Research ArticleOriginal Research

Effect of the Anatomic Reservoir on Low-Flow Oxygen Delivery Via Nasal Cannula: Constant Flow Versus Pulse Flow With Portable Oxygen Concentrator

Steven Zhou and Robert L Chatburn
Respiratory Care August 2014, 59 (8) 1199-1209; DOI: https://doi.org/10.4187/respcare.02878

Abstract

BACKGROUND: The FIO2 for a nasal cannula with constant flow (CF) depends on the anatomic reservoir (AR), which is affected by changes in frequency and end-expiratory flow. Conversely, pulse flow (PF) devices do not require the AR. The purpose of this study was to compare the FIO2 delivered by a nasal cannula supplied by CF via oxygen tank with that delivered by PF delivered via portable oxygen concentrator. Hypotheses were (1) a lung model of COPD with non-zero end-expiratory flow decreases FIO2 for CF more than for PF, and (2) CF and PF perform differently in terms of FIO2 delivery, despite having equivalent settings.

METHODS: Normal and COPD lung models were simulated (IngMar Medical ASL 5000) using published human data: normal: breathing frequency = 15 breaths/min, Rin = 4 cm H2O · s · L–1, Rout = 4 cm H2O · s · L–1, C = 60 mL · cm H2O–1, tidal volume (VT) = 685 mL, Pmax = 11.95 cm H2O, increase = 33%, and release = 28; COPD: breathing frequency = 20 breaths/min, Rin = 12 cm H2O · s · L–1, Rout = 25 cm H2O · s · L–1, C = 66 mL · cm H2O–1, VT = 685 mL, Pmax = 24.52 cm H2O, increase = 35%, and release = 23%. CF was 1–5 L/min. Portable oxygen concentrators used were Solo2 (Invacare), XPO2 (Invacare), FreeStyle (AirSep), Focus (AirSep), One G3 (Inogen), and LifeChoice ActivOx (Inova Labs).

RESULTS: CF produced significantly higher FIO2 at all settings for normal lungs but lower for COPD lungs compared with Solo2. COPD reduced the FIO2 for CF but had a smaller variable effect for PF. Data show there is no equivalency between PF setting and CF rates for the portable oxygen concentrators tested.

CONCLUSIONS: CF oxygen delivery via a nasal cannula is significantly reduced by elimination of the AR in a model of COPD, yielding clinically important decreases in FIO2. PF (delivered with a portable oxygen concentrator) is relatively unaffected. This study supports the recommendation that clinicians and caretakers should titrate the PF setting to each patient's unique oxygen requirements.

Introduction

Long-term oxygen therapy in the home is often delivered using a nasal cannula and either compressed or liquid oxygen sources. More recently, portable oxygen concentrators have been developed and marketed for patients' home use on a prescriptive basis. For these oxygen sources, gas may be delivered either as a constant flow (CF) or pulse flow (PF) (eg, pulsed-dose oxygen-conserving systems). PF devices are intended to either conserve oxygen (for tanks) or conserve battery energy (for portable oxygen concentrators).

The FIO2 during CF nasal cannula oxygen therapy is governed by the interaction of the patient's inspiratory flow pattern and the flow from the oxygen source (Fig. 1). The original mathematical model for this was described in a textbook by Shapiro1 (Table 1, column A). As noted by Boatright and Ward,2 this model is “… not clinically practical (for) calculating the … FIO2 delivered by the nasal oxygen cannula (but) is a useful learning tool.” Specifically, it is very useful for understanding the interactions of the variables that determine FIO2. Two important variables in this model are inspiratory time (TI) and anatomic reservoir (AR). TI is a function of breathing frequency and I-E ratio. As the frequency increases (due to either exercise or disease), TI decreases, which in turn decreases FIO2 (Table 1, column B). We will hereafter refer to this as the frequency effect.

Fig. 1.
Fig. 1.

Idealized waveforms for spontaneous breathing (blue) and constant flow nasal cannula (green) for patient with normal lungs. Areas between flow curves and time axis represent volumes.

View this table:
Table 1.

