sign in sign up
<<
Home , Afterload, Hyperdynamic circulation, Preload (cardiology)

Cardiovascular Effects of Mechanical Ventilation

G. J. DUKE Intensive Care Department, The Northern Hospital, Epping, VICTORIA

ABSTRACT Objective: To review the cardiovascular effects of spontaneous breathing and mechanical ventilation in healthy and pathological states. Data sources: A review of articles published in peer-reviewed journals from 1966 to 1998 and identified through a MEDLINE search on cardiopulmonary interaction. Summary of review: Respiration has a hydraulic influence upon cardiovascular function. Pulmonary and cardiac pathology alter this interaction. Spontaneous inspiration increases right ventricular (RV) and left ventricular (LV) . Mechanical ventilation with positive pressure (MV) reduces LV preload and afterload. The influence of MV upon the cardiovascular system (CVS), particularly in critically ill patients, depends upon the mode of ventilation and the pre-existing cardiac and respiratory status. The influence of these factors is reviewed. Consideration of these parameters will enable the clinician to predict the likely effect of MV and develop strategies to minimise adverse events. Conclusions: Mechanical ventilation has an adverse effect upon the CVS in healthy subjects and in patients with pulmonary pathology, particularly in the presence of preload-dependent LV dysfunction or afterload-induced RV dysfunction. Mechanical ventilation may benefit cardiac function in patients with respiratory failure and afterload-dependent or exercise-induced LV dysfunction. (Critical Care and 1999; 1: 388-399)

Key Words: Mechanical ventilation, cardiovascular physiology, cardiopulmonary interaction, acute respiratory distress syndrome

Respiration and circulation are complementary on cardiac transmural pressure (Ptm) as the predominant physiological processes that interact with each other mechanism.1-3,5 during spontaneous breathing.1,2 The introduction of To appreciate the complex cardiovascular effects of mechanical ventilation (MV), or the presence of MV, especially in critically ill patients, it is prudent first pulmonary and cardiac disease, increases the complex- to identify the cardio-respiratory interactions of ity of this interaction. Research in this area is daunting spontaneous breathing in healthy subjects and in those and interpretation of the data is difficult because these with respiratory and/or cardiac disease, and then and other interrelated variables must be considered. consider the influence of MV in each situation. What is the predominant mechanism underlying this Consideration of cardiovascular and respiratory changes interaction? Explanations have included mechanical over time is also important. An understanding of these (hydraulic),1 neural,3 and humoral mechanisms.4 Phasic interactions should enable the clinician to predict the variation in cardiac function during respiration is closely likely cardiovascular effect of MV in a given clinical linked in time and magnitude to changes in intrathoracic situation. pressure and occurs more rapidly than most neural or humoral processes. Therefore current data support the Thoracic anatomy and transmural pressure hydraulic effect of intrathoracic (pleural) pressure (Ppl) Respiration induces phasic swings in cardiac trans-

Correspondence to: Dr. G. J. Duke, Intensive Care Department, The Northern Hospital, Epping, Victoria 3076 (e-mail: [email protected])

388 Critical Care and Resuscitation 1999; 1: 388-399 G. J. DUKE

Figure 1. Schematic representation of cardiorespiratory relationship. Paw = airway pressure, Palv = alveolar pressure, Ppl = intrathoracic pleural pressure, Pex = extramural stress, Pin = intramural stress, Ptm = transmural pressure, Psystole = aortic mural pressure (Ptm) as a result of the anatomical and Ppl is lowered by the respiratory muscles during functional proximity of respiratory and cardiovascular spontaneous inspiration and increased during the organs. Ptm is the difference between intramural stress application of positive pressure MV, and the resultant (Pin) and extramural stress (Pex), Ptm = Pin - Pex (Figure ∆Ppl is seen by the as a change in extramural 1). pressure (∆Pex). Ppl has a direct influence on cardiac 7,8 1,3 The thorax contains the and pulmonary (epicardial) Pex and thus an influence on LV and RV vasculature (divided into intra- and extra-alveolar volume and function. For example, a fall in Ppl will vessels) and the heart and great vessels (i.e. thoracic usually result in a fall in Pex (rise in Ptm) that will favour and great veins). The proximity of these organs ventricular filling but impede ejection. within the thorax, together with their dynamic The cardiovascular effects of respiration appear to mechanical properties (e.g. volume and elastance) be dependent upon both the magnitude of ∆Ptm and the ensures that changes in volume and Ppl are likely to sensitivity of the cardiovascular system to ∆Ptm. Thus influence cardiac function even during spontaneous the most important clinical variables in breathing. For example, an increase in lung volume cardio-pulmonary interactions include: 1) the produces a non-uniform compression of the lateral cardiovascular status of the subject; 2) the respiratory ventricular wall (Pex) and, although low in magnitude status of the subject, and; 3) the mode of respiration. 6 does influence ventricular function. Accurate measurement of Pex, Pin and ventricular The intra-thoracic cardiovascular system may be volume is important in the assessment of cardio- described as a dual series of pumps (right and left pulmonary interaction. Cardiac Pex is difficult to ventricles) separated from each other by the pulmonary measure accurately.6,9 The use of surrogate measures of vasculature, and from the systemic circulation by the Pex such as oesophageal or pleural pressure may great veins and thoracic aorta. Since the ventricles are in significantly underestimate Pex particularly in the series, the output of the right (RV) provides the presence of pulmonary disease or during positive input (venous return) for the left ventricle (LV) with the pressure MV. Pex varies across the surface of the heart, intervening pulmonary circulation producing a lag depending upon the volume and compliance of the time of (usually) 1-2 beats. adjacent lung, pericardium, and cardiac chamber.9

