help button home button
AJRCCM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by PILLOW, J. J.
Right arrow Articles by RAMSDEN, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by PILLOW, J. J.
Right arrow Articles by RAMSDEN, C. A.
Am. J. Respir. Crit. Care Med., Volume 164, Number 6, September 2001, 1019-1024

In Vitro Performance Characteristics of High-Frequency Oscillatory Ventilators

J. JANE PILLOW, MALCOLM H. WILKINSON, HEATHER L. NEIL, and C. ANDREW RAMSDEN

TVW Telethon Institute for Child Health Research, Perth, Australia; Department of Paediatrics, University of Western Australia, Perth, Australia; King Edward & Princess Margaret Hospital, Perth, Australia; Ritchie Center for Baby Health Research, Institute of Reproduction and Development, Monash University, Clayton, Victoria, Australia; and Department of Paediatrics, Monash University, Victoria, Australia



    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

This study aimed to examine the performance characteristics of four high-frequency oscillatory-type ventilators, using an in vitro model of the intubated neonatal respiratory system. Each ventilator was examined across its operative range of settings and at varying model lung compliance (C) and resistance. The oscillatory pressure waveform was measured at the airway opening (Pao). Tidal volume (VT) and flow were determined from pressure changes within the model lung (Delta PA). The spectral content of the Pao waveform differed between ventilators. The maximum ventilator VT ranged from 3.7 to 11.1 ml at 15 Hz and a mean airway pressure (Paw) of 12 cm H2O to oscillate a model lung (C = 0.4 ml/cm H2O) through a 3.0-mm internal diameter (i.d.) endotracheal tube (ETT). A small drop in C was associated with a decrease in VT and marked increase in Delta PA from 0.1 to 0.8 ml/cm H2O. The influence of C on VT and Delta PA and the pressure cost of ventilation (Delta PA/f·VT2) was dependent on the oscillatory frequency, ETT inner diameter, and the specific ventilator used. Substantive differences exist between oscillatory ventilators that need to be considered in their clinical application. The rapid establishment of optimal lung volume and oscillatory frequency is important in minimizing barotrauma during high-frequency oscillatory ventilation.

Keywords: compliance; high-frequency ventilation; newborn; respiratory distress syndrome; ventilator properties


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

The rapid adoption of high-frequency oscillatory ventilation (HFOV) as a therapeutic tool has been accompanied by a similarly expeditious development of devices designed to deliver this mode of ventilation. Commercially available HFO ventilators differ markedly, some operating only at high frequency (3-22 Hz), and some also providing an option to ventilate patients at more conventional rates (~ 1 Hz) alone, or in combination with HFOV. Five different ventilators (3100 [SensorMedics, Yorba Linda, CA], Humming V [Metran, Saitama, Japan], Stephan Medizintechnik [Gackenbach, Germany], Infant Star [Infrasonics, San Diego, CA], and Dufour [Villeneuve d'Aseq, France]) were used for the clinical trials of HFOV that have been published to date either in final (1) or abstract form (9). A number of other commercially available and custom-designed HFO ventilators are currently in clinical use around the world.

The first report describing significant differences between the in vitro performance of high-frequency ventilators was made by Fredberg and colleagues in 1987 (12). They examined eight ventilators, including a jet ventilator (Life Pulse; Bunnell, Salt Lake City) and three flow interrupters (Bird Military, Emerson interrupter, and the Infrasonics Infant Star) in addition to four oscillators (Gould Texas Research, Senko Hummingbird BMO 20N, Emerson oscillator, and the Metrex Flutter II). Their study showed that there are differences between the oscillatory pressure waveforms and the efficiency of volume delivery (12) both between and within each ventilator category. Some of the differences between categories have diminished over the last decade, however, as modifications such as the inclusion of a Venturi device in the more recently released interrupter devices have resulted in a much closer similarity between the flow interrupter and oscillator groups. In practical terms, the two are commonly classified together as oscillatory ventilators, setting both types of ventilator apart from the somewhat different modality of high-frequency jet ventilation.

Despite the range of oscillatory devices in clinical use, published data on the mechanical performance of newer ventilators remains limited. In 1997, Jouvet and coworkers reported a bench study of the Babylog 8000 (Drägerwerk, Lubeck, Germany), the Dufour OHF 1, and the SensorMedics 3100A (13). They observed differences in the relationship between tidal volume and peak-to-peak pressure amplitude, and that the SensorMedics 3100A was the most powerful, delivering "supraphysiologic" tidal volume (VT) at maximum ventilator power. They also found that in the Dufour OHF 1 ventilator, VT decreased with an increase in compliance from 1 to 5 ml/cm H2O, a finding that contrasted with Fredberg's observation that VT was invariant with compliance over this range. One year later, Hatcher and colleagues (14) reported an in vivo evaluation of current devices including the SensorMedics 3100A, the Humming V (Metran), the Infant Star 950 (Mallinkrodt, St. Louis, MO), and the Babylog 8000 (Drägerwerk) in rabbits before and after saline lavage. Their study again demonstrated important differences between ventilators, although they observed that tidal volume tended to be lower after saline lavage and a reduction in lung compliance. Unfortunately, the in vivo design of that study limited the extent to which the independent effects of lung compliance and resistance could be examined.

