Chest ACCP Education Calendar
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lofaso, F.
Right arrow Articles by Louis, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lofaso, F.
Right arrow Articles by Louis, B.
(Chest. 2006;130:343-349.)
© 2006 American College of Chest Physicians

Bench Evaluation of Flow Limitation Detection by Automated Continuous Positive Airway Pressure Devices*

Frédéric Lofaso, MD, PhD; Gilbert Desmarais, PhD; Karl Leroux; Vincent Zalc; Redouane Fodil, PhD; Daniel Isabey, PhD and Bruno Louis, PhD

* From National Institute for Health and Medical Research (Institut National de la Santé et de la Recherche Médicale), Unit 651 (Drs. Fodil, Isabey, and Louis), Henri Mondor Teaching Hospital, Créteil; Assistance Publique des Hôpìtaux de Paris (Drs. Lofaso and Desmarais and Mr. Zalc), Raymond Poincaré Teaching Hospital, Physiology-Functional Testing, and Technological Innovations Centre, Garches, France; and People with Polio and Disabilities (Association d’Entraide des Polios et Handicapés) [Mr. Leroux], Puteaux, France.

Correspondence to: Frédéric Lofaso, MD, PhD, Service de Physiologie-Explorations Fonctionnelles, Hôpital Raymond Poincaré, 92380 Garches, France; e-mail: f.lofaso{at}rpc.ap-hop-paris.fr

Abstract

Study objective: Automatic continuous positive airway pressure (CPAP) devices that adjust the pressure delivered to the patient are now available to treat sleep-disordered breathing. Sophisticated auto-CPAP devices can detect and correct flattened inspiratory flow contours (FIFCs) associated with subtle upper airway obstruction. However, evaluations of their performance are made difficult by differences across patients and devices. We performed a bench study of five commercially available auto-CPAP devices using a breath waveform simulator to evaluate sensitivity for detecting flattened inspiratory flow.

Design: Five degrees of FIFC were simulated. In addition, normal and abnormal flow contours from patients published in the literature were evaluated.

Measurements and results: One device showed autotriggering leading to CPAP increases, and another device varied the CPAP level independently from the presence of an FIFC. The three remaining devices differed regarding the detection of FIFCs and the means used to increase CPAP.

Conclusion: Based on the characteristics of each patient, physicians must choose among devices with different thresholds of FIFC detection and different pressure responses to detection. Therefore, physicians need details on the algorithms used in auto-CPAP devices. Manufacturers should supply detailed algorithms.

Key Words: continuous positive airway pressure • flow limitation • obstructive sleep apnea

Continuous positive airway pressure (CPAP), which was introduced in 1981 by Sullivan et al,1 has considerably improved the treatment and prognosis of patients with obstructive sleep apnea syndrome. In practice, the optimal CPAP level is a tradeoff between pressure-related side effects and effectiveness in preventing upper airway obstruction during sleep.2 This optimal level is generally determined during a total or split-night sleep study. Follow-up is needed to verify that the selected level remains appropriate for the patient’s needs, since the minimal effective pressure can vary over time depending on body weight changes, sleep deprivation, nasal obstruction, and alcohol or hypnotic agent use.3 In addition, the minimal pressure can change during a given night according to body position and/or sleep stage.3

Greater effectiveness in relieving upper airway obstruction and improved acceptability of CPAP are being sought via the development of devices that use indirect noninvasive assessment of upper airway obstruction to continuously adjust the pressure around the minimal level that prevents abnormal breathing and arousal.4 Nearly all these auto-CPAP devices detect events such as apneas, hypopneas, and snoring. Bench studies have established that devices differ markedly regarding their sensitivity for detecting apnea and hypopnea,567 as well as snoring.8