389 G. J. DUKE Critical Care and Resuscitation 1999; 1: 388-399

Indirect measurement of cardiac Pin (e.g. estimation of left ventricular end-diastolic Pin by catheter) may also be misleading.10-13 More direct measurements of Pex ,Pin, and chamber volume are preferred but usually restricted to animal models. Ventricular interdependence (VI) is another important physiological concept, anatomically based on the adjacent LV and RV sharing a common pericardial sac (with limited volume and compliance) and a common . VI is commonly used to imply that the pressure/volume characteristics of the RV influence those of the LV.7,14-16

Cardiovascular effects of spontaneous breathing To understand the cardiovascular changes resulting from MV it is important to first understand cardio- pulmonary interactions during spontaneous breathing. 2,3 This topic has been extensively reviewed elsewhere Figure 2. Schematic representation of cardiorespiratory interaction and whilst much of the data comes from animal (usually during spontaneous inspiration and cardiac . Arrows depict dog) models there is a substantial agreement with direction of change. Spontaneous inspiration increases Ptm and available human data.17 reduces Psystole. Paw = airway pressure, Palv = alveolar pressure, P = intrathoracic pleural pressure, P = extramural stress, Spontaneous inspiration in healthy subjects is usually pl ex Pin = intramural stress, Ptm = transmural pressure, Psystole = aortic associated with a small fall in systolic blood pressure (< blood pressure, LV = left ventricle l0 mmHg). Explanations for this observation include: 1) an increase in LV afterload, 2) a decrease in venous Under certain pathological states this inspiratory return, 3) the influence of VI, and, 4) transmission of blood pressure fall is exaggerated and referred to as 3,20,21 reduced intra-thoracic aortic Pex to extrathoracic vessels. ‘’. This may be found where there

Both animal and human data indicate that the first is: 1) a greater inspiratory effort and a greater ∆Ppl and 1- 22,23 24 explanation is probably the most important mechanism. ∆Ptm (e.g. acute asthma or pulmonary oedema ), or: 3,7 25 2) an increased sensitivity to ∆Ptm (e.g. hypovolaemia, 3,21 26 Inspiration occurs as a result of the reduction in Ppl tamponade, or congestive cardiac failure ). which is transmitted to the intra-thoracic organs. Whilst During exhalation systolic blood pressure rises as a the primary purpose of this pressure fall is to expand the result of the return of left ventricular afterload to lungs it is also results in a fall in cardiac Pex (and a rise baseline and aided by the augmentation of LV preload in Ptm). The increase in Ptm facilitates right ventricular from the (delayed effect of the) inspiratory rise in RV diastolic filling - the ‘thoracic pump’ mechanism - until output. the closing pressure of the extra-thoracic veins is There is little evidence to suggest other mechanisms 1,3 reached and the resultant increase in RV end-diastolic have a significant influence. The role of VI is volume (RVEDV) increases via the equivocal.1 Although there is little doubt that increasing 7,16,18 Frank-Starling mechanism. The subsequent rise in RV volume can induce leftward septal shift and increase pulmonary flow increases left ventricular end diastolic LVEDP, there is little data to support a significant 3,19 volume (LVEDV) after a delay of 1-2 beats - a influence during spontaneous breathing14,15 and duration similar to the inspiratory phase of the disagreement as to its clinical significance even in respiratory cycle. This may partially account for the rise pathological states. in blood pressure observed during expiration. Spontaneous breathing has little direct effect on the Spontaneous inspiration induces an increase in LV vasculature. Healthy subjects have a low pulmonary afterload that accounts for the observed inspiratory fall (PVR) which does not appreciably in stroke volume and systolic blood pressure. During alter during tidal breathing.2,27 There is also little inspiration left ventricular Pex, falls and Ptm, rises. Since evidence to suggest intra-pulmonary pooling of blood this ∆Ptm must be ‘overcome’ during systole it creates, during inspiration. Although the extra-alveolar by definition, an increase LV afterload. As a result, LV pulmonary increases, this is opposed by a stroke volume and systolic blood pressure are observed fall in intra-alveolar blood volume.1,28 Whilst inspiratory 1-3,5,6,16 to fall and LV end-systolic volume rises. (Figure changes in aortic Ptm are transmitted to the extra- 2)

390 Critical Care and Resuscitation 1999; 1: 388-399 G. J. DUKE thoracic vessels these are inadequate to explain the drive induce , , oliguria and observed respiratory swing in systolic blood pressure. retention of sodium and water. In summary, there are two predominant cardiovascular effects of spontaneous inspiration: 1) a rise in LV afterload and, 2) a rise in RV preload. An inspiratory fall in systemic blood pressure is produced by the former and limited by the latter. Factors that will potentially accentuate blood pressure fluctuation during spontaneous breathing include: 1) a fall in preload (e.g. hypovolaemia, venodilation), or; 2) a rise in left ventricular afterload as a result of increased inspiratory effort (e.g. bronchospasm3 or spontaneous breathing through an endotracheal tube29).

