Research ArticleOriginal Research

Inspiratory Limb Carbon Dioxide Entrainment During High-Frequency Oscillatory Ventilation: Characterization in a Mechanical Test Lung and Swine Model

Adam W Bostick, Gregory A Naworol, Tyler J Britton, Timothy R Ori, Shawn K French and Stephen Derdak
Respiratory Care November 2012, 57 (11) 1865-1872; DOI: https://doi.org/10.4187/respcare.01563

Abstract

BACKGROUND: High-frequency oscillatory ventilation (HFOV) has been utilized as a rescue oxygenation therapy in adults with ARDS over the last decade. The HFOV oscillating piston can generate negative pressure during the exhalation cycle, which has been termed active exhalation. We hypothesized that this characteristic of HFOV entrains CO2 into the inspiratory limb of the circuit and increases the total dead space. The purpose of this study was to determine if retrograde CO2 entrainment occurs and how it is altered by HFOV parameter settings.

METHODS: An HFOV was interfaced to a cuffed endotracheal tube and connected to a mechanical test lung. Negative pressure changes within the circuit's inspiratory limb were measured while HFOV settings were manipulated. Retrograde CO2 entrainment was evaluated by insufflating CO2 into the test lung to achieve 40 mm Hg at the carina. Inspiratory limb CO2 entrainment was measured at incremental distances from the Y-piece. HFOV settings and cuff leak were varied to assess their effect on CO2 entrainment. Control experiments were conducted using a conventional ventilator. Test lung results were validated on a large hypercapnic swine.

RESULTS: Negative pressure was detectable within the inspiratory limb of the HFOV circuit and varied inversely with mean airway pressure (P̄aw) and directly with oscillatory pressure amplitude (ΔP). CO2 was readily detectable within the inspiratory limb and was proportional to the negative pressure that was generated. Factors that decreased CO2 entrainment in both the test lung and swine included low ΔP, high mean airway pressure, high oscillatory frequency (Hz), high bias flow, and endotracheal tube cuff leak placement. CO2 entrainment was also reduced by utilizing a higher bias flow strategy at any targeted mean airway pressure.

CONCLUSIONS: Retrograde CO2 entrainment occurs during HFOV use and can be manipulated with the ventilator settings. This phenomenon may have clinical implications on the development or persistence of hypercapnia.

Introduction

Over the last decade, high-frequency oscillatory ventilation (HFOV) has been utilized as a rescue oxygenation therapy for adults with severe ARDS.1,2 The use of HFOV has consistently shown short-term improvement in oxygenation parameters, which is attributed to the use of higher mean airway pressure (P̄aw).1,3–5 Furthermore, HFOV is associated with reduced rates of ventilator-induced lung injury in small and large animal models, when compared to conventional ventilation.3,4,6

Recommendations from an expert discussion group in 2007 addressed the issue of balancing acceptable blood gas values with lung-protective ventilation during the use of HFOV in adults.7 The goal was to create a strategy that would provide adequate oxygenation and CO2 clearance, while minimizing injurious stresses on the lung.

Specifically regarding CO2 clearance, it was recommended that oscillatory pressure amplitude (ΔP) remain at 90 cm H2O, and that an arterial pH of 7.25–7.35 be achieved by adjusting respiratory frequency in Hertz (Hz) alone. To promote lung protection by minimizing delivered tidal volume (VT), it is suggested that respiratory frequency be increased to the highest feasible rate that would obtain the targeted pH. In contrast to conventional ventilation, VT is altered during HFOV with changes in respiratory frequency.8–10 At low respiratory frequency and high ΔP, VT can exceed the volume of anatomical dead space in smaller patients, and may contribute to ventilator-induced lung injury. Of note, in patients requiring a respiratory frequency ≤ 7 Hz the placement of an endotracheal tube (ETT) cuff leak is recommended to facilitate CO2 clearance, which might avoid further increases in delivered VT. An observational study reported that 30% of patients on HFOV for the treatment of ARDS required a cuff leak to achieve the suggested pH target; however, this will vary between institutions based on their individual protocols.11