There is general agreement that flattened inspiratory flow contours (FIFCs) predict subtle upper airway obstruction and should be eliminated by CPAP because they often herald snoring, hypopnea, apnea, or arousal.91011 In addition, a study12 in patients with obstructive sleep apnea syndrome showed that determining the CPAP level that eliminated FIFC was associated with better daytime alertness, independently from sleep quality or the apnea-hypopnea index while receiving CPAP. Oddly enough, only a few of the more sophisticated CPAP devices can detect FIFC (Table 1 ), and few studies567 of the effectiveness of FIFC detection are available. The ability to detect different degrees of FIFC has not been evaluated. A profile intermediate between a rounded (normal) and a squared (flattened) contour has been described as a suboptimal unstable condition during CPAP titration.910 Consequently, we conducted a bench study to examine the performance of five commercially available auto-CPAP devices in detecting various shapes of inspiratory flow contour (Table 1). The first part of the study assessed the sensitivity of the auto-CPAP devices for detecting various degrees of FIFC. In the second part of the study, we assessed the responses of the devices to inspiratory flow contours collected from the literature.1314


View this table:
[in this window]
[in a new window]

  		Table 1. The Five Auto-CPAP Devices Tested in the Bench Study

 
Materials and Methods

Experimental Setup
The auto-CPAP device was connected via a standard circuit to a two-chamber test lung (MII Vent Aid TTL; Michigan Instruments; Grand Rapids, MI) [Fig 1 ]. To simulate breathing cycles, the second chamber of the Michigan test lung (driving chamber) was connected to a flow-rate generator that could produce various waveforms previously stored in a microcomputer. This breath waveform simulator was developed in our laboratory. It relies on pressurized air, flow-rate measurement, and a servovalve. The simulator continuously adjusts the servovalve via a microcomputer in order to produce the desired flow rate. A small metal component allowed the driving chamber to displace air into the testing chamber, but not the opposite. Flow rate and pressure were measured at the pressurized chamber inlet. Flow rate was inferred using a pneumotachograph (Fleisch #2; Fleisch; Lausanne, Switzerland) connected to a differential pressure transducer (Validyne MP 45; Validyne Engineering; Northridge, CA; ± 3 cm H2O), and pressure was inferred using a pressure transducer (Validyne MP 45; Validyne Engineering; ± 35 cm H2O). Pressure and flow-rate signal outputs were digitized at 32 Hz (MP100; Biopac Systems; Goleta, CA) and recorded in a microcomputer for further analysis.


Figure 1
View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Experimental setup. The flow-rate generator simulates a long period of normal breathing followed by a period with flow limitation. This flow limitation is potentially detected by the CPAP device, whose response is recorded via flow rate (V) and pressure (P) measurements at the outlet of the pressurized test chamber.

 
Protocols
We studied an alternation of normal flow cycles delivered for 20 min and FIFC cycles delivered for 15 min. We voluntarily chose a long duration (15 min) of inspiratory flow limitation, although shorter durations are the rule in upper airway resistance syndrome, for three reasons: both long and short periods of flow limitation are detected by devices characterized by a short response time, using longer periods of flow limitation provides the opportunity to investigate devices with long response times, and longer periods of flow limitation allow an evaluation of device behavior when flow contour does not seem influenced by the first pressure increase.

We recorded the pressure responses of each of the five auto-CPAP devices. In the first experiment, the normal breathing flow had a rounded inspiratory flow contour with a frequency of eight cycles per minute, an inspiratory time of 1.9 s, and a tidal volume of 500 mL (Fig 2 , top, A). Five degrees of FIFC were simulated, namely, 90% of the maximum inspiratory flow rate (R90), 80% of the maximum inspiratory flow rate (R80), 70% of the maximum inspiratory flow rate (R70), 60% of the maximum inspiratory flow rate (R60), and 50% of the maximum inspiratory flow rate (R50) during the normal (rounded) flow cycle. To produce the FIFC, the normal signal was replaced by a plateau signal when the flow reached the desired percentage of the peak flow on the normal flow contour. Inspiratory time was set to keep the tidal volume at 500 mL. In the second set of experiments, we studied normal (rounded) and FIFC from patients with upper airway resistance syndrome published in the literature.1314 We used a normal breathing cycle as recorded by Guilleminault et al13 (NG) and a flow limitation cycle as recorded by Guilleminault (G1) [Fig 2, bottom, B], as well as a normal breathing cycle as recorded by Clark et al14 (NC), a false flow limitation cycle (FFLC), and two levels of flow limitation (C1 = first flow limitation cycle as recorded by Clark [C1], and second flow limitation cycle as recorded by Clark [C2]). In each set of experiments, each device was tested with CPAP set at 8 cm H2O and 12 cm H2O. The maximum pressure was 12 cm H2O for the 8 cm H2O CPAP level and 16 cm H2O for the 12 cm H2O level.