The cardiovascular effects of mechanical ventilation in healthy subjects Like spontaneous breathing, MV is associated with Figure 3. Schematic representation of cardiorespiratory interaction an inspiratory fall in aortic flow and systolic blood 2,30 during positive pressure inspiration and . Arrows depict pressure, but the mechanism is quite different. Unlike direction of change. Inspiration decreases venous return and RVEDV. spontaneous breathing, MV produces a positive ∆Ppl Paw = airway pressure, Palv = alveolar pressure, Ppl = intrathoracic during inspiration whose cardiovascular effects are the pleural pressure, Pex = extramural stress, Pin = intramural stress, Ptm = transmural pressure, CVP = , RA = right opposite to those seen with the negative ∆Ppl of atrium, RV = right ventricle, IVC = inferior vena cava spontaneous breathing. Proposed mechanisms for the blood pressure fall with MV include: There is little evidence to support a major influence 1) reduced LV preload, from ventricular interdependence (VI).9 While a 2) reduced RV preload, reduction in the LV septal-lateral wall dimension during 3) increased PVR and RV impedance, and MV has been reported by early investigators,14,39 recent 4) ventricular interdependence. investigators have either not demonstrated the presence Most of these mechanisms are present in varying of VI during MV,32,34 or found the reduction in LV degrees but the predominant effect is a reduction in LV dimensions were symmetrical, and no greater in the 5,31,32 preload. septal-free wall axis.9 Some even suggest that With the commencement of MV, the positive airway asymmetrical reduction in the septal-free wall axis pressure results in a fall in right ventricular P and a fall tm would also be consistent with increased lateral wall Pex in venous return (or loss of the ‘thoracic pump’), from the inflation of the adjacent lung and not from 18,32,33 causing a reduction in RVEDV (preload). . Positive VI.40,41 Finally, a number of researchers have airway pressure (Paw) is also transmitted to the demonstrated preservation of LV pressure-volume pulmonary vasculature inducing a (small) rise in relationships during MV implying that any effect from 9,42 pulmonary artery pressure (Ppa) and RV impedance VI is clinically insignificant. (afterload), and so with time RVEDV returns to the By maintaining some spontaneous inspiratory effort, 32 baseline. Although Ppa rises, PVR does not - unless assisted modes of ventilation such as CPAP, PSV, 34,35 intra-thoracic pressure or volume are unusually high. BiPAP and even SIMV, will tend to produce a lower Thus MV induces a fall in transpulmonary flow and mean Paw and Ppl compared with controlled modes of venous return to the LV, reducing LV preload (Figure MV for a similar level of minute volume. These modes 3). have been shown to reduce the adverse cardiovascular In contrast to the transient changes in RVEDV, MV effects of MV.18,43 However, assisted modes of produces a sustained reduction in LVEDV because both respiratory support are unsuitable if the patient has 36,37 38 afterload and preload are reduced. Both aortic severe lung disease or an inadequate respiratory drive pressure and fall progressively during (e.g. from sedation, anaesthesia, or coma) or is at risk inspiration and can be (partially or wholly) reversed by from respiratory muscle fatigue (e.g. a high respiratory 18,32 volume loading, suggesting the fall in preload is rate). more influential. Compensatory increases in sympathetic

391 G. J. DUKE Critical Care and Resuscitation 1999; 1: 388-399

Mechanical or extrinsic PEEP (PEEPe) is commonly some cases vasoconstrictor mediators (e.g. used to recruit alveoli, defend end-expiratory lung thromboxanes).30,45,46 PVR has an inverse hyperbolic volume (EELV), and improve oxygenation during MV. relationship with the lung volume (FRC)2 so that in PEEP increases mean Paw and (usually) Ppl and thus the severe cases, a marked rise PVR may induce RV haemodynamic effects are similar to MV. Some data systolic dysfunction.45 support the potential for humoral and neural mediated Acute respiratory failure resulting from extensive effects of PEEP on cardiac function in addition to its alveolar collapse (e.g. post-operative respiratory failure) hydraulic effects.4,9 The concept of ‘best PEEP’ is based may necessitate the use of MV for respiratory support. upon the balance between the respiratory benefits of In this setting MV may increase both Ppa and PVR PEEP and its adverse cardiovascular (and respiratory) through hyperinflation of healthy lung units.9,30,34 The 44 effects. use of PEEPe to recruit collapsed lung units may Factors that may accentuate the haemodynamic exacerbate these effects.47-49 The adverse haemodynamic effects of MV in healthy subjects include: effects of MV seen in healthy subjects are thus 1) a decrease in venous return25,38 (e.g. hypovolaemia, accentuated in these patients and strategies to recruit and venodilation), defend EELV tend to oppose those strategies which 2) an increase in mean intra-thoracic pressure (e.g. the minimise the adverse cardiovascular effects of MV.2,9 use of large tidal volume or high PEEPe), or MV strategies to defend cardiac output in the face of 3) blunting of compensatory sympathetic reflexes (e.g. reduced lung elastance and volume include: anaesthesia, sedatives). 1) attempts to minimise any unnecessary elevation in Strategies to minimise these effects include: Ppl that may occur with high tidal volume (Vt), Ve, 1) the use of volume expansion to restore LV preload, and inspiratory flow rates, and the judicious use of 2) the use of assisted modes of ventilation to reduce PEEPe, ∆Ppl, and 2) alveolar recruitment and reversal of hypoxia and 3) the avoidance of high mean intra-thoracic pressure acidosis in an effort to reduce PVR, (Ppl) that may occur with a high minute volume, high 3) careful use of volume expansion to improve LV inspiratory flow or PEEPe. preload but avoid pulmonary oedema, and, 4) use of inotropic agents to support cardiac output. Cardio-pulmonary interaction in the presence of In theory, the use of assisted non-invasive modes of pulmonary disease ventilation (e.g. CPAP, BiPAP, PSV) to recruit alveoli Pulmonary disease alters lung mechanics in a and provide respiratory support whilst minimising the number of ways which alter the cardio-pulmonary adverse cardiovascular effects of MV is attractive.50 interactions observed during spontaneous breathing. The However, there is little data to support a successful important changes in lung mechanics include those therapeutic role for this modality in these patients.43 affecting lung volume, airflow resistance, minute volume (Ve), work of breathing (WB), and RV Airflow limitation and dynamic hyperinflation impedance. Airflow limitation (e.g. asthma, COPD) prolongs the expiratory time and opposes alveolar deflation resulting Reduced lung volume in an increase in EELV (i.e. dynamic hyperinflation) and Many pulmonary diseases are associated with a the creation of intrinsic PEEP (PEEPi). Expiratory Ppl reduction in lung compliance and volume. This may and Pex will rise as a result of PEEP. These mechanical occur as a result of 1) bronchial obstruction (e.g. effects reduce LV (and RV) preload.22,51 inflammation, secretions), 2) an increase in closing Dynamic hyperinflation and inspiratory airflow volume (e.g. lung disease, elderly), 3) a reduction in limitation both increase the inspiratory effort to maintain functional residual capacity (FRC, e.g. anaesthesia, alveolar ventilation. The elevated lung elastance of high supine posture, abdominal and thoracic trauma), or 4) an lung volume, together with high inspiratory flow increase in lung elastance (eg. pulmonary oedema, resistance, increases work and the ∆Ppl required to pneumonia, ARDS). maintain alveolar ventilation. Thus the LV afterload