In our study, we use an in vitro lung model to define in greater detail the differing relationship between VT and pressure amplitude in four oscillatory ventilators over the range of lung compliance commonly encountered in the newborn infant. The results are of significance to the clinical application of these ventilators and may aid understanding of the mechanisms contributing to lung injury when these ventilators are used with a "low lung volume" strategy.

    METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Ventilators and in Vitro Lung Model

Four high-frequency ventilators were studied, including the Humming V (HUMV), the SensorMedics 3100A (SM3100A), the Babylog 8000 (BL8000), and the Infant Star 950 (IS950). Observations on the effect of varying ventilator frequency in the Humming V were supplemented with data from an earlier model (Humming II, HUMII) with a broader frequency range (10-25 Hz).

To simulate the mechanical properties of the respiratory system of the intubated human newborn infant with hyaline membrane disease (HMD), we used a lung model comprising an endotracheal tube (11 cm long, 3.0-mm i.d.) sealed into the neck of a 590-ml glass flask (15).

Measurement of Pressure

Pressure at the airway opening (Delta Pao) and in the lung model (Delta PA) was measured with MP15 pressure transducers (Micron Instruments, Simi Valley, CA), amplified, filtered, digitized, and stored as previously described (15). Oscillatory pressure amplitude was determined as the peak-to-trough pressure excursions, with an average value obtained from three such determinations.

Experiment Protocols

The ventilator and model lung conditions used for each experiment protocol are summarized in Table 1. Nominal amplitude was defined as the ventilator-displayed amplitude (Delta Pvent) required to deliver a tidal volume of 4.8 ml under nominal conditions (see Table 1).


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

TABLE 1

 NOMINAL CONDITIONS AND RANGES

Ventilator operation. Each ventilator was set up in accordance with the manufacturer guidelines. Mean airway pressure (<OVL>Paw</OVL>) was maintained at 12 cm H2O. Individual ventilator settings were randomly studied across the available ranges of frequency and ventilator-displayed amplitude (Delta Pvent) while remaining ventilator parameters were held constant. The effect of changing I:E ratio on VT was evaluated in the SM3100A and the HUMV. In the SM3100A, inspiratory time, expressed as a percentage of total cycle time (%IT), was adjusted in steps of 2-3% between 30% (I:E = 1:2.3) and 50% (I:E = 1:1). Internal dip-switch settings were adjusted in the HUMV to enable delivery of I:E ratios of 1:2 and 1:3.

Model lung impedance. Impedance (Z) was changed by altering the compliance (C), resistance (R), and inertance (I) of the lung model. To achieve this, a 3.1-L glass flask was substituted for the 590-ml flask in the model lung and filled with measured water volumes to vary C from 0.15 to 2.15 ml/cm H2O. Resistance and inertance were varied by using three endotracheal tubes (ETTs) (Portex, Hythe, Kent, England) with different internal diameters (2.5, 3.0, and 3.5 mm).

Analysis

Waveform. Pressure waveforms measured at the airway opening with each ventilator were analyzed with fast Fourier transforms (FFTs) with a frequency resolution of 0.24 Hz to determine the frequency content of each waveform and the proportion of the signal power contained within the fundamental frequency (15 Hz) under nominal conditions.

Tidal volume (VT). In the in vitro lung model, VT was calculated by the adiabatic gas equation (12). Efficiency of volume delivery (VT/Delta Pao) as used by Fredberg and coworkers (12), was calculated across a range of frequencies for each ventilator.

    RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Ventilator Operation

Frequency. The operating range of each ventilator is summarized in Table 2. There were differences in the frequency range over which each ventilator could operate, although the 13- to 15-Hz range was common to each ventilator.


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

TABLE 2

 OPERATING RANGES

I:E ratio. The SM3100A, and the HUMV, were the only ventilators in which direct control of the I:E ratio was achievable. In the SM3100A, this was achieved over a continuum whereas in the HUMV, the I:E ratio could be adjusted in discrete steps by changing to internal dip-switch settings. This is not an option available in the routine clinical application of this ventilator. In both the SM3100A and the HUMV, the I:E ratio is inferred as the respective duration of the to-and-fro motions of the diaphragm/piston, rather than from the duration of inspiratory and expiratory flow measured at the airway opening. The IS950 has a fixed absolute inspiratory time (18 ms) and hence the I:E ratio varies with frequency. No indication is given by the BL8000 of what I:E ratio is set; however, when using this ventilator, the I:E ratio calculated from the duration of inspiratory and expiratory flow at the airway opening varied between 1:0.8 and 1:2.3 with both amplitude and frequency (data not shown).

