[Medline]
19. Xiong C, Sjoberg BJ, Sveider P, Ask P, Loyd D, Wranne B. Problems in timing of respiration with the nasal thermistor technique. J Am Soc Echocardiogr 1993;6:210–216.[Medline]
20. Ballester E, Badia JR, Hernandez L, Farre R, Navajas D, Montserrat JM. Nasal prongs in the detection of sleep-related disordered breathing in the sleep apnoea/hypopnoea syndrome. Eur Respir J 1998;11: 880–883.[Abstract]
21. Clark SA, Wilson CR, Satoh M, Pegelow D, Dempsey JA. Assessment of inspiratory flow limitation invasively and noninvasively during sleep. Am J Respir Crit Care Med 1998;158:713–722.[Abstract/Free Full Text]
22. Hosselet JJ, Norman RG, Ayappa I, Rapoport DM. Detection of flow limitation with a nasal cannula/pressure transducer system. Am J Respir Crit Care Med 1998;157:1461–1467.[Abstract/Free Full Text]
23. Condos R, Norman RG, Krishnasamy I, Peduzzi N, Goldring RM, Rapoport DM. Flow limitation as a noninvasive assessment of residual upper-airway resistance during continuous positive airway pressure therapy of obstructive sleep apnea. Am J Respir Crit Care Med 1994;150: 475–480.[Abstract]
24. Ayappa I, Norman RG, Krieger AC, Rosen A, O'Malley RL, Rapoport DM. Noninvasive detection of respiratory effort-related arousals (REras) by a nasal cannula/pressure transducer system. Sleep 2000;23: 763–771.[Medline]

This article has been cited by other articles:

				B. V. Vaughn and P. Giallanza Technical Review of Polysomnography Chest, December 1, 2008; 134(6): 1310 - 1319. [Abstract] [Full Text] [PDF]

				H. Muzumdar and R. Arens Diagnostic Issues in Pediatric Obstructive Sleep Apnea Proceedings of the ATS, February 15, 2008; 5(2): 263 - 273. [Abstract] [Full Text] [PDF]

				K. K. H. WONG, D. JANKELSON, A. REID, G. UNGER, G. DUNGAN, J. A. HEDNER, and R. R. GRUNSTEIN Diagnostic test evaluation of a nasal flow monitor for obstructive sleep apnea detection in sleep apnea research Behav Res Methods, February 1, 2008; 40(1): 360 - 366. [Abstract] [PDF]

				Y Dauvilliers, V Stal, B Abril, P Coubes, S Bobin, J Touchon, P Escourrou, F Parker, and P Bourgin Chiari malformation and sleep related breathing disorders J. Neurol. Neurosurg. Psychiatry, December 1, 2007; 78(12): 1344 - 1348. [Abstract] [Full Text] [PDF]

				A. I. Pack, G. Maislin, B. Staley, F. M. Pack, W. C. Rogers, C. F. P. George, and D. F. Dinges Impaired Performance in Commercial Drivers: Role of Sleep Apnea and Short Sleep Duration Am. J. Respir. Crit. Care Med., August 15, 2006; 174(4): 446 - 454. [Abstract] [Full Text] [PDF]

				C. S. Melendres, J. M. Lutz, E. D. Rubin, and C. L. Marcus Daytime Sleepiness and Hyperactivity in Children With Suspected Sleep-Disordered Breathing Pediatrics, September 1, 2004; 114(3): 768 - 775. [Abstract] [Full Text] [PDF]

				D. S. Hui, F. W. Ko, J. P. Fok, M. C. Chan, T. S. Li, B. Tomlinson, and G. Cheng The Effects of Nasal Continuous Positive Airway Pressure on Platelet Activation in Obstructive Sleep Apnea Syndrome Chest, May 1, 2004; 125(5): 1768 - 1775. [Abstract] [Full Text] [PDF]

				M. J. Tobin Sleep-Disordered Breathing, Control of Breathing, Respiratory Muscles, and Pulmonary Function Testing in AJRCCM 2002 Am. J. Respir. Crit. Care Med., February 1, 2003; 167(3): 306 - 318. [Full Text] [PDF]

				J. M. Montserrat and R. Farre Breathing Flow Disturbances during Sleep: Can They Be Accurately Assessed by Nasal Prongs? Am. J. Respir. Crit. Care Med., August 1, 2002; 166(3): 259 - 260. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Heitman, S. J. Articles by Flemons, W. W. Search for Related Content PubMed PubMed Citation Articles by Heitman, S. J. Articles by Flemons, W. W.