Simple Mathematical Model Showing the Variables Affecting FIO2 for a Low-Flow Cannula

The AR is the volume of oxygen stored in the upper airways (pharynx and nasal passage) due to flushing of the dead space by the cannula flow during the last portion of expiration. The AR is not a real physical space (ie, portion of the anatomic dead space) but rather an imaginary mathematical construct representing the volume of oxygen that may or may not exist in the anatomic dead space due to flushing by the cannula. If there is no flushing at end expiration (as with a pulse flow device) or if the flow from the cannula is opposed by the patient's expiratory flow, then the AR will not exist. Patients with COPD often have non-zero expiratory flows when the next inspiratory effort is made (ie, they have alveolar gas-trapping); that is, patients with COPD have long expiratory time constants and typically experience gas-trapping even during rest due to the fact that expiratory flow does not decay to zero before the next inspiratory effort. Thus, end-expiratory flow could prohibit CF devices from flushing away the upper airway dead space, prevent the AR of oxygen from accumulating, and thus decrease the FIO2 (Fig. 2). As shown in Table 1 (column C), eliminating the AR has an even larger effect on FIO2 than increasing frequency (in this case, decreasing FIO2 from 0.34 to 0.26). We will hereafter refer to this as the reservoir effect.

Fig. 2.
Fig. 2.

Idealized waveforms for spontaneous breathing (blue) and constant flow nasal cannula (green) for patient with COPD. Areas between flow curves and time axis represent volumes. Note that end-expiratory flow does not decay to zero before the next breath, eliminating the atomic reservoir and decreasing the useful inspired oxygen volume.

PF devices built into portable oxygen concentrators are most frequently designed to maintain a constant minute volume of oxygen delivered by cannula as the frequency increases (as opposed to some PF devices that maintain a constant pulse volume as frequency increases). Constant minute volume PF devices therefore should, like CF devices, show a decreasing FIO2 with increasing breathing frequency. For PF devices, FIO2 is largely determined by the pulse volume, and different portable oxygen concentrators deliver highly varied pulse volumes for the same setting (Table 2, compare columns A and B). Furthermore, portable oxygen concentrators do not deliver pure oxygen. Rather, the typical specification is 90 ± 3%. This has a small but noticeable effect on FIO2 (Table 2).

View this table:
Table 2.

Math Models for Pulse Flow Oxygen-Conserving Device

One important difference between CF and PF devices is that no AR is established when using PF devices that are triggered by the patient's inspiratory effort. This may have clinical relevance when delivering oxygen to patients with lung disease. The implication is that a disease condition such as COPD would decrease FIO2 from CF devices but not from PF devices.

The purpose of this study was to compare the FIO2 delivered with a nasal cannula attached either to CF of pure oxygen or to PF from a portable oxygen concentrator (constant minute volume design) generating ∼90% purity. The primary hypothesis was that a lung model of COPD disease state with non-zero end-expiratory flow (ie, gas-trapping) decreases FIO2 for CF more than for portable oxygen concentrators. A secondary hypothesis was that CF and portable oxygen concentrators perform differently in terms of FIO2 delivery despite having equivalent settings.

QUICK LOOK

Current knowledge

The delivered FIO2 during low-flow oxygen delivery via a nasal cannula is a function of the respiratory pattern and the presence of the anatomic reservoir (AR). During pulse dose oxygen delivery, the AR should not impact the delivered FIO2.

What this paper contributes to our knowledge

In a model system mimicking chronic obstructive lung disease, functional elimination of the AR due to the presence of flow at end expiration significantly reduced delivered oxygen via continuous low-flow oxygen by cannula. Pulse dose oxygen delivery was not impacted by the functional loss of the AR.

Methods

Six portable oxygen concentrators were evaluated in this study (Table 3). As the largest device with a weight of ∼20 pounds, the Solo2 portable oxygen concentrator (Invacare, Elyria, Ohio) is often compared separately from the other small portable oxygen concentrators that weigh ∼5 pounds. We decided to include it to illustrate the effects of the higher pulse volumes it is capable of delivering. Although the Solo2 can deliver both CF and PF, we limited it to PF for this study.