Waveform. The complex waveform of SM3100A and IS950 contrast sharply with the quasisinusoidal waveform of the HUMV and BL8000 ventilators as evidenced by the presence and absence, respectively, of harmonic structure on power spectral analysis (Figure 1).


View larger version (20K):
[in this window]
[in a new window]
  		Figure 1.   Power spectral analysis of oscillatory pressure waveform at the airway opening. The power at each harmonic is expressed relative to the signal power at the fundamental frequency (15 Hz). (A) HUMV; (B) BL8000; (C) SM3100A; (D) IS950.

Effect of Ventilator Properties

Amplitude. Delta Pao and Delta Pvent were closely related in the HUMV, IS950, and SM3100A. Although the relationship deviated slightly from linearity for the HUMV and IS950, Delta Pao was always within 10% of Delta Pvent (Figure 2A and 2D). In the SM3100A, however, Delta Pao was typically 50% greater than Delta Pvent (Figure 2C). The BL8000 does not provide an indicator of absolute peak-to-trough pressure excursions at the airway opening; however, there was a nonlinear increase in Delta Pao up to approximately 50% maximum "power," beyond which no further increase in Delta Pao was obtained at the nominal <OVL>Paw</OVL> (Figure 2B).


View larger version (24K):
[in this window]
[in a new window]
  		Figure 2.   Ventilator amplitude as an indicator of peak-to-peak pressure variations at the airway opening. Delta Pvent = amplitude indicated on ventilator display (cm H2O); Delta PAo = peak-trough amplitude measured at the airway opening; % max amp = percentage of maximum ventilator amplitude. (A) HUMV; (B) BL8000; (C ) SM3100A (open squares) and SM3100 (closed squares); (D) IS950; dotted lines represent line of identity.

In all ventilators, VT increased proportionately with increases in amplitude at constant frequency, although in the BL8000 no further increase in VT was observed when ventilator setting is increased above approximately 50% maximum ventilator power. Maximum VT at C = 0.4 ml/cm H2O and a frequency of 15 Hz are presented in Table 2. Under nominal conditions, the maximum VT of the HUMV (11.1 ml) was similar to that of the SM3100A (11.0 ml), whereas the maximum VT of the BL8000 and the IS950 were 33.7 and 56.4%, respectively, of the maximum VT obtained with the SM3100A.

Frequency. In the HUMV and BL8000, the Delta Pao decreased as frequency increased at constant Delta Pvent (Figure 3A and 3B). In the SM3100A, Delta Pao remained relatively constant across the frequency range at constant Delta Pvent (Figure 3C) whereas Delta Pao increased at constant Delta Pvent in the IS950 (Figure 3D).


View larger version (23K):
[in this window]
[in a new window]
  		Figure 3.   Effect of increasing frequency on VT and Delta Pao. Measurements performed at nominal ventilator amplitude: VT, closed symbols (left axis); Delta Pao, open symbols (right axis). (A) Humming II; (B) BL8000; (C ) SM3100A; (D) IS950.

Tidal volume increased as frequency decreased at constant Delta Pvent (Figure 3). Thus at 10 Hz, VT was 158.5% (HUMII), 163.7% (BL8000), 117.5% (IS950), and 143.1% (SM3100A) greater than the nominal VT at 15 Hz. The HUMV did not operate below 13 Hz.

Efficiency. As frequency increased from 10 to 15 Hz, the efficiency of volume delivery for each ventilator decreased by 73.9% (HUMII and BL8000), 67.8% (SM3100A), and 77.8% (IS950). Under nominal conditions, the highest efficiency was observed in the SM3100A.

I:E ratio. Under nominal conditions and maintaining a constant Delta Pvent, decreasing the I:E ratio set on the ventilator from 1:1 to 1:2 decreased VT by 7.0 ± 2.0% (mean ± SD) in the SM3100A and by 8.3 ± 2.1% in the HUMV.

Effect of Lung Mechanical Properties

Effect of compliance. The VT increased asymptotically with compliance to a maximum VT such that
V<SC>t</SC>=V<SC>t</SC><SUB>pl</SUB>(1−e<SUP>−3C/Cthr</SUP>) (1)

where VTpl is the plateau VT and Cthr represents the compliance value at which VT had reached 95% of VTpl at any specific ventilator setting and endotracheal tube size. VTpl and Cthr were determined by least-squares estimation.

At values of C above Cthr, VT was largely independent of C. Below Cthr, however, small reductions in C effected significant falls in VT. This finding was evident in each of the high-frequency oscillatory ventilators although to differing extents. The VT at C = 0.15 ml/cm H2O, for example, was 89.3% (HUMV), 76.1% (BL8000), 56.7% (SM3100A), and 36.3% (IS950) of the plateau VT. The value of Cthr varied among the different ventilators when operated at 15 Hz (see Table 2).