Figure 2
View larger version (40K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Top, A: Breathing flow rate cycles used during the first set of experiments (see text). R100 is the normal breathing cycle. R90, R80, R70, R60, and R50 are the five levels of flow limitation tested. Each normal flow period lasted 20 min, and each flow-limited period lasted 15 min, for a total of 195 min. Bottom, B: Breathing flow rate cycles used during the second set of experiments. NG and NC are two distinct normal breathing cycles issued from the recording of Guilleminault et al13 and Clark et al.14 FFLC, C1, and C2 are different levels of false and true flow limitations issued from the recording of Clark et al.14

 
Results

Major differences in sensitivity and behavior were found across the five auto-CPAP devices. Two devices (PV10i and Remstar Auto) exhibited unexpected responses with all the test protocols. With PV10i, after 5 min of constant pressure, autotriggering occurred, resulting in an CPAP increase that reached the maximum pressure within 340 s. The PV10i therefore delivered the maximum pressure at the beginning of the FIFC periods. With the Remstar Auto, the pressure increased by 1.4 cm H2O over 160 s then returned to the initial level within 60 s. This behavior recurred every 540 s, even during the FIFC periods. Figure 3 reports the findings with the other three devices. None of the devices detected all the simulated FIFCs. Overall, detection was less sensitive with the higher initial pressure setting. The Autoset Spirit failed to detect the C1 FIFC at the higher initial CPAP level. The Goodknight 420E device produced results similar to those of the Autoset Spirit. The Goodknight 418P was less sensitive than the other two devices at the higher CPAP level (Fig 3). The three devices differed in terms of the CPAP increase delivered when a FIFC was detected. The Autoset Spirit increased the CPAP level steadily, at a rate (1.3 cm H2O/min) similar to that observed by Teschler et al,15 up to the maximum pressure, which was maintained for the duration of flow limitation. The two other devices (Goodknight 418P and 420E) increased the CPAP level slightly, by a 0.3 cm H2O step. With these two devices, the greatest CPAP increase was consistently < 1 cm H2O, and the maximum pressure was never reached. Moreover, with these two devices, the CPAP level occasionally returned to its initial level before the end of the flow limitation.


Figure 3
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Flow limitation detected by the auto-CPAP devices during the two sets of experiments, with the initial CPAP level set at 8 cm H2O (left panels) and at 12 cm H2O (right panels). Data indicate the numbers of cycles after which the device increased the CPAP level.

 
Discussion

We compared the responses of several auto-CPAP devices to identical patterns of inspiratory flow contour, which is not possible in patients, given the variability in their breathing pattern disturbances. Our results show differences regarding both sensitivity for detecting FIFC and pressure responses to FIFC detection. The first set of experiments served to evaluate detection sensitivity and the second to assess responses to patients’ flow contours (collected from the literature) representing various degrees of flow limitation associated with upper airway resistance syndrome,1314 as well as normal flow.14