A decrease (or an increase) in lung volume will effect (of inspiratory ∆Ppl) is exaggerated in the increase PVR and RV impedance (i.e. afterload).2,27,30 A presence of airflow limitation, producing a greater fall in pathological increase in lung volume most often occurs systolic blood pressure and ‘pulsus paradox-us’.22,51 in the presence of airflow limitation (see later). A A number of other factors may reduce RV output reduction in lung volume increases the resistance in and trans-pulmonary flow in the presence of airflow extra-alveolar pulmonary vessels1,27 due to hypoxic limitation. Although heightened inspiratory effort will vasoconstriction,35,45 structural distortion,1,45 and in tend to increase RV preload, venous return is limited by

392 Critical Care and Resuscitation 1999; 1: 388-399 G. J. DUKE

2,3 the collapse of the extra-thoracic veins. PEEPi will may result from an elevation in metabolic rate or further decrease venous return. Over-inflation of alveoli deterioration in gas exchange. Increased WB may result during inspiration compresses intra-alveolar vessels and from increased lung elastance or airflow limitation.62 (along with other changes in pulmonary vasculature51) The metabolic (e.g. oxygen) cost of breathing - normally increases PVR and RV impedance.2,23,35 only 1-2% of total body oxygen consumption - may rise 61,65 With severe airflow obstruction the reduction in Ppl to as much as 20% in acute respiratory failure. 23 and Pex may be so severe as to cause pulmonary and Respiration may thus be viewed as a form of (cardiac 30 aortic flow to fall dramatically, or hydrostatic and respiratory) exercise stress. A high level of WB pulmonary oedema to occur.52 Alternatively, dynamic may produce respiratory muscle fatigue and cause acute 51 66-68 hyperinflation and airflow limitation may increase WB respiratory failure. Respiratory muscle blood flow to the extent that respiratory fatigue leads to hypercarbia may be significantly limited in the presence of cardiac and a respiratory arrest. These mechanisms may account failure, further exacerbating respiratory muscle fatigue, for unexplained sudden death in some asthmatics.53 lactic acidosis and respiratory failure.69 In addition, the Great care must be taken when instituting MV in the increased metabolic demand of respiration may induce presence of severe airflow limitation as MV may further myocardial ischaemia.70,71 In this setting, MV and 54 increase dynamic hyperinflation and Ppl. Even in the muscle relaxation may reduce some of the adverse presence of mild airflow limitation, the use of MV metabolic and cardiovascular stresses of spontaneous parameters which inadvertently shorten expiratory time breathing and may improve outcome. 65,69,72-75 (e.g. large Vt, high frequency, prolonged inspiratory time) may produce dynamic hyperinflation. Increased right ventricular impedance Monitoring dynamic hyperinflation during MV is not Pulmonary (PHT) is a frequent difficult. With appropriate volume-measuring devices, complication of acute and chronic pulmonary disease. the trapped volume (Vei) of dynamic hyperinflation can Particular mention is warranted here because severe be measured during a prolonged expiratory pause (30-60 PHT (mean Ppa > 30mmHg) may precipitate acute right seconds). Data from Tuxen and Lane54 suggests that a .2,45,76,77 This may arise in such diverse Vei < 20 mL/kg should minimise the adverse conditions as pulmonary embolism, acute exacerbation haemodynamic effects of MV in the presence of severe of chronic lung disease, or the acute respiratory distress airflow limitation. syndrome (ARDS). Alternatively, alveolar pressure (Palv) may be The compliant, thin-walled RV and pulmonary measured. In the presence of airflow limitation, Palv circulation usually operate as a low-pressure flow cannot be inferred from (upstream) Paw since the later generator. PHT increases RV impedance (i.e. afterload), will be higher than Palv during inspiration, and lower induces acute RV dilatation and, when severe, RV during exhalation.55 By temporarily adding an failure.76 This may significantly reduce pulmonary flow inspiratory or expiratory pause (0.5-1.0 seconds) Paw (i.e. LV preload) and precipitate systemic hypotension. 49 will more accurately indicate peak inspiratory Palv and The initial effects of MV will exacerbate this situation. end-expiratory Palv (or PEEPi) respectively. The transmission of positive Palv via the intra-alveolar Strategies to minimise the haemodynamic effects of vessels will increase mean Ppa whilst compression of MV in patients with mild airflow limitation include the intra-alveolar vessels will increase PVR. In the use of, long-term, however, MV may beneficially reduce PVR 1) assisted modes of ventilation to reduce threshold through the recruitment of collapsed lung units and work, inspiratory effort and improve minute reversal of acidotic and hypoxic pulmonary volume,43,56 vasoconstriction.30,45,77 2) MV parameters which reduce the inspiratory to The combination of systemic hypotension and pulm- expiratory time (I:E) ratio, and onary hypertension has an adverse effect on RV 57,58 78,79 3) avoidance of PEEPe in excess of PEEPi. myocardial oxygen supply and demand. RV However, in the presence of severe airflow limitat- myocardial pressure (PRVM) is determined by ion it is often necessary to monitor trapped lung volume the balance of the driving pressure (i.e. mean arterial (Vei) and PEEPi, and to use a very low I:E ratio with pressure) and the opposing mean ventricular pressure 54,59,60 controlled hypoventilation. (Pin,RV). The critical level for PRVM is around 25- 40mmHg, below which right ventricular ischaemia and 79 Increased work of breathing (WB) acute right heart failure may occur. Most pulmonary diseases are associated with an increase in minute volume (Ve) and/or an increase in 30,61 respiratory effort per breath (WB). High Ve demand PRVM = Pin,RV - MAP