The effect of frequency on the value of Cthr and on the relationship between VT and C, is illustrated in Figure 4A with the SM3100A. Whereas VTpl was rapidly established with increasing compliance at 15 Hz, no such plateau was reached at 5 Hz over a range of C from 0.15 to 2.1 ml/cm H2O. Using three different frequencies at C = 0.15 ml/cm H2O, VT was 32.7% (5 Hz), 48.4% (10 Hz), and 62.5% (15 Hz) of VTpl at each frequency. The calculated value for Cthr decreased with increasing frequency (see Table 3). Whatever frequency was used, the reduction in VT observed below Cthr was associated with a marked increase in Delta PA (Figure 4B). The rise in Delta PA with decreasing C was greatest at the higher frequencies.


View larger version (32K):
[in this window]
[in a new window]
  		Figure 4.   Effect of compliance on VT and Delta PA at different frequencies and ETT sizes. Results are shown for both tidal volume (A and C ) and Delta PA (B and D). (A and B) Results obtained at each of three frequencies, using a 3.0-mm ETT i.d. in the SM3100A: 5 Hz (open circles), 10 Hz (open triangles), and 15 Hz (open squares). (C and D) Effect of varying ETT size: ETT inner diameter of 2.5 mm (closed circles), 3.0 mm (closed triangles), and 3.5 mm (closed squares). The dotted lines represent the Cthr at each frequency or ETT size.


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

TABLE 3

 DETERMINATION OF Cthr IN THE SM3100A

Effect of resistance and inertance. The effect of ETT inner diameter on the value of Cthr, and on the relationship between VT and C, is shown in Figure 4C for the SM3100A. As endotracheal tube size increased, the effect of low C value on VT increased proportionately such that at 15 Hz, VT at C = 0.15 ml/cm H2O was 87.5% (2.5-mm i.d.), 56.6% (3.0-mm i.d.), and 35.2% (3.5-mm i.d.), respectively, of plateau VT. The ETT inner diameter had a negligible effect, however, on the value of Cthr. Notably, however, the rise in Delta PA with decreasing C was present with all ETTs (Figure 4D).

Pressure Cost of Ventilation

The potential consequences of changes in both VT and Delta PA with changes in compliance at differing frequencies and endotracheal tube sizes was estimated in terms of the cyclical alveolar pressure cost per unit ventilation [Delta PA/(f · VT2)] and is illustrated in Figure 5. The lowest value for Delta PA/(f · VT2) was achieved at the highest frequency and largest ETT. The observed differences were most marked at low compliances, whereas only small differences in Delta PA/(f · VT2) were noted between frequencies or different endotracheal tube sizes when compliance was set at values observed in the healthy neonatal lung (16).


View larger version (18K):
[in this window]
[in a new window]
  		Figure 5.   Effect of compliance on the pressure cost of ventilation. The changes in pressure amplitude are expressed per unit ventilation (f · VT2) over a range of compliances at (A) frequencies of 5 Hz (open circles), 10 Hz (open triangles), and 15 Hz (open squares), using a 3.0-mm i.d. ETT and (B) at 15 Hz, using an ETT inner diameter of 2.5 mm (closed circles), 3.0 mm (closed triangles), and 3.5 mm (closed squares). The SM3100A was used for each set of observations.

    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

This study provides a detailed assessment of four of the most commonly used high-frequency oscillators and demonstrates considerable variability between machines in oscillatory pressure waveform, and the often complex interactions between ventilator settings and resulting VT and Delta Pao. Our observations highlight the potential difficulties that may be encountered when trying to compare the effects and outcomes of ventilator strategies when using different devices. The most interesting finding of our study relates to load dependence of both VT and Delta PA during HFOV. This is the first in vitro study to comprehensively document these changes, and to describe the extent to which they are influenced by frequency and ETT diameter.

Critique of Methodology

Our in vitro lung model is similar to that described by other investigators (12) and was chosen because its mechanical properties resemble those of the intubated human newborn infant with severe hyaline membrane disease (12). The nominal frequency of 15 Hz falls within the range of frequencies used clinically and could be achieved by all ventilators under study. The constant mean airway pressure of 12 cm H2O is in the range most infants would be exposed to at some stage of their course on HFOV. While this study did not specifically examine the effects of varying <OVL>Paw</OVL>, other reports published subsequent to the completion of our studies suggest that an increase in <OVL>Paw</OVL> is associated with a decrease in VT in most high-frequency oscillators with the exception of the BL8000 (14).

Performance Characteristics

The general principles applied to generate pressure oscillations vary between high-frequency ventilators and have been described in detail elsewhere (14, 17). Information about the performance characteristics of high-frequency oscillatory ventilators is limited to the in vitro study undertaken by Fredberg and colleagues in 1987 (12), followed a decade later by similar laboratory (13) and in vivo (14) performance evaluations of more contemporary oscillators. The seven oscillatory ventilators in the first study (12) have since been superseded by more current models. Jouvet and coworkers (13) described the in vitro behavior of the SM3100A, the BL8000, and the Dufour OHF 1; however, this study included a limited evaluation of the load dependence of VT delivery during HFOV. The study by Hatcher and coworkers (14) included SM3100A, HUMV, BL8000, and IS950 high-frequency ventilators to ventilate adult rabbits before and after saline bronchoalveolar lavage. The current study uses an in vitro lung model similar to that employed by Fredberg and coworkers (12) and Jouvet and colleagues (13); however, it offers a more thorough and previously unpublished evaluation of the load dependence of VT delivery and other previously unpublished performance characteristics, thus providing novel data of significant importance to the clinician.

Waveform. Differences in the shape of the oscillatory pressure waveform produced by different ventilators have been reported elsewhere (14) and are highlighted in the current study, which contrasts the complex harmonic content of the IS950 and SM3100A ventilators with the predominantly fundamental frequency composition of the pressure waveforms produced by the HUMV and the BL8000. The clinical significance of differences in harmonic content and shape of the Pao waveform has yet to be evaluated.

Pressure amplitude. The potential difficulties in drawing comparisons between ventilators on the basis of recorded values for pressure amplitude displayed by the oscillator (Delta Pvent) is highlighted by the differences between Delta Pvent and Delta Pao (Figure 2) and the influence of oscillatory frequency on this relationship (Figure 3). These interrelationships are further complicated by the differences in efficiency of tidal volume delivery between the ventilators. A more useful comparative index would be tidal volume, provided that the frequency response characteristics of flow-measuring devices used were uniform and sufficient to handle the high flows and frequencies associated with HFOV.

Frequency. The decrease in VT that we observed with increasing frequency and decreasing %IT at constant Delta Pvent has been described previously (18, 19). Substantially larger VT results when ventilators are operated at lower frequencies are due to the diminished effect of the flow-dependent ETT resistance and the increased time available for flow. Tidal volume delivery at low frequency will ultimately be limited by the time constant of the lung. Differences between ventilators in the nature of the change in VT with frequency arise as a result of the differences in the Delta Pao/Delta Pvent relationship with increasing frequency (Figure 3). In the clinical situation, however, gains in VT below 10 Hz are offset by increased transmission of Delta Pao to the alveolus (20) and the potential volutrauma and barotrauma that this may cause.

Differences in the efficiency of volume delivery (VT/Delta Pao) primarily reflect the diversity of ventilator waveforms and illustrate the potential hazards in utilizing Delta Pvent to draw parallels between different ventilators. Ventilators with a higher efficiency of volume delivery will require lower Delta Pao to achieve the same degree of gas exchange. This presumes, however, that the relationship between frequency, VT, and gas exchange is the same for different oscillatory flow waveforms. A more pertinent comparison of the effectiveness of gas exchange and ventilatory requirements between the different ventilators would include VT and a means of assessing CO2 elimination.

Impedance. Our finding of a strong dependence of VT on C (Figure 4) at low compliance contrasts sharply with the apparent insensitivity of VT to compliance of the respiratory system in the in vitro study by Fredberg and coworkers (12). Although Hatcher and coworkers (14) have also observed that VT was reduced at low compliance, our in vitro study is the first to define this relationship in detail across a range of compliances and ventilatory frequencies associated with the use of HFOV for neonatal respiratory disease. When C falls below Cthr (~ 0.5 ml/cm H2O at 15 Hz), further declines in C are associated with substantial increases in transmission of the pressure at the airway opening to the interior of the model lung. While Delta PA is inversely proportional to C above Cthr in the region of constant VT, much greater changes in PA were observed below Cthr in the region of decreasing flow and reduced VT. Previous measurements of Delta PA during HFOV have been reported only for the normal lung, with the exception of one study by Kamitsuka and colleagues, who used alveolar capsules to measure alveolar pressure during HFOV in rabbits before and after saline lavage (21). Although their experiments were primarily designed for another purpose, their results suggest that Delta PA increases after saline lavage compared with pressures recorded in the healthy state.

The transmission of Delta Pao to the interior of the lung may be exaggerated in our model, which lacks a complex branching airway. Under nominal conditions, however, and using published values of resistance in the neonate with respiratory distress (29.5 cm H2O · s/L) (16), the resistance of our model (56 cm H2O · s/L) would account for at least two-thirds of the total resistance of the intubated respiratory system. Furthermore, increasing the resistance of our model showed that the same dependence of Delta PA on C was evident.

These findings can be explained by understanding how the different components of impedance contribute to changes in Delta PA and volume. The impedance of the respiratory system may be divided into frequency-dependent resistive and elastic properties of the tissues, which dominate below the resonant frequency (f0, frequency at which the inertive and elastic properties cancel out) and the inertive and resistive properties of the airways and ETT, which represent the main source of impedance above f0 (22). The tissues are represented by the glass flask (C) in our model, and the inertance (I) and resistance are represented by the ETT. In our system, the value obtained for inertance using a 3.0-mm ETT was 0.24 cm H2O · s2/L. Given that
f<SUB>0</SUB>=<FR><NU>1</NU><DE>2Π(IC)<SUP>1/2</SUP></DE></FR> (2)

the f0 of our lung model ranged from 7.1 Hz at a compliance of 2.1 ml/cm H2O to 32.5 Hz at a compliance of 0.1 ml/cm H2O. In scenarios where the oscillatory frequency exceeded f0 (e.g., high compliance), the ETT resistance and inertance were the dominant parameters limiting amplitude of the PA and flow waveforms. In contrast, when model conditions were dominated by a low compliance (< Cthr), f0 was substantially higher than the frequency used for HFOV, and hence C significantly influenced the magnitude of both VT and Delta PA. Likewise, given the flow dependence of the ETT resistance during HFOV (23), the use of lower frequencies (and therefore lower flows), and the substitution of larger ETT inner diameter, will also both effectively reduce the resistive contribution of the ETT, and increase the influence of lung compliance on the overall impedance of the respiratory system.

A similar argument may explain the observed variability between ventilators with regard to the percent change in VT with C, occurring below Cthr at any given frequency and ETT inner diameter. Differences in circuits and compressible volume at the piston/diaphragm may contribute significantly to the overall system compliance, which in turn affects VT delivery. The impedance of each ventilator and associated circuitry proximal to the endotracheal tube was not measured in our study.

Clinical relevance. The findings of our in vitro study clearly demonstrate that the clinical effects of manipulating ventilator settings may differ with each HFOV device. The Delta Pvent required to establish a particular VT will vary between machines, and the effect of altering frequency would result in very different effects on VT and PCO2. The differences are accentuated by features specific to particular devices, such as the nonlinear relationship between percent ventilator amplitude and Delta Pao in the BL8000. These effects need to be known and understood to ensure appropriate and optimal use of each ventilator in the clinical situation.

The dependence of VT on C across both ranges of frequency and variations between ventilators has particular relevance to the recovery period of HMD and after surfactant treatment and initial lung volume optimization, when substantial increases in VT could occur without changing ventilator amplitude, thus increasing the potential for volutrauma and hypocarbia. Although the increased stability of VT across a range of C values at higher frequency further illustrates the potential of HFOV to protect the lung from volutrauma, the differences observed between ventilators is concerning. This may be particularly important in ventilators such as the SM3100A and the IS950, where this effect was particularly evident and where monitoring of VT is not available. Although real-time monitoring of VT during HFOV is now integrated into the BL8000 circuit, the accuracy of this device has not been tested.

Finally, our study suggests that in the presence of reduced compliance, oscillation of alveolar pressure may be significantly larger than previously thought, and is coupled with a concomitant reduction in VT. In a landmark theoretical paper, Venegas and Fredberg have shown that the pressure cost of achieving flow (Delta PA/f · VT) decreases with increasing frequency (24). Elimination of CO2 during HFOV, however, is estimated to be proportional to f · VT2 (18, 19, 25, 26). Evaluating the clinical significance of changes in compliance on the pressure cost of achieving CO2 exchange is therefore most appropriately assessed as alveolar pressure cost per unit ventilation [Delta PA/(f · VT2)] as illustrated in Figure 5. Using this approach, it is clear that Delta PA/(f · VT2) is lower at 15 Hz than at 5 Hz in the poorly compliant lung, whereas there is little difference in this parameter when the two frequencies are used to ventilate the normally compliant lung. These findings are consistent with the theoretical study of Venegas and Fredberg (24). It is important to note, however, that although HFOV at 15 Hz afforded significantly greater protection from alveolar pressure swings compared with 5 Hz in the noncompliant lung, these remained significantly higher than those observed when ventilating a normally compliant lung at any of the three frequencies examined. This assumes, however, that CO2 removal is equally related to f · VT2 across the spectrum of frequencies used during HFOV.

The independent effect of a high Delta PA on subsequent development of chronic lung injury during HFOV ventilation is unknown. High Delta PA are likely to result in the presence of reduced lung compliance, which may occur at both low and high lung volumes, although these will be lower during HFOV than during ventilation at more conventional frequencies. Superimposing increased Delta PA on a constant high background mean airway pressure could create peak pressures that approach those used during conventional ventilation, only more often, and potentially cause barotrauma to susceptible areas within the parenchyma of the developing lung. This in vitro study supports the concept of a rapid establishment of optimal rather than high lung volumes to minimize barotrauma during HFOV (1, 17, 27), and the use of frequencies that are appropriate for the mechanical parameters of the ventilated lung.

    Footnotes

Correspondence and requests for reprints should be addressed to Jane Pillow, Ph.D., Portex Unit, Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK. E-mail: jpillow{at}btinternet.com

(Received in original form May 2, 2000 and in revised form May 8, 2001).

Acknowledgments: Supported by the Financial Markets Foundation for Children; the Royal Australasian College of Physicians S.W. Shields Research Fellowship; and Newborn Services, Monash Medical Center. The generous loan of high-frequency ventilators from Senko Medical Instruments, Metran, SensorMedics Corporation, Drägerwerk, and Infrasonics is also gratefully acknowledged.
    References
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

1. Gerstmann D, Minton S, Stoddard R, Meredith K, Monaco F, Bertrand JM, Battisti O, Langhendries J, Francois A, Clark R. The Provo multicenter early high-frequency oscillatory ventilation trial: improved pulmonary and clinical outcome in respiratory distress syndrome. Pediatrics 1996; 98: 1044-1057 [Abstract/Free Full Text].

2. Clark R, Gerstmann D, Null D Jr,, deLemos R. Prospective randomized comparison of high-frequency oscillatory and conventional ventilation in respiratory distress syndrome. Pediatrics 1992; 89: 5-12 [Abstract/Free Full Text].

3. Ogawa Y, Miyasaka K, Kawano T, Imura S, Inukai K, Okuyama K, Oguchi K, Togari H, Nishida H, Mishina J. A multicenter randomised trial of high frequency oscillatory ventilation as compared with conventional mechanical ventilation in preterm infants with respiratory failure. Early Hum Dev 1993; 32: 1-10 [Medline].

4. HIFI Study Group. High-frequency oscillatory ventilation compared with conventional mechanical ventilation in the treatment of respiratory failure in preterm infants. N Engl J Med 1989;320:88-93.

5. HiFO Study Group. Randomized study of high-frequency oscillatory ventilation in infants with severe respiratory distress syndrome. J Pediatr 1993;122:609-619.

6. Plavka R, Kopecky P, Sebron V, Svihovec P, Zlatohlavkova B, Janus V. A prospective randomized comparison of conventional mechanical ventilation and very early high frequency oscillatory ventilation in extremely premature newborns with respiratory distress syndrome. Intensive Care Med 1999; 25: 68-75 [Medline].

7. Rettwitz-Volk W, Veldman A, Roth B, Vierzig A, Kachel W, Varnholt V, Schlosser R, von Loewenich V. A prospective, randomized, multicenter trial of high-frequency oscillatory ventilation compared with conventional ventilation in preterm infants with respiratory distress syndrome receiving surfactant (see comments). J Pediatr 1998; 132: 249-254 [Medline].

8. Froese A, Butler P, Fletcher W, Byford L. High-frequency oscillatory ventilation in premature infants with respiratory failure: a preliminary report. Anesth Analg 1987; 66: 814-824 [Abstract/Free Full Text].

9. Thome U, Kossel H, Lipowsky G, Porz F, Furste H, Genzel-Boroviczeny O, Troger J, Oppermann H-C, Hogel J, Pohlandt F, HFOV Study Group. High frequency oscillatory ventilation (HFOV) compared with high rate intermittent positive pressure ventilation (IPPV) as first line therapy for premature infants with respiratory insufficiency. A prospective randomized multicenter trial. Pediatr Res 1998;43:300A.

10. Ramanathan R, Ruiz I, Tantivit P, Cayabyab R, deLemos R. High frequency oscillatory ventilation compared to conventional mechanical ventilation in preterm infants with respiratory distress syndrome. Pediatr Res 1995; 37: 347A .

11. Lombet J, Claris O, Debauche C, Putet G, Rigo J, Verellen G, Salle B. High frequency oscillation (HFO) versus conventional mechanical ventilation (CMV) for respiratory distress syndrome (RDS). Pediatr Res 1996;40:540:A149.

12. Fredberg J, Glass G, Boynton B, Frantz I III.. Factors influencing mechanical performance of neonatal high-frequency ventilators. J Appl Physiol 1987; 62: 2485-2490 [Abstract/Free Full Text].

13. Jouvet P, Hubert P, Isabey D, Pinquier D, Dahan E, Cloup M, Harf A. Assessment of high-frequency neonatal ventilator performances. Intensive Care Med 1997; 23: 208-213 [Medline].

14. Hatcher D, Watanabe H, Ashbury T, Vincent S, Fisher J, Froese A. Mechanical performance of clinically available, neonatal, high-frequency, oscillatory-type ventilators. Crit Care Med 1998; 26: 1081-1088 [Medline].

15. Pillow JJ, Neil H, Wilkinson M, Ramsden C. Effect of I:E ratio on mean alveolar pressure during high-frequency oscillatory ventilation. J Appl Physiol 1999; 87: 407-414 [Abstract/Free Full Text].

16. Dorkin H, Stark A, Werthammer J, Strieder D, Fredberg J, Frantz I III.. Respiratory system impedance from 4 to 40 Hz in paralyzed intubated infants with respiratory disease. J Clin Invest 1983; 72: 903-910 .

17. Gerstmann D, deLemos R, Clark R. High-frequency ventilation: issues of strategy. Clin Perinatol 1991; 18: 563-580 [Medline].

18. Rossing T, Slutsky A, Lehr J, Drinker P, Kamm R, Drazen J. Tidal volume and frequency dependence of carbon dioxide elimination by high frequency ventilation. N Engl J Med 1981; 305: 1375-1379 [Abstract].

19. Slutsky A, Kamm R, Rossing T, Loring S, Lehr J, Shapiro A. Effects of frequency, tidal volume and lung volume on CO2 elimination by high frequency (2-30Hz), low tidal volume ventilation. J Clin Invest 1981; 68: 1475-1494 .

20. Frantz IDI, Close RH. Alveolar pressure swings during high frequency ventilation in rabbits. Pediatr Res 1985; 19: 162-166 [Medline].

21. Kamitsuka MD, Boynton BR, Villanueva D, Vreeland PN, Frantz ID III.. Frequency, tidal volume, and mean airway pressure combinations that provide adequate gas exchange and low alveolar pressure during high frequency oscillatory ventilation in rabbits. Pediatr Res 1990; 27: 64-69 [Medline].

22. Hantos Z, Daroczy B, Suki B, Nagy S, Fredberg JJ. Input impedance and peripheral inhomogeneity of dog lungs. J Appl Physiol 1992; 72: 168-178 [Abstract/Free Full Text].

23. Gavriely N, Solway J, Loring S, Butler J, Slutsky A, Drazen J. Pressure- flow relationships of endotracheal tubes during high-frequency ventilation. J Appl Physiol 1985; 59: 3-11 [Abstract/Free Full Text].

24. Venegas JG, Fredberg JJ. Understanding the pressure cost of ventilation: why does high-frequency ventilation work? Crit Care Med 1994;22(9 Suppl):S49-S57.

25. Jaeger M, Kurzweg U, Banner M. Transport of gases in high-frequency ventilation. Crit Care Med 1984; 12: 708-710 [Medline].

26. Fredberg J. Augmented diffusion in the airways can support pulmonary gas exchange. J Appl Physiol 1980; 49: 232-238 [Abstract/Free Full Text].

27. Froese AB. Role of lung volume in lung injury: HFO in the atelectasis-prone lung. Acta Anaesth Scand Suppl 1989; 90: 126-130 [Medline].




This article has been cited by other articles:


Home page
 Arch. Dis. Child. Fetal Neonatal Ed.Home page
J A Leipala, S Iwasaki, A Milner, and A Greenough
Accuracy of the volume and pressure displays of high frequency oscillators
Arch. Dis. Child. Fetal Neonatal Ed., March 1, 2004; 89(2): F174 - F176.
[Abstract] [Full Text] [PDF]

Home page
 Evid. Based Nurs.Home page
C. Cheng and P. Shah
High frequency oscillatory ventilation was not more effective than conventional ventilation in preterm infants
Evid. Based Nurs., April 1, 2003; 6(2): 41 - 41.
[Full Text] [PDF]

Home page
 Am. J. Respir. Crit. Care Med.Home page
S. Derdak, S. Mehta, T. E. Stewart, T. Smith, M. Rogers, T. G. Buchman, B. Carlin, S. Lowson, J. Granton, and the Multicenter Oscillatory Ventilation
High-Frequency Oscillatory Ventilation for Acute Respiratory Distress Syndrome in Adults: A Randomized, Controlled Trial
Am. J. Respir. Crit. Care Med., September 15, 2002; 166(6): 801 - 808.
[Abstract] [Full Text]

Home page
 NEJMHome page
A. R. Stark
High-Frequency Oscillatory Ventilation to Prevent Bronchopulmonary Dysplasia -- Are We There Yet?
N. Engl. J. Med., August 29, 2002; 347(9): 682 - 684.
[Full Text] [PDF]

Home page
 Am. J. Respir. Crit. Care Med.Home page
M. J. TOBIN
Critical Care Medicine in AJRCCM 2001
Am. J. Respir. Crit. Care Med., March 1, 2002; 165(5): 565 - 583.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by PILLOW, J. J.
Right arrow Articles by RAMSDEN, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by PILLOW, J. J.
Right arrow Articles by RAMSDEN, C. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Proc. Am. Thorac. Soc. Am. J. Respir. Cell Mol. Biol.
Copyright © 2001 American Thoracic Society