Comparison With Other Studies
Previous bench studies have evaluated the effectiveness of FIFC detection by selected CPAP devices. Farré et al5 observed that only the precursors of the Autoset Spirit were able to detect a square-like and U-shaped inspiratory flow contour with a tidal volume equal to 70% of the normal cycle and a flattening index (see below) between 0.01 and 0.012151617; whereas three devices, including Goodknight 418P and the Remstar Auto precursor investigated in our study, failed to detect these events. Abdendi et al6 observed that an Autoset Spirit precursor and the Goodknight 418P were able to detect some but not all FIFCs, whereas Remstar Auto and PV10i consistently missed FIFCs. Bliss et al7 reported that among four unspecified devices, only one detected FIFCs. Our study is in accordance with these data, with several devices failing to detect FIFCs, in contradiction with the claims of the manufacturers. In addition, we differentiated five degrees of FIFC and simulated flow contours described in the literature in order to obtain detailed information on device performance. None of the five devices detected all FIFCs. Autoset software for CPAP devices uses a flattening index of the inspiratory flow contour, described in the literature151617 as a measure of the root median square deviation from the normalized mean inspiratory flow rate over the middle 50% of inspiratory time. A value of zero would indicate a square wave flow-time profile. The manufacturers state that a normal awake value exceeds 0.25 U.151617 A flattening index value of 0.15 has been suggested as the critical threshold below which flow limitation becomes significant.151617 Table 2 gives the flattening index values corresponding to each of our contour flow simulations. Autoset Spirit was able to detect flow limitation to 80% of maximum (R80), corresponding to a flattening index of 0.18. Thus, our results with this device were not far from the threshold of 0.15 suggested in the literature.151617


View this table:
[in this window]
[in a new window]

  		Table 2. Flattening Index Used by the Software of the Autoset CPAP Devices of our Flow Contour Simulations*

 
Sensitivity for Detecting FIFCs
None of the five devices detected all FIFCs. One device (Remstar Auto) consistently failed to detect FIFC, probably because it was not able to detect the simulated normal breathing, so that the FIFC detection mechanism was not activated. Another device (PV10i) was probably excessively sensitive and therefore lacked specificity, as the maximum pressure level was determined before the delivery of FIFC. With the other three devices, sensitivity for detecting flattening decreased at the higher CPAP level. A similar result has been reported for the detection of snoring.8 At the higher CPAP level, a marked decrease in sensitivity occurred with the Goodknight 418P device. With all five devices, inspiratory flow is measured and characterized inside the machine. Therefore, flow measurements are done on the flow delivered by the device, which is the sum of the patient’s inspiratory flow and of the leakage through the calibrated hole placed in or near the mask to reduce rebreathing.18 An increase in the CPAP level is associated with increased leakage and, therefore, with a greater difference between the patient’s inspiratory flow signal and the flow delivered and measured by the auto-CPAP device. This may explain why flattening was less easily detected in our study at the higher CPAP level. Nevertheless, the effect of increasing the CPAP level varied across the devices, suggesting a role for the quality of the flow transducer signal and of the signal treatment process.

The Autoset Spirit showed the best sensitivity, without autotriggering during the long periods of simulated normal breathing. However, this result does not necessarily mean that the Autoset Spirit is the best device. The patient’s characteristics and the strategy chosen by the physician influence the performance of CPAP devices. Overlap occurs between normal individuals and patients with upper airway obstruction: normal individuals may exhibit nonrounded flow contours1920 detected by the software in Sullivan Autoset devices (ResMed),20 whereas patients with upper airway resistance syndrome do not consistently exhibit FIFCs during nonapneic/hypopneic obstructive events followed by arousal.21 Therefore, any auto-CPAP device may be either too sensitive or too specific for a given patient. The behavior of the patient/auto-CPAP tandem must be evaluated by a sleep study to check that the device is optimal for the patient’s characteristics and the physician’s objectives. For example, in patients who require a high level of CPAP but cannot tolerate it well, this high level could be used as the maximum pressure in order to reduce discomfort, provided the device has high sensitivity for detecting even moderately flattened contours. In contrast, when the required CPAP level is modest but tends to vary because of irregular alcohol use, changes in nasal patency, or upper airway edema, the relatively low level of CPAP determined under optimal conditions could be used as the minimal pressure; the pressure can be increased to a reasonable extent provided the device is moderately sensitive to flattening, so that it responds only to unquestionable flow limitation, in order to avoid the use of excessive pressure.