393 G. J. DUKE Critical Care and Resuscitation 1999; 1: 388-399

Where cardiac function (see previously). Counteracting this is Pin,RV = 1/3 (Ppa,sys - CVP) + CVP the fact that non-distensible (i.e. collapsed or Ppa,sys = systolic pulmonary artery pressure consolidated) lung segments will ‘protect’ Ppl and 8,34,83 MAP = cardiac Ptm from the elevation in Paw. CVP = central venous pressure. Right ventricular dysfunction during MV is un- avoidable due to the combined effects of positive Paw Strategies to avoid RV ischaemia include defence of and Ppl and the presence of PHT (which is often severe) RV preload (with volume loading), defence of MAP and an elevated PVR.46 Not surprisingly, the 79,80 (with pressor agents) and monitoring of PRVM (e.g. cardiovascular effects of MV for ARDS are via a pulmonary artery catheter). Unfortunately most predominantly borne by the RV.82,84 RV dysfunction, inotropic agents also have some pulmonary elevated PVR and VI (if present), will all reduce LV vasoconstrictor properties, and those with significant preload. vasodilatory properties (e.g. isoprenaline, dobutamine The effects of ARDS on LV function are determined and ) may be best avoided. Noradrenaline by the combined effect of reductions in preload, appears to have the best profile for the RV by producing afterload and contractility.85 Contractility may be less pulmonary vasoconstriction for similar levels of impaired by humoral myocardial depressant factors and inotropic effect and support of myocardial perfusion.80,81 myocardial ischaemia.46 Peripheral vascular failure (systemic vasodilatation) reduces LV afterload and Acute lung injury (ALI), ARDS and MV favours a hyperdynamic circulation so that the clinical Special mention of MV in acute lung injury (ALI) observation of a warm periphery and a bounding and ARDS is warranted because the complexity of the may mask the presence of significant LV dysfunction. cardio-respiratory interaction of MV with lung disease is The addition of MV has little cardiovascular benefit epitomised in a patient with ARDS. The dynamic in ARDS.86 Strategies to minimise the haemodynamic interaction of pathological processes and pulmonary effects of MV tend to oppose many of the ventilatory mechanics demonstrates that care should be taken to strategies used to overcome the pulmonary dysfunction identify these variables when interpreting clinical and of ARDS. Where possible MV parameters which research data. minimise mean Paw and total PEEP (PEEPe plus PEEPi) The pathological changes in pulmonary and will favour cardiac function as well as limiting cardiovascular function with ALI and ARDS are ventilator-induced lung injury. Maintenance of RV documented elsewhere.46,82 In summary, all the preload (volume loading), RV contractility (inotropic mechanical factors discussed previously are seen in agents) and systemic blood pressure (pressor agents), various degrees with ALI and ARDS. For example, together with frequent clinical and haemodynamic increased elastance, alveolar collapse, loss of EELV, assessment (e.g. pulmonary artery catheter, airflow limitation, , and elevated , etc.) are essential. Clinical goals will metabolic rate and WB. Humoral inflammatory often need to be guided by response to these therapeutic mediators produce pulmonary vasoconstriction, challenges and will differ for each patient depending on myocardial depression, and systemic hypotension. the severity and progression of the disease. Control of oxygen flux, organ blood flow and cellular The elusive search for an ideal MV mode for ARDS respiration may also be deranged. patients - which supports ventilation without adverse A number of different modes of MV may be utilised cardiac (and respiratory) side effects - has included the depending upon the stage and severity of ALI. These use of high frequency ventilation,87 inverse ratio vary from non-invasive modes (e.g. face-mask PSV and ventilation,89 extra-corporeal oxygenation,90 partial CPAP) for ALI, through to assisted modes (e.g. SIMV liquid ventilation,91 inhaled pulmonary vasodilators (e.g. and PSV) and to the extremes of (pressure or volume) nitric oxide92 and prostaglandins93), prone ventilation, 74 controlled modes with significant levels of PEEPe being phasic abdomino-thoracic pressure support, and extra- required for ARDS. thoracic negative-pressure ventilators.94,95 Many of these ARDS is characterised by markedly increased lung options are worthy of consideration in the presence of elastance so that during MV, a greater inspiratory severe cardiovascular dysfunction in response to MV airway pressure (Paw,i) is required to maintain adequate but evidence of improved outcome is lacking. alveolar ventilation, and high levels of PEEPe are often utilised to prevent airway collapse and aid recruitment Cardio-pulmonary interaction in the presence of of alveoli. Healthy lung units exposed to high Paw,i will cardiac disease be subject to overdistension. To the extent that Paw is Cardiovascular pathology is a common finding in transmitted to Ppl, MV will decrease Ptm and impair patients who require MV. Some of these factors have