View this table:
Table 3.

Specifications for Portable Oxygen Concentrators Used in This Study From Manuals

As noted above, some manufacturers declare the portable oxygen concentrator settings to be equivalent to corresponding values of CF. Therefore, in this study, portable oxygen concentrator settings of 1–5 were compared with settings of CF values of 1–5 L/min from a compressed oxygen source. All portable oxygen concentrators were allowed to warm up for at least 5 min before beginning experimentation.

To determine whether these portable oxygen concentrators should be classified as constant minute volume devices (as opposed to constant pulse volume devices), we made measurements of pulse volume using an oxygen-conserving test system (Hans Rudolph) at two breathing frequencies (see lung model descriptions below). This device was also used to verify that the oxygen purity was within manufacturers' specifications.

Lung Models

This experiment was conducted using a nasal cannula (AirLife, CareFusion, San Diego, California) and artificial nares connected to an ASL 5000 lung simulator (IngMar Medical, Pittsburgh, Pennsylvania) as shown in Figure 3. Two lung models were created with the ASL 3.3 software: normal and COPD (Table 4). Lung models were created in the ASL software by (1) specifying the mechanical properties of the respiratory system, that is, resistance and compliance (Fig. 4), and (2) specifying the parameters of the mathematical function that represents the force (pressure) generated by the simulated respiratory muscle (Fig. 5). For these simulations, we used a sinusoidal muscle pressure function (Pmus) and specified the amplitude (Pmax), rate of increase, hold, and decrease in Pmus as percentages of the time for one cycle of inspiration and expiration (TTot). Figure 4 is a screenshot of the ASL 5000 software that was used to set these parameters.

Fig. 3.
Fig. 3.

Photo of experimental setup showing model nares connected to the IngMar ASL 5000 lung simulator.

View this table:
Table 4.

Lung Model Parameters Used With ASL 5000 Lung Simulator

Fig. 4.
Fig. 4.

Screen shot of the ASL 5000 lung simulator showing settings for lung mechanics.

Fig. 5.
Fig. 5.

Screen shot of ASL 5000 lung simulator software that allows setting of the parameters that define inspiratory effort during simulated breathing.

A normal patient was defined as being medically stable and devoid of ailments limiting pulmonary function. A COPD patient was defined as medically stable but diagnosed with COPD. Lung model parameters (eg, resistance, compliance, patient effort profile) were derived from previously published studies.

Normal Lung Model.

For the normal model, breathing frequency was found from the average of the data obtained by Gallagher et al3 and Im Hof et al,4 ∼13 breaths/min, which we arbitrarily rounded up to 15 breaths/min. Inspiratory resistance was taken from Younes and Riddle,5 who reported passive resistance of 4 cm H2O/L/s. Expiratory resistance was taken from Tobin6 and was assumed to be the same as normal patients: 4 cm H2O/L/s. Younes et al7 report an average passive elastance of 17 cm H2O/L, which was rounded to a compliance value of 60 mL/cm H2O. Data taken from Gallagher et al3 and Im Hof et al4 yielded an average tidal volume (VT) of 685 mL. Pmax (maximum muscle pressure amplitude generated by the simulated respiratory muscles) parameters were as follows; percent increase was found by averaging percent inspiration of data obtained by Gallagher et al3 and Im Hof et al4 using the equation TI/Ttot. TI was averaged to be 1.6 s, and Ttot was averaged to be 4.935 s, yielding an increase of ∼33%. Hold (delay between inspiration, expiration, and breaths) was based on analysis of proportional assist pressure waveforms, which showed a hold of 0% of TTot.8 Percent release was taken from analysis of proportional assist pressure8 waveforms, approximated to be 28% of Ttot. These model parameters are summarized in Table 4 and Figures 4 and 5.

For a constant Pmax setting, as resistance increases, peak flow of the lung model decreases, and hence VT decreases. Using the ASL 3.3 software, the Pmax setting was manually adjusted during each experimental run to maintain the target VT for the lung model (∼685 ± 5 mL).