During the exhalation cycle of the 3100B HFOV (SensorMedics/CareFusion, San Diego, California), the oscillating piston can generate negative pressure relative to the mean airway pressure. This property of HFOV is often referred to as “active exhalation”; however, its contribution to expiratory flow and gas exchange has not been well characterized. The 3100B HFOV utilizes bias flow in an attempt to direct all expired gases into the expiratory limb of the circuit. Its efficiency, however, to compensate for the negative pressures generated by the oscillating piston is unknown. Because of these characteristics, we hypothesized that negative pressure generated by the oscillating piston within the inspiratory limb of the 3100B HFOV circuit promotes the entrainment of expired CO2. Retrograde CO2 entrainment, if present, would increase the total dead space and could contribute to hypercapnia.

The purpose of this research was to determine if retrograde CO2 entrainment into the inspiratory limb of the 3100B HFOV occurs, and to characterize how CO2 entrainment is effected by manipulations of the different ventilator parameters.

QUICK LOOK

Current knowledge

High-frequency oscillation is used to treat refractory hypoxemia across a range of patients, from neonates to adults. A unique characteristic of high-frequency oscillation is known as active exhalation, created by the negative displacement of a diaphragm or piston.

What this paper contributes to our knowledge

Retrograde carbon dioxide movement into the patient circuit is common during high-frequency oscillation and may exacerbate hypercarbia. Carbon dioxide rebreathing can be reduced by reducing ΔP; by using a higher mean airway pressure, higher oscillatory frequency, and higher bias flow; and by the presence of an endotracheal tube cuff leak.

Methods

HFOV Inspiratory Limb Pressure Measurements

A SensorMedics 3100B HFOV was interfaced to a cuffed 8.0 mm ETT positioned within an artificial trachea (Fig. 1). The artificial trachea was attached to a test lung (5600i, Michigan Instruments, Grand Rapids, Michigan) set with a compliance of 0.02 L/cm H2O (approximating ARDS compliance). A pressure transducer (TruWave, Edwards Lifesciences, Irvine, California) with a maximum frequency response of 200 Hz was adapted to fit the temperature port on the inspiratory limb of the HFOV circuit, located 2.5 cm from the Y-piece. The pressure transducer was zeroed relative to atmospheric pressure. Pressure tracings were displayed on a hemodynamic monitor (Solar 8000i, General Electric, Fairfield, Connecticut). Peak positive and negative pressure values were identified with the pressure display cursor and recorded while adjusting P̄aw or ΔP.

Fig. 1.
Fig. 1.

Mechanical lung design. A pressure transducer was fitted 2.5 cm from the Y-piece during pressure measurements. During retrograde CO2 entrainment measurements, the pressure transducer was removed and the gas analyzing line was inserted at 76 cm from the Y-piece, and positioned at locations of interest during the different experiments. Y = Y-piece. ETT = endotracheal tube.

HFOV settings during the P̄aw experiment were as follows: ΔP 90 cm H2O, bias flow 30 L/min, inspiratory time (I-time) 33%, frequency 7 Hz. HFOV settings during the ΔP experiment were as follows: P̄aw 34 cm H2O, bias flow 30 L/min, I-time 33%, frequency 7 Hz.

CO2 Entrainment During HFOV in Mechanical Lung Model

The 3100B HFOV and mechanical test lung were set up as previously described (with the exception of the pressure transducer). A gas analyzer (RGM 5250, Datex-Ohmeda, Madison, Wisconsin) with a sample flow rate of 190 ± 40 mL/min, response time ≤ 400 ms at a flow of 200 mL/min) was utilized to measure inspiratory limb CO2 entrainment. The gas analyzing line was inserted at a point 76 cm from the Y-piece within the inspiratory limb of the HFOV circuit, and then positioned at the carina. CO2 was insufflated (7% CO2 from a size H tank) into the test lung at a flow of 0.5 L/min to attain 40 mm Hg CO2 at the carina. The CO2 insertion point into the system was located between the artificial trachea and the carina. The settings of the 3100B HFOV during this calibration period were: P̄aw 34 cm H2O, bias flow 30 L/min, I-time 33%, frequency 7 Hz, ΔP 90 cm H2O, FIO2 0.21. The settings used to calibrate CO2 insufflation also served as the baseline settings for each of the subsequent experiments. Gas sampling at the carina was performed prior to initiation of all experiments to confirm that carinal CO2 was 40 mm Hg. Insufflated CO2 flow did not require adjustment between experiments and was not altered during experiments in order to mimic steady state CO2 production.

The gas analyzer was withdrawn back from the carina, into the inspiratory limb of the HFOV circuit, to a maximum of 89 cm from the Y-piece, confirming the presence of retrograde CO2 entrainment. After confirming CO2 was detectable within the inspiratory limb, the gas analyzer was placed at 51 cm from the Y-piece position and the different parameters (P̄aw, bias flow, respiratory frequency, ΔP) of the 3100B HFOV were independently manipulated to assess each setting's effect on CO2 entrainment. Evaluation of CO2 entrainment was also performed while simultaneously increasing bias flow and manipulating the mean pressure adjustment to maintain a constant P̄aw of 34 cm H2O. All experiments were performed with and without a 5 cm H2O cuff leak. Cuff leaks were created by increasing bias flow to raise P̄aw by 5 cm H2O, followed by removal of air from the ETT pilot balloon until the original P̄aw was achieved.

CO2 Entrainment During Conventional Ventilation in Mechanical Lung Model

Conventional ventilator experiments with a ventilator (Servo-i, Maquet, Bridgewater, New Jersey) were performed to validate the experimental design. Two modes were utilized: volume control continuous mandatory ventilation (VC-CMV) and Bi-Vent (pressure controlled ventilation on 2 independently adjustable levels allowing unrestricted spontaneous breathing on both levels). CO2 insufflation was performed during these control experiments by 2 different methods: 40 mm Hg CO2 (measured at end-exhalation) was attained at the carina prior to manipulating the ventilator parameters (CO2 flow of 0.1 L/min); and CO2 flow of 0.5 L/min to match the same flow as used during the HFOV experiments (producing a carinal CO2 of > 107 mm Hg, which is above the limit of the gas analyzer). Test lung compliance was set at 0.02 L/cm H2O.

The VC-CMV settings during CO2 insufflation were as follows: VT 420 mL (simulating 6 mL/kg/ideal body weight in a patient with an ideal body weight of 70 kg), PEEP 10 cm H2O, respiratory rate 15 breaths/min, FIO2 0.4, inspiratory/expiratory ratio 1:3. Once carinal CO2 was established, the gas analyzer was withdrawn back into the inspiratory limb and measurements of CO2 entrainment were obtained at incremental distances. The analyzer was then positioned 8 cm from the Y-piece to assess the effect on CO2 entrainment, while independently manipulating the different ventilator parameters as follows: PEEP 5–25 cm H2O, respiratory rate 10–30 breaths/min, VT 100–700 mL.

The Servo-i ventilator was then placed in Bi-Vent mode with baseline settings as follows: Phigh 30 cm H2O, Plow 5 cm H2O, inspiratory/expiratory ratio = 4:1, FIO2 0.4. The gas analyzer was placed 8 cm from the Y-piece and effects on CO2 entrainment were assessed as was done during the VC-CMV experiment. The ventilator parameters were independently manipulated as follows: Phigh 20–34 cm H2O, Plow 0–20 cm H2O, inspiratory/expiratory ratio 1:5–5:1.