394 Critical Care and Resuscitation 1999; 1: 388-399 G. J. DUKE already been mentioned. The effect of MV is MV may be beneficial in these patients by reducing 65 particularly dependent upon the type and severity of metabolic demand and WB and by favourably altering cardiac dysfunction. When ventricular dysfunction is LV mechanics (e.g. decrease the afterload). Consistent preload-dependent (e.g. hypovolaemia, ischaemia, with this hypothesis is the observation that patients with restrictive cardiomyopathy, tamponade, and valvular stress-induced ischaemia may experience difficulty in stenoses), MV will generally cause a further reduction in weaning from supported (MV) to spontaneous cardiac output. Poor ventricular compliance (resulting breathing.71,72,99,100 Even long-term nocturnal CPAP in from myocardial ischaemia, hypertrophy or fibrosis) is a severe cardiac failure patients (e.g. New York Heart particular situation that often causes a rise in Pin and Ptm Association class III & IV) has been shown to produce a but not preload (LVEDV). General measures to prevent sustained improvement in LV and respiratory reserve.101 hypotension include the use of MV parameters that In subjects with afterload-induced RV dysfunction minimise mean Paw together with volume loading and, (e.g. severe pulmonary hypertension, acute PE, COPD, when necessary, the use of inotropic agents. or RV infarct) MV may also adversely effect the balance When left ventricular dysfunction is afterload- of RV oxygen supply and demand. As discussed above, dependent, MV may improve cardiac output. Two treatment of reversible pulmonary vasoconstriction (e.g. important groups are worthy of mention: 1) cardiac- from hypoxia or acidosis) and defence of coronary induced acute respiratory failure (e.g. cardiogenic perfusion pressure with pressor agents may be pulmonary oedema), and, 2) respiratory-induced acute beneficial. cardiac failure in the presence of severe systolic dysfunction. Acute cardiogenic pulmonary oedema most commonly occurs as a result of ischaemic LV diastolic dysfunction leading to a hydraulic shift of fluid from the intravascular compartment into the lung parenchyma.96 Increased inspiratory effort (WB) produces a large 70 negative Ppl that increases LV afterload and may lead to respiratory muscle fatigue68,75 and respiratory acidosis. Myocardial ischaemia and afterload impair LV systolic function and cardiac output, initiating a compensatory sympathetic response. RV dysfunction may also occur as a result of adverse changes in afterload (e.g. hypoxic pulmonary vasoconstriction), preload (e.g. hypovolaemia), and myocardial oxygen supply (e.g. coronary hypoperfusion). MV is useful in cardiogenic pulmonary oedema because it reverses hypoxia, reduces metabolic demand 72 Figure 4. Schematic representation of the effect of positive pressure and improves afterload (Figure 4). Many subjects with inspiration in the presence of systolic dysfunction. Arrows depict acute cardiogenic pulmonary oedema are hypovolaemic direction of change. Inspiration decreases afterload (Ptm) and and due consideration must be given to the adverse increases (Psystole). Paw = airway pressure, Palv = preload effects of MV.97 Noninvasive modes of assisted alveolar pressure, Ppl = intrathoracic pleural pressure, Pex = extramural stress, Pin = intramural stress, Ptm = transmural pressure, LV = left breathing (e.g. face-mask CPAP) provide high inspired ventricle oxygen, reduce WB and improve LV afterload and have been shown to be equally effective as MV, with lower a rate of intubation, compared with pharmacological Summary therapy alone.98 MV has significant haemodynamic side-effects, the Respiratory-induced acute cardiac failure occurs in nature of which depend upon the cardiac and respiratory the presence of severe systolic dysfunction (e.g. severe status of the subject, the mode of ventilatory support and cardiomyopathy or ischaemic heart disease). Since the ventilatory parameters chosen. In general, MV spontaneous breathing is a form of exercise30 an increase reduces RV and LV preload and improves LV afterload. in WB during respiratory failure will increase the MV will increase the risk of adverse cardiovascular metabolic (oxygen) demand of the respiratory muscles effects in the presence of acute or chronic pulmonary by as much as ten-fold.68,69,75 If myocardial reserve is disease, especially in association with preload- insufficient to meet this demand acute myocardial dependent LV function, or afterload-induced RV ischaemia or systolic failure may occur.