COPD Lung Model.

For the COPD lung model frequency, O'Donnell et al9 reported an average breath cycle time of 3.1 s, which corresponded with a frequency of ∼19 breaths/min, rounded up to 20. Inspiratory and expiratory resistance values were taken from Mead et al.10 Kondili et al11 reported respiratory system compliance in slices (series of volume intervals for determining nonlinear dynamic respiratory system mechanics) ranging from 50 to 79 mL/cm H2O. The average of 5 slices when PEEP = 0 is ∼66 mL/cm H2O. Data for VT were taken from O'Donnell et al9 and found to be 690 mL.9 To maintain consistency between the two models, a common VT of 685 mL was chosen. The Pmax setting was manually adjusted during each experimental run to maintain the target VT for the lung model (∼685 ± 5 mL). Pmus waveform parameters were analyzed, yielding the following settings; O'Donnell et al9 reported that for unassisted resting patients with COPD, the average TI was 1.1 s, and the total cycle time was 3.1 s. Thus, the Pmus rise time was estimated to be 1.1/3.1 = 0.35 or 35%. Analysis of waveforms reported by Tassaux et al12 indicated a Pmus hold setting of 0% and a release setting of 23% at a breathing frequency of 16 breaths/min (which was the closest to a frequency of 20 breaths/min that we could find in the literature). These model parameters are summarized in Table 4 and Figures 4 and 5.

Data Analysis

The main outcome variable of this study was FIO2, as measured by the ASL 5000 lung simulator. Prior to the beginning of experimentation, the lung simulator O2 sensor was calibrated according to the manufacturer's instructions using compressed oxygen.

For each lung model (normal and COPD), data for 25 breaths were collected for each oxygen flow setting (after the oxygen concentration in the lung model stabilized, typically after 5 breaths). For CF, these included 1–5 L/min. For the portable oxygen concentrators, all numerical settings available were used. Measured FIO2 values for the last 10 of the 25 breaths of each oxygen setting were averaged, and the SD was calculated using the ASL 3.3 simulator software. Mean FIO2 values were compared using two-way analysis of variance (multiple comparisons were performed with the Holm-Sidak procedure) or Kruskal-Wallis analysis of variance on ranks (multiple-comparison procedures were performed using the Tukey test), with P < .05 considered significant. Mean FIO2 for the AirSep Focus included only one portable oxygen concentrator setting, so a t test was used.

Results

Classification of Portable Oxygen Concentrators

Figure 6 shows that all portable oxygen concentrators maintained an essentially constant minute volume of oxygen delivery via the cannula as frequency increased from 15 to 20 breaths/min. This means that they all decreased their pulse volumes as the frequency increased. In contrast, devices that maintain a constant pulse volume show a dramatic increase in minute volume as frequency increases.13

Fig. 6.
Fig. 6.

Results of pulse volume measurement and calculation of minute volume of oxygen delivery. All devices maintained virtually constant minute volume as frequency was changed from 15 breaths/min (normal lung model) to 20 breaths/min (COPD lung model).

Hypothesis 1: Effect of Disease State

Study data for both normal and COPD lung models (described in methods above) are summarized in Figure 7. For all experimental conditions, the resultant FIO2 values were lower for the COPD lung model than for the normal lung model (P < .001). The FIO2 difference increased as the CF or portable oxygen concentrator setting increased and was most pronounced at a flow of 5 L/min. However, if we define a clinically important difference as ≥ 2%, then the FIO2 for CF was decreased by COPD for all oxygen flows greater than 1 L/min. The only clinically important difference for portable oxygen concentrators was at the highest setting for the Solo2 (decrease of 4%).

Fig. 7.
Fig. 7.

Hypothesis 1: Graphs of measured FIO2 (expressed as percentage) during simulated breathing. Negative numbers above curves represent the difference in FIO2 (COPD lung model minus normal lung model) for the highest and lowest FIO2 values. POC = portable oxygen concentrator.

Hypothesis 2: Evaluation of CF and Portable Oxygen Concentrator Setting Equivalency

Our secondary hypothesis was that CF and portable oxygen concentrators perform differently in terms of FIO2 delivery despite having equivalent settings. Study data are reproduced in Figure 8 and are grouped by normal and COPD lung models (lung parameters described in methods). Equivalent settings produced different FIO2 values for both lung models (P < .001). Post hoc comparison results are shown in Table 5.

Fig. 8.
Fig. 8.

Hypothesis 2: Comparison of constant flow (CF) and portable oxygen concentrator (POC) settings for normal (A) and COPD lung models (B).

View this table:
Table 5.

Results for Hypothesis 2. Data are FIO2 for constant flow minus the FIO2 for the indicated portable oxygen concentrator.

Discussion

To our knowledge, this is the first published study to compare the effects of eliminating the AR (using a model of COPD) on oxygen delivery via a nasal cannula using either CF or PF. The data support the hypothesis that FIO2 in the COPD disease model is lower than in the normal model. Furthermore, FIO2 with CF was decreased more in the disease model than PF using portable oxygen concentrators. We speculate that this is because FIO2 with CF is reduced by two factors, the frequency effect and the reservoir effect, whereas PF is affected only by the frequency effect. We further speculate that PF oxygen-conserving devices that maintain a constant pulse volume as frequency increases (ie, increasing the minute volume delivery of oxygen) would be less affected or perhaps not affected at all by this COPD lung model.

This study also suggests that portable oxygen concentrator performance varies based on the minute volume of oxygen produced. In general, the higher the minute volume of oxygen, the higher the FIO2 achieved. All of these devices delivered the pulse volume in less than ∼ 0.6 s. Therefore, a reduction in TI caused by a small increase in breathing frequency (from 15 to 20 breaths/min) for our study lung models is not expected to waste oxygen by delivering it into the dead-space volume at the end of a breath.

Finally, our data show that there is no such thing as equivalency between portable oxygen concentrator setting and CF rate for the devices tested. Therefore, the numerical settings on these portable oxygen concentrators should be interpreted as merely indicating increased relative minute volumes levels (ie, the higher the number, the higher minute volume of oxygen delivered via the cannula). These observations emphasize the importance of physicians or caretakers titrating the portable oxygen concentrator setting to each patient's unique oxygen requirements.

Our data are consistent with other studies in humans. LeBlanc et al14 tested the ability of three portable oxygen concentrators to maintain SpO2 ≥ 90% during exercise in patients with COPD. The Eclipse 3 (chart SeQual Technologies, Ball Ground, Georgia), which delivers the largest oxygen bolus (pulse volume), yielded significantly higher SpO2 during the walking exercise. In a similar fashion, our results confirm that the larger the pulse volume, the higher the FIO2.

A major limitation of this study was that our results were entirely dependent on the parameters of the simulations we designed. However, we believe that this is the first study using the IngMar ASL 5000 lung simulator that described every parameter of the model including not just mechanics but also the settings for the Pmus waveform. Furthermore, these parameters were based on an analysis of actual human data published in the literature. Therefore, we believe that our study design is particularly strong in its ability to model real-world clinical conditions.

Another limitation was that our simulated dead space (volume of nares and tube connecting the nose to the ASL 5000) was relatively small (40 mL) but still large enough to form an AR. However, the aim of this study was not to formulate a predictive model for patient FIO2 but rather to offer insight into portable oxygen concentrator performance and the effects of normal and COPD disease states on FIO2. This study also offers insight into the relative effectiveness of CF and PF for treating COPD.

We speculate that the subject of our study has never been addressed because PF devices have historically been seen as valuable for their oxygen-conserving properties. Oxygen waste and conservation have played a significant role in the formulation of PF devices. In 1978, Auerbach et al15 developed the first intermittent-demand nasal cannula, which delivered flow throughout inspiration. For this device, oxygen flow began when negative pressure was detected and when positive pressure ceased. Tiep et al16 invented the first prototype for contemporary PF oxygen-conserving systems in 1985. This battery-operated device delivered 25 mL at the beginning of inspiration. Their study demonstrated that a PF device could provide the same oxygen saturation to patients using oxygen only at 230 mL/min compared with CF at 2 L/min.

Subsequent studies sought to examine the efficacy and effectiveness of these oxygen-conserving devices. Bower et al17 treated hypoxemic patients with a constant minute volume demand oxygen-conserving system during rest, exercise, and sleep. This study found that the system produced arterial oxygenation equivalent to CF, but with oxygen savings of up to 60%. Another study by Tiep et al18 found similar results. During treadmill exercise, hypoxemic patients achieved an equivalent arterial oxygen saturation (SaO2) of ∼90% using oxygen at 1,600 mL/min for steady flow (CF) but only 210 mL/min using the oxygen-conserving device.

In another exercise study, Senn et al19 established similar SpO2 values for patients at 25–30% maximum work load using both CF and an oxygen-conserving device, with oxygen compared at equivalent flow settings. However, there was no explanation of how these settings were derived (a common trend). This relationship between PF settings and CF was further explored by Carter et al,20 who compared COPD patient oxygenation using equivalent flow settings of CF and a demand flow device. At rest, a PF setting of 1 was found to be equivalent to CF at 1 L/min (SpO2), with an oxygen-conserving ratio of 4.6. During exercise, a PF setting of 4 resulted in higher SpO2 than a CF of 4 L/min (SpO2 92% vs 89%), with an oxygen-conserving ratio of 7.2.

Braun et al21 again compared COPD patient oxygenation using 5 different PF devices with a constant pulse volume or constant minute volume. At rest, PF was found to be equivalent to CF. During exercise, PF performed worse than CF (SpO2 89.6% vs 90.3%). Constant pulse volume also performed better than constant minute volume. This study concluded that oxygen-conserving devices vary in their ability to maintain SpO2 during exercise. Roberts et al22 found PF with a pulse volume of 35 mL with a pulse time of 200 ms to be equivalent to CF at 2 L/min. During the 6-min walk test, PF performed worse than CF (SpO2 78% vs 81%). PF was also found to be inferior in terms of patients' walking distance given supplemental oxygen. Fuhrman et al23 found significant differences in demand oxygen delivery system performance for COPD patients. At lower oxygen flows, all four devices improved SpO2 compared with CF. This study found that performance was better with PF at the start of inspiration. Palwai et al24 showed significant differences in patient oxygenation at a PF setting of 2, with only one device equaling oxygenation with CF at 2 L/min. Kerby et al25 examined the long-term mean SpO2 values of hospitalized patients with diseases requiring oxygen therapy. This study found that PF and CF yielded equivalent mean SpO2 values over the course of a day and a night.

In all of these studies, we see a consistent lack of information about how equivalent settings are derived. Perhaps it is not surprising then that the large variety of PF devices available today have a wide range of pulse volumes associated with the same setting values. Our study indicates one source of this variability, that is, patients with varying degrees of COPD-induced lung impairment (that variably affect the AR) will require different pulse volumes to attain the same oxygenation. Thus, studies using different kinds of patients will naturally derive different pulse volumes for the same settings.

In summary, when using a nasal cannula to deliver oxygen, the resultant FIO2 could be reduced if the AR is reduced by COPD disease state, depending on whether CF or PF is used. Specifically, CF is subject to both the frequency effect and the reservoir effect, whereas PF (with a constant minute volume device) is subject only to the frequency effect. To further complicate this issue, differing PF characteristics (eg, pulse volume and pulse time) among different devices with the same settings would increase the variability of the results. Settings on PF devices are whole numbers that indicate increasing pulse volumes with increasing setting number (and presumably higher FIO2 delivery). Some manufacturers go so far as to claim that the setting numbers are equivalent to CF values expressed in L/min (eg, a setting of 2 on a portable oxygen concentrator is supposed to deliver the same inspired oxygen as CF at 2 L/min). Although this notion has often been refuted in the scientific literature, it persists in marketing literature and promotional materials.26 This can have dangerous repercussions on oxygen therapy, as clinicians and patients may believe that higher oxygen flows are delivered than in reality.

Conclusion

Our study shows that oxygen delivery via a nasal cannula is affected not only by the type of device used, that is, constant versus pulse flow, but also by the presence of end-expiratory flow (and hence absence of the AR), as in the case of COPD. As such, our data may help to explain differences in device performance found in human studies. Our data also show that there is no such thing as equivalency between portable oxygen concentrator setting and CF rate for the devices tested. Therefore, the numerical settings on these portable oxygen concentrators should be interpreted as merely indicating increased relative minute volume levels.

Footnotes

  • Correspondence: Robert L Chatburn MHHS RRT-NPS FAARC, Respiratory Institute, M-56, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, Ohio 44195. E-mail: chatbur{at}ccf.org.

  • Mr Chatburn has disclosed relationships with Philips Respironics, Inogen, Invacare, Hamilton, Covidien, Draeger, CareFusion, Newport, IngMar Medical, and Radiometer America Breathe. Mr Zhou has disclosed no conflicts of interest.

REFERENCES

  1. 1.↵
    1. Shapiro BA
    . Clinical application of blood gases. Chicago: Year Book Medical Publishers; 1973;105-112.
  2. 2.↵
    1. Hess DR,
    2. MacIntyre NR,
    3. Mishoe SC,
    4. Galvin WF,
    5. Adams AB
    1. Boatright J,
    2. Ward J
    . Therapeutic gases: management and administration. In Hess DR, MacIntyre NR, Mishoe SC, Galvin WF, Adams AB, editors. Respiratory care: principles and practice, 2nd edition. Sudbury: Jones & Bartlett Learning; 2012:282.
  3. 3.↵
    1. Gallagher CG,
    2. Sanii R,
    3. Younes M
    . Response of normal subjects to inspiratory resistive unloading. J Appl Physiol 1989;66(3):1113-1119.
    OpenUrlAbstract/FREE Full Text
  4. 4.↵
    1. Im Hof V,
    2. West P,
    3. Younes M
    . Steady-state response of normal subjects to inspiratory resistive load. J Appl Physiol 1986;60(5):1471-1481.
    OpenUrlAbstract/FREE Full Text
  5. 5.↵
    1. Younes M,
    2. Riddle W
    . Relation between respiratory neural output and tidal volume. J Appl Physiol 1984;56(4):1110-1119.
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    1. Tobin MJ
    . Principles and practice of mechanical ventilation, 3rd edition. New York: McGraw-Hill Professional; 2012.
  7. 7.↵
    1. Younes M,
    2. Riddle W,
    3. Polacheck J
    . A model for the relation between respiratory neural and mechanical outputs. III. Validation. J Appl Physiol 1981;51(4):990-1001.
    OpenUrlAbstract/FREE Full Text
  8. 8.↵
    1. Younes M,
    2. Puddy A,
    3. Roberts D,
    4. Light RB,
    5. Quesada A,
    6. Taylor K,
    7. et al
    . Proportional assist ventilation. Results of an initial clinical trial. Am Rev Respir Dis 1992;145:121-129.
    OpenUrlPubMed
  9. 9.↵
    1. O'Donnell DE,
    2. Sanii R,
    3. Anthonisen NR,
    4. Younes M
    . Expiratory resistive loading in patients with severe chronic air-flow limitation. An evaluation of ventilatory mechanics and compensatory responses. Am Rev Respir Dis 1987;136(1):102-107.
    OpenUrlPubMed
  10. 10.↵
    1. Mead J,
    2. Lindgren I,
    3. Gaensler EA
    . The mechanical properties of the lungs in emphysema. J Clin Invest 1955;34(7 Part 1):1005-1016.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Kondili E,
    2. Alexopoulou C,
    3. Prinianakis G,
    4. Xirouchaki N,
    5. Georgopoulos D
    . Pattern of lung emptying and expiratory resistance in mechanically ventilated patients with chronic obstructive pulmonary disease. Intensive Care Med 2004;30(7):1311-1318.
    OpenUrlPubMed
  12. 12.↵
    1. Tassaux D,
    2. Gainnier M,
    3. Battisti A,
    4. Jolliet P
    . Impact of expiratory trigger setting on delayed cycling and inspiratory muscle workload. Am J Respir Crit Care Med 2005;172(10):1283-1289.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Chatburn RL,
    2. Williams TJ
    . Performance comparison of 4 portable oxygen concentrators. Respir Care 2010;55(4):433-442.
    OpenUrlAbstract/FREE Full Text
  14. 14.↵
    1. Leblanc CJ,
    2. Lavallée LG,
    3. King JA,
    4. Taylor-Sussex RE,
    5. Woolnough A,
    6. McKim DA
    . A comparative study of 3 portable oxygen concentrators during a 6-minute walk test in patients with chronic lung disease. Respir Care 2013;58(10):1598-1605.
    OpenUrlAbstract/FREE Full Text
  15. 15.↵
    1. Auerbach D,
    2. Flick MR,
    3. Block AJ
    . A new oxygen cannula system using intermittent-demand nasal flow. Chest 1978;74(1):39-44.
    OpenUrlCrossRefPubMed
  16. 16.↵
    1. Tiep B,
    2. Nicotra MB,
    3. Carter R,
    4. Phillips R,
    5. Otsap B
    . Low-concentration oxygen therapy via a demand oxygen delivery system. Chest 1985;87(5):636-638.
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Bower JS,
    2. Brook CJ,
    3. Zimmer K,
    4. Davis D
    . Performance of demand oxygen saver system during rest, exercise, and sleep in hypoxemic patients. Chest 1988;94(1):77-80.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Tiep BL,
    2. Carter R,
    3. Nicotra B,
    4. Berry J,
    5. Phillips RE,
    6. Otsap B
    . Demand oxygen delivery during exercise. Chest 1987;91(1):15-20.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Senn S,
    2. Wanger J,
    3. Fernandez E,
    4. Cherniack RM
    . Efficacy of a pulsed oxygen delivery device during exercise in patients with chronic respiratory disease. Chest 1989;96(3):467-472.
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Carter R,
    2. Tashkin D,
    3. Djahed B,
    4. Hathaway E,
    5. Nicotra MB,
    6. Tiep BL
    . Demand oxygen delivery for patients with restrictive lung disease. Chest 1989;96(6):1307-1311.
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Braun SR,
    2. Spratt G,
    3. Scott GC,
    4. Ellersieck M
    . Comparison of six oxygen delivery systems for COPD patients at rest and during exercise. Chest 1992;102(3):694-698.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Roberts CM,
    2. Bell J,
    3. Wedzicha JA
    . Comparison of the efficacy of a demand oxygen delivery system with continuous low flow oxygen in subjects with stable COPD and severe oxygen desaturation on walking. Thorax 1996;51(8):831-834.
    OpenUrlAbstract/FREE Full Text
  23. 23.↵
    1. Fuhrman C,
    2. Chouaid C,
    3. Herigault R,
    4. Housset B,
    5. Adnot S
    . Comparison of four demand oxygen delivery systems at rest and during exercise for chronic obstructive pulmonary disease. Respir Med 2004;98(10):938-944.
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Palwai A,
    2. Skowronski M,
    3. Coreno A,
    4. Drummond C,
    5. McFadden ER Jr.
    . Critical comparisons of the clinical performance of oxygen-conserving devices. Am J Respir Crit Care Med 2010;181(10):1061-1071.
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Kerby GR,
    2. O'Donohue WJ,
    3. Romberger DJ,
    4. Hanson FN,
    5. Koenig GA
    . Clinical efficacy and cost benefit of pulse flow oxygen in hospitalized patients. Chest 1990;97(2):369-672.
    OpenUrlCrossRefPubMed
  26. 26.↵
    1. McCoy RW
    . Options for home oxygen therapy equipment: storage and metering of oxygen in the home. Respir Care 2013;58(1):65-85. Hypothesis 2: Comparison of constant flow and portable oxygen concentrator (POC) settings for normal (A) and COPD (B) lung models.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

 

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section

Related Articles

Cited By...

Keywords