Behavior of the Devices After Detection of FIFCs
This study demonstrated marked differences in the behavior of CPAP devices in response to detected FIFCs. Contrary to the Autoset Spirit, the two Goodknight devices never reached the maximum pressure, and they returned to the initial pressure before the end of the FIFC period. Thus, these devices seem to interpret consistent inspiratory flow flattening occurring cycle after cycle as a normal breathing pattern rather than as a sign of recurrent arousal, as shown during slow-wave sleep.1921 However, it has been suggested that the optimal CPAP level in patients with obstructive sleep apnea syndrome is the level that avoids abnormal inspiratory effort,910 ie, that produces a rounded inspiratory flow contour.91012

The consequences of prolonged periods of flow limitation remain unknown.22 Findings reported by Meurice et al12 suggest a need for completely eliminating flow contour flattening in patients with persistent daytime sleepiness despite CPAP treatment. The apparent Goodknight algorithm (regular consistent flattening is interpreted as normal) needs to be examined in the light of this report. Therefore, further studies are required to determine which algorithm is optimal for sleep apnea patients receiving auto-CPAP.

Perspectives and Conclusion
With the newest generation of auto-CPAP devices, there will be an increased need for polysomnography and/or polygraphy during CPAP to identify the best auto-CPAP device for each individual patient and to determine the optimal settings. Future generations of auto-CPAP devices will probably be able to detect specific inspiratory flow patterns that characterize patient subsets20 or the specific flow pattern abnormality that characterizes a given patient. The choice of the device and of the settings will depend primarily on the patient’s characteristics and on the strategy chosen by the physician. Therefore, manufacturers must specify the characteristics of abnormal flow pattern detection and the algorithms used by the devices. Since the study of Farré et al5 showing that auto-CPAP algorithms were not disclosed by the manufacturers, little improvement has occurred in the information that manufacturers supply to physicians. The Autoset Spirit device probably used the Autoset software, which has been briefly described by Gugger et al,16 Berthon-Jones et al,17 and Teschler et al.15 Physicians should be able to change the threshold according to the flow pattern abnormality in the individual patient. Because no information was available on the algorithms used by the other auto-CPAP devices in our study, we cannot make hypotheses regarding the reasons for the differences seen across the five devices. Neither can we determine which device may be optimal for which type of patient.

Finally, the term flow limitation detection used by manufacturers to describe auto-CPAP devices masks differences in detection sensitivity and in responses to flow limitation. Therefore, the physician has an important role to play by choosing the best device for each patient. In the future, CPAP devices will probably become more sophisticated to allow greater customization and better performance. The manufacturers will have to provide detailed descriptions of the algorithms used by each device, to ensure that physicians can make the best choices for their patients.

Footnotes

Abbreviations: C1 = first flow limitation cycle as recorded by Clark; C2 = second flow limitation cycle as recorded by Clark; CPAP = continuous positive airway pressure; FFLC = false flow limitation cycle; FIFC = flattened inspiratory flow contour; G1 = flow limitation cycle as recorded by Guilleminault; NC = normal breathing cycle as recorded by Clark; NG = normal breathing cycle as recorded by Guilleminault; R50 = 50% of the maximum inspiratory flow rate; R60 = 60% of the maximum inspiratory flow rate; R70 = 70% of the maximum inspiratory flow rate; R80 = 80% of the maximum inspiratory flow rate; R90 = 90% of the maximum inspiratory flow rate

This study was supported by the Association d’Entraide des Polios et Handicapés.

All authors have reported to the ACCP that no significant conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Received for publication December 7, 2005. Accepted for publication February 21, 2006.