395 G. J. DUKE Critical Care and Resuscitation 1999; 1: 388-399 dysfunction. Optimal MV modes and parameters to 15. Maughan WL, Kallman CH, Shoukas A. The effect of minimise this will depend upon the pulmonary right ventricular filling on the pressure-volume pathophysiology present, and may change over time in relationship of the ejecting canine left ventricle. Circ the same subject. Conversely, MV may benefit cardiac Res 1981;49:382-388. 16. Robotham JL, Lixfeld W, Holland L, MacGregor D, function in patients with respiratory failure associated Bryan AC, Rabson J. Effects of respiration on cardiac with afterload-dependent or exercise-induced LV performance. J Appl Physiol 1978;44:703-709. dysfunction. 17. Scharf SM, Brown R, Tow DE, Parisi AF. Cardiac effects of increased lung volume and decreased pleural pressure in man. J Appl Physiol 1979;47:257-262. Received 22.2.99 18. Pinsky MR. Determinants of pulmonary arterial blood Accepted 18.10.99 flow variation during respiration. J Appl Physiol 1984;56:1237-1245. REFERENCES 19. Guntheroth WG, Gould R, Butler J, Kinnen E. Pulsatile 1. Permutt S. Wise RA. Handbook of Physiology - The flow in pulmonary artery, capillary and vein in the dog. Respiratory System III. Chapter 36 Mechanical Cardiovasc Res 1974;8:330-337. interaction of respiration and circulation. Oxford 20. Shabetai R. Fowler NO, Gueron M. The effects of University Press; p647-656. respiration on aortic pressure and flow. Am Heart J 2. Pinsky MR. Heart Lung Interactions. Ch 29. 1963;65:525-533. Pathophysiologic Foundations of Critical Care. 1993. 21. Shabetai R, Fowler NO, Fenton JC, Masangkay M. Editors: Pinsky MR, Dhainaut J-F. Publisher: Williams Pulsus paradoxus. J Clin Invest 1965;44:1882-1898. & Wilkins. 22. Jardin F, Farcot JC, Boisante L, Prost JF, Gueret P, 3. Wise RA, Robotham JL, Summer WR. Effects of Bourdarias JP. Mechanism of paradoxical pulse in spontaneous ventilation on the circulation. Lung bronchial asthma. Circulation 1982;66:887-894. 1981;159:175-186. 23. Jardin F, Dubourg O, Margairaz A, Bourdarias JP. 4. Grindlinger GA, Manny J, Justice R, Dunham B, Shepro Inspiratory impairment in right ventricular performance D, Hechtman HB. Presence of negative inotropic agents during asthma. Chest 1987;92:789-795. in canine plasma during positive end-expiratory 24. Sharp JT, Griffith GT, Bunnell IL et al. Ventilatory pressure. Circ Res 1979;45:460-467. mechanics in in man. J Clin Invest 5. Scharf SM, Brown R, Saunders N, Green LH. 1957;35:111-117. Hemodynamic effects of positive pressure inflation. J 25. Perel A, Pizov R, Cotev S. Systolic blood pressure Appl Physiol 1980;49:124-131. variation is a sensitive indicator of hypovolaemia in 6. Lloyd TC. Respiratory system compliance as seen from ventilated dogs subject to graded haemorrhage. the cardiac fossa. J Appl Physiol 1982;53:57-62. Anesthesiology 1987;67:498-502. 7. Summer WR, Permutt S, Sagawa K, Shoukas AA, 26. Naughton MT, Rahman MA, Hara K, Floras JS, Bradley Bromberger-Barnea B. Effects of spontaneous TD. Effect of continuous positive airway pressure on respiration on canine left ventricular function. Circ Res intrathoracic and left ventricular transmural pressures in 1979;45:719-728. patients with congestive heart failure. Circulation 8. Robotham JL, Mitzner W. A model of the effects of 1995;91:1725-1731. respiration on left ventricular performance. J Appl 27. Hughes JM, Glazier JB, Maloney JE, West JB. Effect of Physiol 1979;46:411-418. extra-alveolar vessels on distribution of blood flow in 9. Smith PK, Tyson GS, Hammon JW, et al. the dog lung. J Appl Physiol 1968;25:701-712. Cardiovascular effects of ventilation with positive 28. Brower R, Wise R, Hassapoyannes C, et al. The effect of airway pressure. Ann Surg 1982;195:121130. lung inflation on pulmonary venous return. Federation 10. Raper RF, Sibbald WJ. Misled by the Wedge? The Proc 1983;42:594. Swan-Ganz Catheter and Left Ventricular Preload. Chest 29. Bersten AD, Rutten AJ, Vedig AE, Skowronski GA. 1986;89:427-435. Additional work of breathing imposed by endotracheal 11. Calvin JE, Driedger AA, Sibbald WJ. The hemodynamic tubes, breathing circuits, and intensive care ventilators. effect of rapid fluid infusion in critically ill patients. Crit Care Med. 1989;17:671-677. Surgery 1981;90:61-76. 30. Pinsky MR. Clinical applications of cardiopulmonary 12. O'Quin R. Marini JJ. Pulmonary artery occlusion 1997;48:587-603. pressure: Clinical physiology, measurement and 31. Morgan BC, Martin WE, Hornbein TF, Crawford EW, interpretation. Am Rev Resp Dis 1983;128:319-326. Guntheroth WG. Hemodynamic effects of intermittent 13. Pinsky MR, Vincent J-L, De Smet J-M. Estimating left positive pressure respiration. Anesthesiology ventricular filling pressure during positive end- 1966;27:584-590. expiratory pressure in humans. Am Rev Respir Dis 32. Rankin JS, Olsen CO, Arentzen CE, et al. The effects of 1991;143:25-31. airway pressure on cardiac function in intact dogs and 14. Taylor RR, Covell JW, Sonnenblick EH, Ross J Jr. man. Circulation 1982;66:108-120. Dependence of ventricular distensibility on filling of the opposite ventricle. Am J Physiol 1967;213:711-718.