CO2 Entrainment During HFOV in Swine Model

Once retrograde CO2 entrainment during HFOV was characterized with the mechanical lung model, a feasibility study was performed on a 75 kg swine (Sus scrofa). The use of this swine was approved by the Wilford Hall Medical Center Institutional Animal Care and Use Committee (IACUC) board and was performed during an ongoing training protocol. The swine utilized for this experiment did not receive artificial lung injury as directed by the training protocol. The swine was anesthetized with isoflurane and intubated with a 7.0 mm ETT. A fentanyl infusion was then used to continue analgesia-sedation as the swine was transitioned to a 3100B HFOV. In a similar fashion to the test lung HFOV experiment, the gas analyzing line was inserted into the inspiratory limb of the circuit and CO2 was measured at incremental distances from the Y-piece. The initial HFOV settings were identical to the baseline settings used during the mechanical test lung experiment (with the exception of FIO2 of 1.0 during the swine experiment, as directed by the IACUC training protocol). The swine was briefly hypoventilated between experiments by increasing respiratory frequency for 10 seconds to achieve mild hypercapnia (PaCO2 52.5–63.8 mm Hg). Once retrograde CO2 entrainment was confirmed within the inspiratory limb, the gas analyzing line was placed 25 cm from the Y-piece position and the HFOV parameters were independently manipulated (P̄aw, bias flow, respiratory frequency, ΔP) to assess their effects on CO2 entrainment. Despite maximal ETT cuff inflation, there was 10 mm Hg CO2 detectable within the swine's oropharynx. The effect of an additional 5 cm H2O ETT cuff leak was also assessed.

Statistical Analysis

All retrograde CO2 entrainment measurements were performed in duplicate for both the 3100B HFOV and conventional ventilator experiments on the test lung. Data were averaged and rounded to the nearest whole number. Retrograde CO2 entrainment results were identical during the HFOV test lung experiments (≤ 1 mm Hg). Due to restrictions of the swine training protocol, only one set of data for retrograde CO2 entrainment in the swine model was collected. All inspiratory limb negative pressure measurements were performed in triplicate.

Results

HFOV Inspiratory Limb Pressure Measurements

Negative pressure was readily measured within the inspiratory limb of the HFOV circuit and varied inversely with P̄aw and directly with ΔP. Representative data from a single experiment are depicted in Figure 2. Retrograde CO2 entrainment increased when more negative pressure was generated within the inspiratory limb of the circuit (Fig. 3).

Fig. 2.
Fig. 2.

Inverse effects of mean airway pressure (P̄aw) and ΔP on inspiratory limb negative pressure. Increasing P̄aw reduces negative pressure (other settings: ΔP 90 cm H2O, frequency 7 Hz, bias flow 30 L/min, inspiratory time 33%). Raising ΔP increases negative pressure (other settings: P̄aw 34 cm H2O, frequency 7 Hz, bias flow 30 L/min, inspiratory time 33%).

Fig. 3.
Fig. 3.

Effect of inspiratory limb negative pressure on retrograde CO2 entrainment. Retrograde CO2 entrainment increases (at 51 cm from the Y-piece) when more negative pressure is generated by either reducing mean airway pressure (P̄aw) or raising ΔP. The data points represent the means of duplicate experiments (measurements varied ≤ 1 mm Hg).

CO2 Entrainment During HFOV in Mechanical Lung Model

With the 3100B HFOV on baseline settings, retrograde CO2 entrainment was detectable as far back as 89 cm from the Y-piece. Entrained CO2 steadily decreased 1 mm Hg for every 2.5–5.0 cm from the Y-piece and dissipated to 0 mm Hg at 89 cm from the Y-piece (Fig. 4).

Fig. 4.
Fig. 4.

Inspiratory limb CO2 pressure and distance of CO2 entrainment at baseline settings in the mechanical lung and swine models. The high-frequency oscillatory ventilation (HFOV) settings were: mean airway pressure P̄aw 34 cm H2O, ΔP 90 cm H2O, frequency 7 Hz, bias flow 30 L/min (without cuff leak), inspiratory time 33%. Despite maximal cuff inflation, there was a persistent 10 mm Hg CO2 leak within the swine's oropharynx.

The effect of increasing P̄aw using either bias flow or the mean pressure adjustment, with a 5 cm H2O cuff leak in place, is depicted in Figure 5. In both cases, increasing P̄aw reduced retrograde CO2 entrainment. Increasing P̄aw by adjusting bias flow had a greater effect on reducing CO2 entrainment, when compared to utilizing the mean pressure adjustment. Removal of the 5 cm H2O cuff leak increased the amount of CO2 entrainment in both cases, and the trends were comparable to those seen with a cuff leak (data not shown).

Fig. 5.
Fig. 5.

Effect of mean airway pressure (P̄aw) adjustments on retrograde CO2 entrainment. The P̄aw was manipulated by either the mean pressure adjustment or by bias flow (17–60 L/min). Adjustments were made relative to baseline settings: P̄aw 34 cm H2O, ΔP 90 cm H2O, frequency 7 Hz, bias flow 30 L/min, inspiratory time 33%. Mechanical lung data were obtained at 51 cm from the Y-piece, and swine data at 25 cm from the Y-piece.

The effect on retrograde CO2 entrainment when increasing bias flow at a constant P̄aw of 34 cm H2O is shown in Figure 6. Retrograde CO2 entrainment was reduced with a higher bias flow strategy. A 5 cm H2O cuff leak further reduced CO2 entrainment once bias flow was increased above 30 L/min.

Fig. 6.
Fig. 6.

Effect of bias flow manipulations at constant mean airway pressure P̄aw on retrograde CO2 entrainment. Data were obtained at 51 cm from the Y-piece. Other high-frequency oscillatory ventilation settings remained at baseline: mean airway pressure P̄aw 34 cm H2O, ΔP 90 cm H2O, frequency 7 Hz, inspiratory time 33%.

The effect of increasing ΔP is depicted in Figure 7. In the absence of a cuff leak, CO2 entrainment became detectable once ΔP reached 70 cm H2O, and continued to rise as ΔP was increased. A 5 cm H2O cuff leak reduced CO2 entrainment yet maintained a similar trend.

Fig. 7.
Fig. 7.

Effect of ΔP manipulations on retrograde CO2 entrainment. Data were obtained 51 cm from the Y-piece in the mechanical lung model and 25 cm from the Y-piece in the swine model. Other HFOV settings remained at baseline: mean airway pressure 34 cm H2O, bias flow 30 L/min, frequency 7 Hz, inspiratory time 33%.

The effect of increasing respiratory frequency is depicted in Figure 8. CO2 entrainment was decreased as respiratory frequency was increased. In the presence of a cuff leak, retrograde CO2 entrainment was further reduced at respiratory frequencies less than 10 Hz.

Fig. 8.
Fig. 8.

Effect of respiratory frequency manipulations on retrograde CO2 entrainment. Data were obtained at 51 cm from the Y-piece in the mechanical lung model and at 25 cm from the Y-piece in the swine model. Other HFOV settings remained at baseline: mean airway pressure 34 cm H2O, ΔP 90 cm H2O, bias flow 30 L/min, inspiratory time 33%.

CO2 Entrainment During Conventional Ventilation in Mechanical Lung Model

In both the VC-CMV and Bi-Vent modes, no retrograde CO2 entrainment was detectable beyond 8 cm from the Y-piece while insufflating CO2 at either 0.1 L/min or 0.5 L/min. The absence of inspiratory limb CO2 entrainment during this experiment confirmed that this phenomenon is unique to HFOV and validated the mechanical test lung design. Further validation of the experimental design was that no CO2 entrainment was detectable during HFOV when the piston was turned off.

CO2 Entrainment During HFOV in Swine Model

With the swine on baseline HFOV settings and the ETT cuff maximally inflated, 10 mm Hg of CO2 was detectable within the Y-piece. Due to the ETT size used during the swine experiment, there was a 10 mm Hg CO2 leak detectable within the swine's oropharynx, despite maximal cuff inflation. Despite this persistent leak, inspiratory limb CO2 entrainment was identified as far back as 89 cm from the Y-piece (see Fig. 4). A 5 cm H2O cuff leak (in addition to the persistent leak) was then placed, which reduced but did not eliminate retrograde CO2 entrainment.

The effect of manipulating P̄aw, ΔP, and respiratory frequency are shown in Figures 5, 7, and 8. As observed during test lung experiments, CO2 entrainment during the swine experiment was directly proportional to ΔP and inversely proportional to P̄aw and respiratory frequency.

Discussion

Our findings demonstrate that carbon dioxide is readily detectable within the inspiratory limb of the SensorMedics 3100B HFOV in both mechanical lung and swine models. Retrograde CO2 entrainment was identified as far back as 76 cm from the Y-piece in both models, suggesting that total dead space may extend well into the inspiratory limb of the circuit. We speculate that an increase in total dead space by this phenomenon may contribute to clinical hypercapnia under some circumstances. This is a unique characteristic of the 3100B HFOV that, to our knowledge, has not previously been reported. In contrast, CO2 rebreathing has been identified during noninvasive ventilation and has prompted modifications to the circuit (eg, single circuit bi-level positive airway pressure systems).12,13

The phenomenon of retrograde CO2 entrainment is complex and has not been fully explored. Our experimental model was designed to verify that retrograde CO2 entrainment does occur and to show the effect of ventilator parameter changes on the degree of CO2 entrainment. We have also shown that inspiratory limb negative pressure correlates directly with the degree of CO2 entrainment; however, this does not imply that inspiratory limb pressure must be negative for this phenomenon to occur. There are most likely other HFOV characteristics that contribute to inspiratory limb CO2 entrainment, such as pressure differentials between the ETT and the inspiratory and expiratory limbs of the circuit during oscillatory piston cycles. In addition, our data demonstrated frequency dependence of CO2 entrainment, which suggests that variable expiratory phase time constants may contribute.

These data indicate that retrograde CO2 entrainment during HFOV is proportional to the negative pressure generated within the inspiratory limb of the circuit. During HFOV, retrograde CO2 entrainment is reduced (not eliminated) by increasing P̄aw, decreasing ΔP, placement of a cuff leak, increasing respiratory frequency, and increasing bias flow at any targeted P̄aw. Whether retrograde CO2 entrainment has a physiologic effect is still unknown. The presence of this phenomenon suggests that ventilator parameter changes that alter VT alone may not always result in the anticipated PaCO2 response. There is an equilibrium between minute ventilation and total dead space (induced by HFOV settings) that ultimately determines CO2 clearance. This may explain why increasing ΔP or lowering respiratory frequency (ie, increasing VT) does not always result in the expected PaCO2 reduction. To our knowledge, there are no studies available that have directly measured PaCO2 during 3100B HFOV setting changes while simultaneously examining for retrograde CO2 entrainment. We are now in the process of doing such a study. A majority of the current HFOV guidelines are based on the assumption that increasing VT will consistently reduce PaCO2, yet there are no studies that directly measure PaCO2 to corroborate this assumption. Previous work has looked at endotracheal tube CO2 levels as a surrogate of PaCO2 and related them to CO2 removal; however, the high-frequency oscillator used during that study is not similar to the 3100B HFOV.14,15

If retrograde CO2 entrainment is physiologically important, then reducing contamination of inspired gas may improve the effectiveness of CO2 clearance. This could be achieved to some degree by altering HFOV strategies; however, more efficient reductions might be obtained with simple modifications to the circuit.

Limitations

A limitation to our mechanical lung model experiments was assuring CO2 insufflation was comparable between the HFOV and conventional ventilator control experiments. We reconciled this by insufflating CO2 during conventional ventilation at 0.1 L/min to achieve 40 mm Hg at the carina (identical partial pressure as during the HFOV experiment), as well as insufflating at 0.5 L/min (identical flow as during HFOV experiment). CO2 at the carina was above the limit of the RGM 5250 gas analyzer (> 107 mm Hg) when insufflating at 0.5 L/min. In neither case was retrograde CO2 entrainment detected beyond 8 cm from the Y-piece. Furthermore, no CO2 was detected within the inspiratory limb of the 3100B HFOV circuit when the piston was inactivated (eg, no negative pressure within the inspiratory limb of the circuit).

There were also quantitative differences in the amount of CO2 entrainment between the mechanical lung model and swine. We attribute this finding to a persistent cuff leak during the swine experiment, despite maximal ETT cuff inflation (oropharyngeal CO2 of 10 mm Hg). Additionally, there are inherent differences in CO2 production with a live animal. Despite the lower levels of entrained CO2 measured during the swine experiment, the trends in CO2 entrainment with manipulations of HFOV parameters were similar when compared to the mechanical test lung.

A further limitation of our experimental design was that the compliance of the test lung and the swine lung were different. The test lung was set to a compliance of 0.02 L/cm H2O to simulate ARDS. The swine lung was un-injured and presumed to have normal compliance, as directed by the training protocol. The differing compliances between the test lung and swine lung may have contributed to the variation of CO2 entrainment seen in our experiment. The effect of varying compliance on CO2 entrainment is not known.

Our study was not designed to correlate carinal CO2 changes or in vivo effects on PaCO2 with ventilator parameter manipulations of retrograde CO2 entrainment. Because of this limitation, we urge caution in extrapolating this information to a clinical setting. At our institution we routinely use an RGM 5250 gas analyzer during HFOV to document the adequacy of ETT cuff leaks, and now monitor for the presence of inspiratory limb CO2 entrainment. In support of our experimental findings, we have recently observed retrograde CO2 entrainment in a hypercapnic patient during HFOV.

Our findings have potential clinical implications during use of HFOV. For example, raising P̄aw by increasing bias flow, as opposed to the mean pressure adjustment, will result in less CO2 entrainment. In contrast, lowering P̄aw using the mean pressure adjustment will result in a smaller increase in CO2 entrainment, when compared to using bias flow. Furthermore, at any targeted P̄aw, entrained CO2 is minimized by using a higher bias flow strategy. Additionally, retrograde CO2 entrainment could be clinically important during P̄aw weaning, particularly when using a high fixed ΔP (eg, 90 cm H2O).

Conclusions

In summary, this research confirms that retrograde CO2 entrainment does occur within the inspiratory limb of the 3100B HFOV circuit, and can be altered by manipulating HFOV parameters. Future research should address the incidence of this phenomenon in patients and its clinical implications.

Acknowledgments

We thank Roger Shih MD for data organization and figure creation, and Steven Venticinque MD for design and construction of the artificial trachea interface.

Footnotes

  • Correspondence: Adam W Bostick MD, Pulmonary and Critical Care Medicine, Wilford Hall Medical Center, 59MDW/SGO5P, 2200 Bergquist Drive, Suite 1, Lackland Air Force Base, San Antonio TX 78236. E-mail: bostickaw{at}gmail.com.

  • This research was partly supported by the Clinical Investigation Directorate of Wilford Hall Medical Center.

  • Dr Bostick presented a version of this paper at the International Conference of the American Thoracic Society, held May 13–18, 2011, in Denver, Colorado.

  • The authors have disclosed no conflicts of interest.

  • The opinions expressed in this document are solely those of the authors and do not represent an endorsement by or the views of the United States Air Force, the Department of Defense, or the United States Government.

  • See the Related Editorial on Page 1984

References

  1. 1.↵
    1. Sud S,
    2. Sud M,
    3. Friedrich J,
    4. Meade M,
    5. Ferguson N,
    6. Wunsch H,
    7. Adhikari N
    . High frequency oscillation in patients with acute lung injury and respiratory distress syndrome (ARDS): systematic review and meta-analysis. BMJ 2010;340:c2327. DOI: 10.1136/bmj.c2327
    OpenUrlAbstract/FREE Full Text
  2. 2.↵
    1. Derdak S,
    2. Cannon J
    . “Rescue oxygenation therapies” for severe pH1N1-associated acute respiratory distress syndrome (editorial). Crit Care Med 2010;38(11):2257–2258.
    OpenUrlPubMed
  3. 3.↵
    1. Imai Y,
    2. Nakagawa S,
    3. Ito Y,
    4. Kawano T,
    5. Slutsky A,
    6. Miyasaka K
    . Comparison of lung protection strategies using conventional and high-frequency oscillatory ventilation. J Appl Physiol 2001;91(4):1836–1844.
    OpenUrlAbstract/FREE Full Text
  4. 4.↵
    1. Yoder B,
    2. Siler-Khodr T,
    3. Winter V,
    4. Coalson J
    . High frequency oscillatory ventilation. Effects on lung function and airway cytokines in the immature baboon model for neonatal chronic lung disease. Am J Respir Crit Care Med 2000;162(5):1867–1876.
    OpenUrlPubMed
  5. 5.↵
    1. Derdak S,
    2. Mehta S,
    3. Stewart T,
    4. Smith T,
    5. Rogers M,
    6. Buchman T,
    7. et al
    . High frequency oscillatory ventilation for acute respiratory distress syndrome in adults: a randomized, controlled trial. Am J Respir Crit Care Med 2002;166(6):801–808.
    OpenUrlCrossRefPubMed
  6. 6.↵
    1. Muellenbach R,
    2. Markus K,
    3. Said H,
    4. Klosterhalfen B,
    5. Zollhoefer B,
    6. Wunder C,
    7. et al
    . High-frequency oscillatory ventilation reduces lung inflammation: a large-animal 24-h model of respiratory distress. Intensive Care Med 2007;33(8):1423–1433.
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Fessler H,
    2. Derdak S,
    3. Ferguson N,
    4. Hager D,
    5. Kacmarek R,
    6. Thompson T,
    7. Brower R,
    8. et al
    . A protocol for high-frequency oscillatory ventilation in adults: results from a roundtable discussion. Crit Care Med 2007;35(7):1649–1654.
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Sedeek K,
    2. Takeuchi M,
    3. Suchodolski K,
    4. Kacmarek R
    . Determinants of tidal volume during high frequency oscillation. Crit Care Med 2003;31(1):227–231.
    OpenUrlCrossRefPubMed
  9. 9.
    1. Hager D,
    2. Fuld M,
    3. Kaczka D,
    4. Fessler H,
    5. Brower R,
    6. Simon B
    . Four methods of measuring tidal volume during high frequency oscillatory ventilation. Crit Care Med 2006;34(3):751–757.
    OpenUrlCrossRefPubMed
  10. 10.↵
    1. Hager D,
    2. Fessler H,
    3. Kaczka D,
    4. Shanholtz C,
    5. Fuld M,
    6. Simon B,
    7. Brower R
    . Tidal volume delivery during high-frequency oscillatory ventilation in adults with acute respiratory distress syndrome. Crit Care Med 2007;35(6):1522–1529.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Fessler H,
    2. Hager D,
    3. Brower R
    . Feasibility of very high-frequency ventilation in adults with acute respiratory distress syndrome. Crit Care Med 2008;36(4):1043–1048.
    OpenUrlCrossRefPubMed
  12. 12.↵
    1. Lofaso F,
    2. Brochard L,
    3. Touchard D,
    4. Hang T,
    5. Harf M,
    6. Isabey D
    . Evaluation of carbon dioxide rebreathing during pressure support ventilation with airway management system (BiPAP) devices. Chest 1995;108(3):772–778.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Schettino G,
    2. Chatmongkolchart S,
    3. Hess D,
    4. Kacmarek R
    . Position of exhalation port and mask design affect CO2 rebreathing during noninvasive positive pressure ventilation. Crit Care Med 2003;31(8):2178–2182.
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Rossing T,
    2. Slutsky A,
    3. Lehr J,
    4. Drinker P,
    5. Kamm R,
    6. Drazen J
    . Tidal volume and frequency dependence of carbon dioxide elimination by high-frequency ventilation. N Engl J Med 1981;305(23):1375–1379.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Nishimura M,
    2. Imanaka H,
    3. Tashiro C,
    4. Taenaka N,
    5. Yoshiya I
    . Capnometry during high-frequency oscillatory ventilation. Chest 1992;101(6):1681–1683.
    OpenUrlCrossRefPubMed
Back to top

In this issue

 

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section

Related Articles

Cited By...

Keywords