References

  1. Sullivan, CE, Issa, FG, Berthon-Jones, M, et al (1981) Reversal of obstructive sleep apnoea by continuous positive airway pressure applied through the nares. Lancet 1,862-865[ISI][Medline]
  2. American Thoracic Society.. Indications and standards for use of nasal continuous positive airway pressure (CPAP) in sleep apnea syndromes: official statement adopted March 1944. Am J Respir Crit Care Med 1994;150,1738-1745[ISI][Medline]
  3. Polo, O, Berthon-Jones, M, Douglas, NJ, et al Management of obstructive sleep apnoea/hypopnoea syndrome. Lancet 1994;344,656-660[CrossRef][ISI][Medline]
  4. Berthon-Jones, M Feasibility of a self-setting CPAP machine. Sleep 1993;16,S120-S121;discussion S121–S123[ISI][Medline]
  5. Farré, R, Montserrat, JM, Rigau, J, et al Response of automatic continuous positive airway pressure devices to different sleep breathing patterns: a bench study. Am J Respir Crit Care Med 2002;166,469-473[Abstract/Free Full Text]
  6. Abdenbi, F, Chambille, B, Escourrou, P Bench testing of auto-adjusting positive airway pressure devices. Eur Respir J 2004;24,649-658[Abstract/Free Full Text]
  7. Bliss, P, McCoy, R, Eiken, T Exploring auto-adjust nasal PAP. Sleep Rev 2003;,42-46
  8. Lofaso, F, Leroux, K, Quera-Salva, MA, et al Snoring detection during auto-nasal continuous positive airway pressure. Eur Respir J 2002;19,108-112[Abstract/Free Full Text]
  9. Condos, R, Norman, RG, Krishnasamy, I, et al 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]
  10. Montserrat, JM, Ballester, E, Olivi, H, et al Time-course of stepwise CPAP titration: behavior of respiratory and neurological variables. Am J Respir Crit Care Med 1995;152,1854-1859[Abstract]
  11. Ayappa, I, Norman, RG, Hosselet, JJ, et al Relative occurrence of flow limitation and snoring during continuous positive airway pressure titration. Chest 1998;114,685-690[Abstract/Free Full Text]
  12. Meurice, JC, Paquereau, J, Denjean, A, et al Influence of correction of flow limitation on continuous positive airway pressure efficiency in sleep apnoea/hypopnoea syndrome. Eur Respir J 1998;11,1121-1127[Abstract]
  13. Guilleminault, C, Stoohs, R, Duncan, S Snoring (I): Daytime sleepiness in regular heavy snorers. Chest 1991;99,40-48[Abstract/Free Full Text]
  14. Clark, SA, Wilson, CR, Satoh, M, et al Assessment of inspiratory flow limitation invasively and noninvasively during sleep. Am J Respir Crit Care Med 1998;158,713-722[Abstract/Free Full Text]
  15. Teschler, H, Berthon-Jones, M, Thompson, AB, et al Automated continuous positive airway pressure titration for obstructive sleep apnea syndrome. Am J Respir Crit Care Med 1996;154,734-740[Abstract]
  16. Gugger, M, Mathis, J, Bassetti, C Accuracy of an intelligent CPAP machine with in-built diagnostic abilities in detecting apnoeas: a comparison with polysomnography. Thorax 1995;50,1199-1201[Abstract]
  17. Berthon-Jones, M, Lawrence, S, Sullivan, CE, et al Nasal continuous positive airway pressure treatment: current realities and future. Sleep 1996;19,S131-S135[ISI][Medline]
  18. Lofaso, F, Brochard, L, Touchard, D, et al Evaluation of carbon dioxide rebreathing during pressure support ventilation with airway management system (BiPAP) devices. Chest 1995;108,772-778[Abstract/Free Full Text]
  19. Rees, K, Kingshott, RN, Wraith, PK, et al Frequency and significance of increased upper airway resistance during sleep. Am J Respir Crit Care Med 2000;162,1210-1214[Abstract/Free Full Text]
  20. Aittokallio, T, Saaresranta, T, Polo-Kantola, P, et al Analysis of inspiratory flow shapes in patients with partial upper-airway obstruction during sleep. Chest 2001;119,37-44[Abstract/Free Full Text]
  21. Guilleminault, C, Poyares, D, Palombini, L, et al Variability of respiratory effort in relation to sleep stages in normal controls and upper airway resistance syndrome patients. Sleep Med 2001;2,397-405[CrossRef][Medline]
  22. Thomas, RJ Flow-limitation and sleep: exploring the interface of a complex and dynamic system. Sleep Med 2001;2,375-377[CrossRef][Medline]



This article has been cited by other articles:


Home page
 ChestHome page
L. K. Brown
Autotitrating CPAP: How Shall We Judge Safety and Efficacy of a "Black Box"?
Chest, August 1, 2006; 130(2): 312 - 314.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lofaso, F.
Right arrow Articles by Louis, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lofaso, F.
Right arrow Articles by Louis, B.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS