Inherent Instability of Pressure Support Ventilation? |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Pressure support ventilation (PSV) is almost universally employed in the management of actively breathing ventilated patients with acute respiratory failure. In this partial support mode of ventilation, a fixed pressure is applied to the airway opening, and flow delivery is monitored by the ventilator. Inspiration is terminated when measured inspiratory flow falls below a set fraction of the peak flow rate (flow cutoff); the ventilator then cycles to a lower pressure and expiration commences. We used linear and nonlinear mathematical models to investigate the dynamic behavior of pressure support ventilation and confirmed the predicted behavior using a test lung. Our mathematical and laboratory analyses indicate that pressure support ventilation in the setting of airflow obstruction can be accompanied by marked variations in tidal volume and end-expiratory alveolar pressure, even when subject effort is unvarying. Unstable behavior was observed in the simplest plausible linear mathematical model and is an inherent consequence of the underlying dynamics of this mode of ventilation. The mechanism underlying the observed instability is "feed forward" behavior mediated by oscillatory elevation in end-expiratory pressure. In both mathematical and mechanical models, unstable behavior occurred at impedance values and ventilator settings that are clinically realistic.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Keywords: dynamics; mathematical model; mechanical ventilation
Pressure support ventilation (PSV) is almost universally employed in the management of actively breathing ventilated patients with acute respiratory failure. PSV can be used as a primary mode of ventilatory support, to offset resistive loads imposed by endotracheal intubation, or to provide graded respiratory assistance during liberation from mechanical ventilation ("weaning") (1, 2). In this partial support mode of ventilation, a fixed pressure is applied to the airway opening, and flow delivery is monitored by the ventilator. Inspiration is terminated when measured inspiratory flow falls below a set value or below a specific fraction of the peak flow rate (flow cutoff, commonly 25-45% of peak inspiratory flow); the ventilator then cycles to a lower pressure (positive airway pressure, or CPAP) and expiration commences. PSV allows the patient to influence the depth and duration of each breath. Although patient control may underlie much of the variability in respiratory pattern observed during PSV, in this report we demonstrate that, under certain circumstances, irregularity may arise from dynamic instability of the mode of ventilation itself.
We previously investigated the dynamic behavior of pressure
support noninvasive ventilation (PSV applied noninvasively via a facemask), focusing on the consequences of a leak to atmosphere proximal to the airway opening (3, 4). Both mathematical analysis and laboratory investigation using a mechanical analog of the respiratory system indicated that, in the setting of an airway leak and airflow obstruction, PSV can display unstable dynamic behavior, characterized by wide swings in inspiratory time
and autoPEEP
despite unvarying patient effort. In this study, we extend our previous analysis to address dynamic stability of PSV in the absence of an airway leak, and to investigate the influence of "noise"corresponding to measurement errors in
peak flow or flow cutoffon system stability.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Mathematical Model
We analyzed four mathematical models of PSV that are detailed in the online data supplement: (1) linear (simple) PSV, (2) noisy PSV, (3) nonlinear PSV, and (4) variable frequency (Table 1). The models all assume that airway pressure rises abruptly to a constant inspiratory pressure (Pset) until the total flow applied by the ventilator falls to a predetermined fraction of peak flow ("flow cutoff," denoted by K); the applied pressure then cycles to the set value of CPAP. Patient effort is assumed constant and negligible. Two models incorporate pressure support ventilation applied at a fixed frequency to a single compartment having a constant compliance C, inspiratory resistance RI, and expiratory resistance RE. The first model ("simple PSV," Model 1) is the simplest possible representation of PSV. In the second model ("noisy PSV," Model 2), a random error term in the flow cutoff parameter (equivalent to a proportionate error in the sensed peak flow) is imposed on each breath. The random error term varies from breath to breath, but has a fixed and predetermined maximum value. In the third model ("nonlinear PSV," Model 3), the flow resistive pressures were assumed to vary according to the Rohrer equation:
|
(1) |
|
In the fourth model ("variable frequency," Model 4), the instantaneous respiratory frequency is assumed to vary as a linear function of the inspiratory time of that breath, as was observed by Laghi and coworkers (5). Modeling details are provided in the online data supplement. For the work presented here, input variables were selected to simulate airflow obstruction.
Mechanical Model
A test lung model was used as a mechanical analog. A purpose-built solenoid valve system was used to trigger a commercially available ventilator commonly used in adult intensive care (Puritan Bennett 840; Nellcor Puritan Bennett, Lenexa, KS) at fixed frequencies over a range of ventilator settings and impedance parameters. Airway flow and test lung chamber pressures were continuously monitored. Further details may be found in the online data supplement.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Mathematical Predictions
Fixed frequency Model 1 predicts regions of unstable ventilator output and dynamic behavior within the clinical range of ventilator settings and impedance parameter values. Specifically, when the flow cutoff is set below a critical value, the model predicts that the system will cycle (or oscillate) from breath to breath among multiple values for tidal volume and end-expiratory pressure (Figure 1). Increasing set frequency or time constant (compliance × resistance) raises the critical value for flow cutoff (data not shown). Similar behavior was seen with many other impedance configurations and ventilator settings.
|
The mechanism underlying the observed instability is revealed when the predicted end-expiratory pressures (EEP) of consecutive breaths for a given configuration are examined (Model 1; Figure 2). At higher values for flow cutoff, the EEPs converge to a stable value. However, as the flow cutoff threshold is progressively lowered, a critical value for flow cutoff is encountered, beyond which sequential EEP values diverge. Because applied inspiratory pressure is fixed, divergent EEP levels engender variation in the inspiratory driving pressure gradient for the subsequent breath, causing variation in sequential tidal volumes and still wider oscillations of EEP. Model 1 thus represents a class of unstable linear systems. The divergence of EEP (minimum and maximum end-expiratory pressures) is bounded by the set PEEP below and inspiratory pressure above. In this model, reducing EEP by decreasing compliance, resistance, or frequency lowers the critical value for flow cutoff at which instability occurs (data not shown). These behaviors are governed by the iterative function relating the end-expiratory pressure values of successive breaths (see Appendix, online data supplement).
|
0.32, sequential
values of EEP converge.
As K is lowered below
0.32, sequential values
of EEP diverge, and the
system becomes profoundly unstable.
When a small random error is imposed on the value of the
flow cutoff, the system behavior changes dramatically (Model
2). As shown for EEP in Figure 3, small errors in flow cutoff
(corresponding to errors in sensed peak flow of 
7%) render
the system more unstable, even for a nominal flow cutoff that
leads to convergent behavior in the "noise-free" model. The
behavior remains bounded, but the range of EEP is large and
apparently aperiodic. Many other configurations behave in a
similar fashion. When the system is stable (high flow cutoff),
superimposed noise does not lead to unstable behavior of such
large magnitude. Similar results were obtained when random
noise was imposed on the sequential values for EEP or total
cycle time (data not shown).
|
Incorporation of a resistive pressure drop which varies according to the Rohrer equation also results in unstable behavior, with pronounced oscillation in tidal volume and end-expiratory pressure. When using a value of 2 cm H2O (s2/L2) for the leading coefficient of the (dV/dt)2 term (corresponding to an endotracheal tube of size 8 mm), oscillatory behavior developed at a higher value of flow cutoff than seen in the linear model, a situation that was worsened as the coefficient of the quadratic term was increased (Figure 4) (6). If flow across the resistive elements is assumed to be fully turbulent (a = 0) the system displays classic "period doubling" behavior, followed by a transition to chaotic behavior (data not shown) (7).
|
Incorporation of a variable frequency (Model 4) stabilizes the behavior observed in the fixed frequency model. As the frequency is made increasingly sensitive to inspiratory time, the variability in tidal volume and end-expiratory pressure declines, until the delivered support becomes stable and monotonous. Further increases in frequency responsiveness have no effect on system stability. However, this stabilizing effect is entirely overcome if the base frequency is increased slightly; for the impedance configuration presented in Figure 1, the variable frequency model is unstable within the clinical range of flow cutoff for base respiratory frequencies above ~ 22 breaths/min. An increased base frequency destabilizes the variable frequency model by attenuating the decreases in EEP afforded by inspiratory time (TI)-driven cycle prolongation, thereby reestablishing breath-to-breath dynamic coupling.
Mechanical Simulation
Nonuniform behavior was reproducibly observed in numerous physical simulations using input values well within the clinical range; two examples are presented in Figure 5. When a critical value for flow cutoff was reached, tidal volumes and end-expiratory pressures cycled through much wider ranges than those observed with flow cutoff values exceeding the critical level. The oscillatory behavior predicted by the model was also observed, and exacerbation of this behavior by increased resistance, compliance, or frequency was confirmed (data not shown). The critical value for flow cutoff varied depending on whether it was approached from higher values (system transitioning from a stable to an unstable state), or from below (system transitioning from an unstable to a stable state) (Figure 5). Directional dependence of the destabilization threshold was reproducible across and within many different impedance configurations and ventilator settings. The magnitude of the shift in the critical value was significant (3% to 20%); the direction of the shift (up or down) varied between configurations. Directional dependence of the destabilization threshold was also consistent with the predictions of Model 1 (see below). Finally, if the flow cutoff threshold was lowered sufficiently, the compartmental EEP became large enough to preclude uniform triggering with a triggering impulse of fixed magnitude, and "triggering failures" and dropped breaths were observed (data not shown).
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Interpretation
Our main findings are that pressure support ventilation in a
leak-free system (such as occurs with traditional mechanical ventilation delivered through a cuffed endotracheal tube) can develop dynamic instability in the setting of airway obstruction, and that system noisesuch as that arising from errors in measurement of peak flow or flow cutoffaccentuates this behavior. Instability is exacerbated in the setting of a flow-dependent
resistive element, and, as Model 4 demonstrates, may be partially ameliorated by patient feedback. The underlying mechanism for unstable behavior appears to be variability in the level
of end-expiratory alveolar pressure (EEP), which in a pressure-targeted system influences the tidal volume and end-expiratory
pressure of the subsequent breath. In these models, unstable
behavior does not arise from variation in patient effort or from
external factors, but is an inherent consequence of the underlying dynamics of this mode of ventilation. Theoretically predicted behaviors were confirmed in a mechanical model, which
also displayed directional variation in the critical value for instability. Observed instability is worsened by factors that reduce
expiratory time and increase end-expiratory alveolar pressure:
longer system time constants (compliance × resistance) and increased respiratory frequency (Table 2).
|
Our analysis differs from that of Yamada and Du (10), who addressed neural:mechanical synchrony of the expiratory trigger by modeling events during inspiration, and suggested that adjustment of the expiratory trigger sensitivity based on respiratory system mechanics may be necessary to attain neuromechanical synchrony. In contrast, we focused on purely dynamic effects arising from "feed forward" interaction between the inspiratory and expiratory limbs during pressure support ventilation, and have found that pressure support ventilation may become "unstable" independent of neural activity. The two studies complement each other.
The unstable behavior we observed in the "noise-free" mathematical models and the mechanical simulations may seem surprising, as all three systems are governed by a simple set of differential equations. However, these systems are recursive, or iterative, in nature when a series of breaths is examined. Oscillatory and even "chaotic" behavior of deterministic recursive mathematical systems has clear precedent in the physics and mathematics literature (7).
As shown in the Appendix (see online data supplement), the simplest model of PSV presented can be described as a recursive function in which the end-expiratory pressure of a given breath is a function of the preceding breath. System instability arises largely through modulation of the inspiratory pressure gradient for inflation (Pset-EEP), with inspiratory time remaining relatively stable. This stability of inspiratory time is in contrast to our previous work, where inspiratory time varied markedly; the difference may relate to the fact that our prior models of noninvasive ventilation incorporated a leak to atmosphere ("open system"), which allowed greater variability in inspiratory time (3, 4). In the presence of a mask leak, the rate of decline in inspiratory flow is slowed by flow through the mask leak, at times permitting much longer inspiratory phases. The "open" versus "closed" system distinction may also explain the different responses to variability in inspiratory time (variable frequency model).
The propensity for the system to develop oscillatory behavior (Figure 1) renders it sensitive to noise. When imposed on a dynamic system predisposed to oscillation, random noise of relatively small magnitude leads to severe system instability, a phenomenon previously described in the analysis of respiratory rhythm (11). It is important to note that the apparently chaotic behavior displayed by the noisy PSV model (Figure 2) actually represents a stochastic process in which the systems underlying proclivity for oscillation can be thought of as amplifying imposed noise.
The misbehavior observed in the nonlinear model is not unexpected, as the parent linear model is itself unstable. Incorporation of a nonlinear element into the linear model results in higher order polynomial terms in the recursion relationship, a phenomenon known to increase the propensity to instability (12).
Limitations of the Model
The mathematical models used are the simplest plausible
models of pressure support ventilation. The rate of rise of inspiratory airway pressure was assumed to be abrupt, untapered, and always uniform (an "ideal square wave"). Furthermore, the models are unicompartmental, and the complex
issue of viscoelastance is not addressed. The variable frequency model (Model 4) incorporates only a simple feedback
loop, albeit one that is derived from clinical data (5), and does
not address more detailed conscious or neuroregulatory modulation of the ventilatory pattern. The impedance values and ventilator settings we used in the examples are clearly arbitrary but reasonable analogs for the clinical setting. We assumed "perfect" triggering of the ventilator, with negligible
patient effort to initiate each breath, no contribution of patient inspiratory effort to inspiratory events, and no active expiratory efforts. Evidence suggests that patient effort is important both during triggering/inspiration and during expiration
in patients with chronic obstructive pulmonary disease (COPD)
(13, 14). Parenthetically, preliminary investigation of a model
including active patient triggering suggests that contributions from effort may worsen, not improve, instability. We did not
address the issue of sudden airway opening or closure, and
modeled the system with a constant compliance (15, 16). It is
possible that dynamic variation in effective compliancesuch
as that arising from sudden airway opening or closure, or variations in gas trappingcould further destabilize system dynamic behavior; this complex question is beyond the scope of
the current discussion. Finally, we did not incorporate higher
order kinetic interactions, such as inertance.
As is common practice in the physical sciences, we investigated simple models to ascertain qualitative behaviors and highlight potential areas of further investigation. Assessing the quantitative modulation of these qualitative behaviors by incorporating higher order nonlinearity (viscoelastance, multiple compartments, more complex patient feedback, inertance, etc.) requires significant assumptions regarding these complicating factors, which may not be universally valid in the clinical setting. Furthermore, the incorporation of viscoelastance may render the equations refractory to closed-form, readily tractable solutions.
We did not study these behaviors in patients. Although dynamic instability may theoretically contribute to patient:ventilator dyssynchrony, without further data we cannot attribute the irregular behavior observed in the clinical setting to the underlying dynamic processes we describe. MacIntyre and Ho (17) examined the effect of breath termination criteria on patient-ventilator interaction; however, their study focused on patients with lower compliances than modeled here. As noted above, lower compliance would be expected to obviate dynamic instability. When configured with the mean compliance value and ventilator settings reported by MacIntyre and Ho, the current model (linear PSV) predicts relatively stable dynamic behavior. When confronted with unstable ventilator output, the patient may respond by actively "overriding" the ventilator. As shown by Model 4, the natural variation in total cycle length observed empirically may reduce instability (5). Furthermore, at present it would be difficult to ascertain whether clinically observed instability arises from unstable patient behavior (due to ventilatory drive or agitation), from dynamic effects, or both. Our analyses indicate that effort-independent unstable behavior may exist; further characterization of the predicted dynamics is required before their role or importance in the clinical setting can be clearly delineated.
Clinical Implications
The implications of dynamic instability are unclear. Such interactions may cause mechanical dyssynchrony, manifest as
"dropped" or missed triggering attempts, mismatching of neural and mechanical inspiratory time, or unstable patient responses due to attempts to entrain a variable level of ventilatory support. Triggering failure likely due to elevated autoPEEP
has been observed in the clinical setting, and represents an extreme example of dynamic instability (18). In addition to
these purely mechanical consequences, dyssynchrony may
have adverse effects on sleep architecturea possibility consistent with the demonstration of increased arousals during pressure support, but not volume-cycled, ventilation (21, 22).
The potential adverse effects of patient-ventilator dyssynchrony must be weighed against evidence suggesting that "biologically variable ventilation" (BVV) improves oxygenation in
models of acute lung injurya finding felt to be due to more
effective lung recruitment arising from the phenomenon of stochastic resonance (23). Important differences exist between
the experimental models reported in those studies and the focus of the current analysis. The work of Mutch and coworkers
(23) focused on a recruitable model of acute lung injury,
with oxygenation, lung recruitment, and lung protection as primary concerns; a setting different from obstructive lung disease. It is possible that the salutary effects they observed may
be of less importance in obstructive lung disease. It is also possible that spontaneous "amplification" of respiratory pattern
variability arising from dynamic interactions between ventilator algorithm/settings and the specific impedance characteristics of a given patient may prove more effective in lung recruitment than imposition of an archetyped pattern.
Our experimental data indicate that the system may exhibit directional effects in the development of dynamic instability (Figure 5). When the mathematical model was configured to "use" the terminal EEP level for the preceding value of K as the initial end expiratory pressure for calculations at the subsequent value of K, similar directionality was observed (data not shown). This suggests that the end-expiratory pressure "memory" can determine the subsequent (and persistent) dynamic state of the system, thus affecting system behavior on a much longer time scale than seen with the breath-to-breath effects noted above. Such a phenomenon has been described in the biophysics and mathematics literature (8). Longer term, perhaps regional, "dynamic memory" arising from end-expiratory pressure may have implications for respiratory phenomena distinct from those we have examined, such as lung recruitment and regional lung unit emptying, in heterogeneous lungs (28, 29). Finally, because a given breath can influence subsequent breaths only through the "memory" of end-expiratory pressure, an expiratory time constant that is "long" relative to the total respiratory cycle time (as seen in obstructive disease) is crucial to the spontaneous development of instability in this model.
Similarly, the consequences of effectively amplifying the
natural variability of spontaneous breathing remain unknown.
Breathing patterns of normal subjects and diseased patients
with chronic obstructive lung disease are known to be highly
variable (30, 31). Although the potentially oscillatory dynamics of PSV may interact with concurrent or time-lagged respiratory control systems to amplify unstable behavior and promote dyssynchrony, such a hypothesis is currently speculative.
It is equally possible that a variable ventilatory patternprovided it remains contained within limitsmay facilitate overall patient tolerance of mechanical ventilation.
Conclusion
These results suggest that pressure support ventilation applied in the context of airway obstruction may be inherently unstable, leading to substantial breath-to-breath variation in autoPEEP and tidal volume. The predicted instabilities arise entirely independently of patient effort or volition. The unstable behavior is mediated by end-expiratory pressure, suggesting that it is most likely to occur when the respiratory system time constant is long relative to the ventilatory frequency, as in chronic obstructive lung disease. The experimental model also showed directional effects in the onset of instability. Unstable ventilatory support and dynamics could affect patient comfort directly, require active (patient initiated) termination of inspiration, or impose breath-to-breath variability in the effort required for inspiratory triggering. Unstable behavior was observed with impedance value/ventilator setting combinations that are clinically realistic. Our results suggest that further theoretical, laboratory, and clinical investigation of dynamic instability during pressure support ventilation, and other modes of partially supported ventilation, is warranted.
| |
Footnotes |
|---|
Correspondence and requests for reprints should be addressed to John R. Hotchkiss, Jr., M.D., Pulmonary and Critical Care Section, Regions Hospital, 640 Jackson Street, St. Paul, MN 55101. E-mail: John.R.Hotchkiss{at}HealthPartners.com
(Received in original form January 8, 2001 and accepted in revised form September 21, 2001).
John R. Hotchkiss is a Scientist Development Grantee of the American Heart Association. Supported in part by NIH SCOR 50152 and HealthPartners Research Foundation.This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org
Acknowledgments: The authors thank Peter L. Bliss, BME, for invaluable help in designing and constructing the solenoid system, as well as Li Chen, Ph.D., for assistance in data collection.
| |
References |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
1.
Macintyre NR.
Respiratory function during pressure support ventilation.
Chest
1986;
89:
677-683
2. Brochard L, Pluskwa F, Lemaire F. Improved efficacy of spontaneous breathing with inspiratory pressure support. Rev Respir Dis 1987; 136: 411-415 .
3.
Hotchkiss JR,
Adams AB,
Dries DJ,
Marini JJ,
Crooke PS.
Dynamical
behavior during noninvasive ventilation: chaotic support?
Am J Respir
Crit Care Med
2001;
163:
374-378
4. Crooke PS, Hotchkiss JR, Marini JJ. Linear and nonlinear models for non-invasive ventilation. Math Comput Modelling. (In press)
5.
Laghi F,
Karamchandani K,
Tobin MJ.
Influence of ventilator settings in
determining respiratory frequency during mechanical ventilation.
Am
J Respir Crit Care Med
1999;
160:
1766-1770
6. Wright PE, Marini JJ, Bernard GR. In vitro versus in vivo comparison of endotracheal tube airflow resistance. Am Rev Respir Dis 1989; 140: 10-16 [Medline].
7. May RM. Simple mathematical models with very complicated dynamics. Nature 1976; 261: 459-467 [Medline].
8. Olsen LF, Degn H. Chaos in biological systems. Q Rev Biophys 1985; 18: 165-225 [Medline].
9. Ott E. Strange attractors and chaotic motion of dynamical systems. Rev Modern Phys 1981; 53: 655-671 .
10.
Yamada Y,
Du H-L.
Analysis of the mechanisms of expiratory asynchrony in pressure support ventilation: a mathematical approach.
J
Appl Physiol
2000;
88:
2143-2150
11. Glass L, Graves C, Petrillo GA, Mackey MC. Unstable dynamics of a periodically driven oscillator in the presence of noise. J Theor Biol 1980; 86: 455-475 [Medline].
12. Hubbard JH, West BH. Differential equations: a dynamical systems approach. New York: Springer-Verlag; 1991.
13. Younes M. Patient-ventilator interaction with pressure assisted modalities of ventilatory support. Semin Respir Med 1993; 14: 299-322 .
14. Jubran A, Van De Graff WB, Tobin MJ. Variability of patient-ventilator interaction with pressure support ventilation in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995; 152: 129-136 [Abstract].
15. Guerin C, LeMasson S, de Varax R, Milic-Emili J, Fournier G. Small airway closure and positive end-expiratory pressure in mechanically ventilated patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1997; 155: 1949-1956 [Abstract].
16.
Gaver DP III,,
Samsel RW,
Solway J.
Effects of surface tension and viscosity on airway reopening.
J Appl Physiol
1990;
69:
74-85
17.
MacIntyre NR,
Ho LI.
Effects of initial flow rate and breath termination
criteria on pressure support ventilation.
Chest
1991;
99:
134-138
18.
Sottiaux T.
Patient-ventilator interactions during volume support ventilation: asynchrony and tidal volume instabilitya report of three
cases.
Respir Care
2001;
46:
255-262
[Medline].
19. Fabry B, Guttmann J, Eberhard L, Bauer T, Haberthur, Wolff G. An analysis of desynchronization between the spontaneously breathing patient and ventilator during inspiratory pressure support. Chest 1995;107:1387-1394.
20. Nava S, Bruschi C, Rubini F, Palo A, Iotti G, Braschi A. Respiratory response and inspiratory effort during pressure support ventilation in COPD patients. Intensive Care Med 1995; 21: 871-879 [Medline].
21. Parsatharathy S, Malik I, Jubran A, Tobin M. Does ventilator mode influence sleep quality in critically ill patients? Am J Respir Crit Care Med 2001; 165: A483 .
22. Parsatharathy S, Jubran A, Tobin M. Does sleep influence breathing pattern in mechanically ventilated patients? Am J Respir Crit Care Med 2001; 165: A302 .
23.
Mutch WA,
Harms S,
Graham MR,
Kowalski SE,
Girling LG,
Lefevre GR.
Biologically variable or naturally noisy mechanical ventilation recruits atelectatic lung.
Am J Respir Crit Care Med
2000;
162:
319-323
24. Lefevre GR, Kowalski SE, Girling LG, Thiessen DB, Mutch WA. Improved arterial oxygenation after oleic acid lung injury in the pig using a computer-controlled mechanical ventilator. Am J Respir Crit Care Med 1996; 154: 1567-1572 [Abstract].
25. Mutch WA, Harms S, Lefevre GR, Graham MR, Girling LG, Kowalski SE. Biologically variable ventilation increases arterial oxygenation over that seen with positive end-expiratory pressure alone in a porcine model of acute respiratory distress syndrome. Crit Care Med 2000; 28: 2457-2464 [Medline].
26. Suki B, Alencar AM, Sujeer MK, Lutchen KR, Collins JJ. Life support system benefits from noise. Nature 1998; 393: 127-128 [Medline].
27. Suki B, Barabasi AL, Hantos Z, Petak F, Stanley HE. Avalanches and power-law behavior in lung inflation. Nature 1994; 368: 615-618 [Medline].
28. Georgopoulos D, Mitrouska I, Markopoulou K, Patakas D, Anthonisen NR. Effects of breathing patterns on mechanically ventilated patients with chronic obstructive pulmonary disease and dynamic hyperinflation. Intensive Care Med 1995; 21: 880-886 [Medline].
29. Georgopoulos D, Gomez A, Mink S. Factors determining lobar emptying during maximal and partial forced deflations in nonhomogeneous airway obstruction in dogs. Am J Respir Crit Care Med 1994; 149: 1241-1247 [Abstract].
30.
Tobin MJ,
Chadha TS,
Jenouri G,
Birch SJ,
Gazeroglu HB,
Sackner MA.
Breathing patterns. 1. normal subjects.
Chest
1983;
84:
202-205
31.
Tobin MJ,
Chadha TS,
Jenouri G,
Birch SJ,
Gazeroglu HB,
Sackner MA.
Breathing patterns. 2. Diseased subjects.
Chest
1983;
84:
286-294
This article has been cited by other articles:
 |
|
 |
|
  E. Kondili, G. Prinianakis, and D. Georgopoulos Patient-ventilator interaction Br. J. Anaesth., July 1, 2003; 91(1): 106 - 119. [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Tobin Critical Care Medicine in AJRCCM 2002 Am. J. Respir. Crit. Care Med., February 1, 2003; 167(3): 294 - 305. [Full Text] [PDF]
|
|
|
|
|
|
S. Parthasarathy and M. J. Tobin Effect of Ventilator Mode on Sleep Quality in Critically Ill Patients Am. J. Respir. Crit. Care Med., December 1, 2002; 166(11): 1423 - 1429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Schonhofer and S. Sortor-Leger Equipment needs for noninvasive mechanical ventilation Eur. Respir. J., October 1, 2002; 20(4): 1029 - 1036. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Proc. Am. Thorac. Soc. | Am. J. Respir. Cell Mol. Biol. |