396 Critical Care and Resuscitation 1999; 1: 388-399 G. J. DUKE

33. Cournand A, Motley HL, Werko L, et al. Physiological influence of ventilation with PEEP. Anesthesiology studies of the effect of breathing on cardiac output in 1990;73:964-975. man. Am J Physiol 1947;149:162-174. 50. Bradley TD, Holloway RM, McLaugh. Cardiac output 34. Culver BH, Marini JJ, Butler J. Lung volume and response to continuous positive airway pressure in pleural pressure effects on ventricular function. J Appl congestive heart failure. Am Rev Respir Dis Physiol 1981;50:630-635. 1992;145:377-382. 35. Hakim TS, Chang HK, Michel RP. Pressure flow 51. Jardin F. Bourdarias J-P. Acute Asthma. Ch30. relationships in the pulmonary circulation during lung Pathophysiologic Foundations of Critical Care. 1993. inflation. Federation Proc 1983;42:594. Editors: Pinsky MR, Dhainaut J-F. Publisher: Williams 36. Buda AJ, Pinsky MR, Ingels NB Jr, Daughters GT 2d, & Wilkins. Stinson EB, Alderman EL. The effect of intrathoracic 52. Stalcup SA, Mellins RB. Mechanical forces producing pressure on left ventricular performance. N Engl J Med pulmonary edema in acute asthma. New Engl J Med 1979;301:453-459. 1977;297:592-596. 37. Fessler HE, Brower RG, Wise RA, Permutt S. 53. Benatar SR. Fatal asthma. New Engl J Med. Mechanism of reduced LV afterload by systolic and 1986;314:423-429. diastolic pleural pressure. J Appl Physiol 1988 65:1244- 54. Tuxen DV, Lane S. The effects of ventilatory pattern on 1250. hyperinflation, airway pressures, and circulation in 38. Beaussier M, Coriat P. Perel A, et al. Determinants of mechanical ventilation of patients with severe airflow systolic pressure variation in patients ventilated after obstruction. Am Rev Repir Dis 1987;136:872-879. vascular surgery. J Cardiothor Vasc Anesth 1995;9:547- 55. Valta P, Corbeil C, Chasse M, Braidy J, Milic-Emili J. 551. Mean airway pressure as an index of mean alveolar 39. Jardin F, Farcot JC, Boisante L, Curien N, Margairaz A, pressure. Effect of expiratory flow limitation. Am J Bourdarias JP. Influence of positive end-expiratory Respir Crit Care Med 1996;153:1825-1830. pressure on left ventricular performance. N Engl J Med 56. Keenan SP, Kernerman PD, Cook DJ, Martin CM, 1981;304:387-392. McCormack D, Sibbald WJ. Effect of noninvasive 40. Cassidy SS, Mitchell JH, Johnson RL Jr. Dimensional positive pressure ventilation on mortality in patients analysis of right and left ventricles during positive admitted with acute respiratory failure: a meta-analysis. pressure ventilation in dogs. Am J Physiol Crit Care Med 1997;25:1685-1692. 1982;242:H549-H556. 57. Marini JJ. Should PEEP be used in airflow obstruction. 41. Cassidy SS, Ramanathan, M. Dimensional analysis of Am Rev Respir Dis 1989;140:1-3 the left ventricle during PEEP: relative septal and lateral 58. Tobin MJ, Lodato. PEEP, auto-PEEP, and waterfalls. wall displacements. Am J Physiol 1984;246:H792- Chest 1989;96:449-451. H805. 59. Dariloi R. Peret C. Mechanical controlled 42. Fewell JE, Abendschein DR, Carlson CJ, Rapaport E, hypoventilation in status asthmaticus. Am Rev Respir Murray JF. Continuous positive-pressure ventilation Dis 1984;129:385-387. does not alter ventricular pressure-volume relationship. 60. Tuxen DV. Detrimental effects of positive Am J Physiol 1981;241:H821-H826. end-expiratory pressure during controlled mechanical 43. Duke GJ, Bersten AD. Non-invasive ventilation in adult ventilation of patients with severe airflow obstruction. acute respiratory failure. Part I. Critical Care & Am Rev Respir Dis 1989;140:5-9. Resuscitation. 1999;1:187-198. 61. Field S. Kelly SM, Macklem PT. The oxygen cost of 44. Grace MP, Greenbaum DM. Cardiac performance in breathing cardiorespiratory disease. Am Rev Respir Dis response to PEEP in patients with cardiac dysfunction. 1982;126:9-13. Crit Care Med 1982;10:358-360. 62. Robertson CH, Foster GH, Johnson RL. The 45. Rattes M. Calvin JE. Acute Pulmonary Hypertension. relationship of respiratory failure to the oxygen Ch 19. Pathophysiologic Foundations of Critical Care. consumption of, lactate production by, and distribution 1993. Editors: Pinsky MR, Dhainaut J-F. Publisher: of blood flow among respiratory muscles during Williams & Willkins. increased inspiratory resistance. J Clin Invest 46. Bersten AD, Sibbald WJ. Acute lung injury in septic 1977;59:31-42. shock. Critical Care Clinics, 1989;5:49-79. 63. Viires N, Sillye G, Aubier M, Rassidakis A, Roussos C. 47. Raper RF, Sibbald WJ. Increased right ventricular Regional blood flow distribution in dog during induced compliance in response to continuous positive airway hypotension and low cardiac output. J Clin Invest pressure. Am Rev Respir Dis 1992;145:771-775. 1983;72:935-947. 48. Imai T, Uchiyama M, Maruyama N, Yoshikawa D, 64. Kanak R. Fahey PJ, Vanderwarf. Oxygen cost of Fujita T. Influence of constant sustained positive airway breathing. Changes dependent upon mode of mechanical pressure on right ventricular performance. Intensive ventilation. Chest 1985;87:126-127. Care Med 1993;19:8-12. 65. Manthous CA, Hall JB, Kushner R, Schmidt GA, Russo 49. Zwissler B. Forst H. Messmer K. Local and global G, Wood LD. The effect of mechanical ventilation on function of the right ventricle in a canine model of oxygen consumption in critically ill patients. Am J pulmonary microembolism and oleic acid edema: Respir Crit Care Med 1995;151:210-214.

397 G. J. DUKE Critical Care and Resuscitation 1999; 1: 388-399

66. Moxam J. Respiratory muscle fatigue: mechanisms, 82. Cook DJ, Raffin TA. Acute hypoxemic respiratory evaluation and therapy. BJA 1990;65:43-53. failure. Ch 25 Pathophysiologic Foundations of Critical 67. Cohen CA, Zagelbaum G, Gross D, Roussos C, Care. 1993. Editors: Pinsky MR, Dhainaut J-F. Macklem PT. Clinical manifestations of inspiratory Publisher: Williams & Wilkins. muscle fatigue. Am J Med. 1982;73:308-316. 83. Romand J-A, Shi W. Pinsky MR. Cardiopulmonary 68. Mancini DM, Henson D, LaManca J, Levine S. effects of positive pressure ventilation during acute lung Evidence of reduced respiratory muscle endurance in injury. Chest 1995;108:1041-1048. patients with heart failure. J Am Coll Cardiol 84. Sibbald WJ, Driedger AA, Myers ML, Short AI, Wells 1994;24:972-981. GA. Biventricular function in the adult respiratory 69. Aubier M, Lecocguic Y, Murciano D, Pariente R. distress syndrome. Chest 1983;84:126-134. Function of the respiratory muscles in acute cardiac 85. Dhainaut JF, Devaux JY, Monsallier JF, Brunet F, decompensation. Schweiz Med Wochenschr Villemant D, Huyghebaert MF. Mechanisms of 1985;115:190-193. decreased left ventricular preload during continuous 70. Scharf SM, Bianco JA, Tow DE, Brown R. The effects positive pressure ventilation in ARDS. Chest of large negative intrathoracic pressure on left 1986;90:74-80. ventricular function in patients with coronary artery 86. Artigas A, Bernard GR, Carlet J. et al. The disease. Circulation 1981;63:871-875. American-European Conference on ARDS, Part 2. 71. Rasanen J. Nikki P. Heikkila J. Acute myocardial Intensive Care Med 1998; 24:378-398. infarction complicated by respiratory failure. Chest 87. Standiford TJ, Morganroth ML. High-frequency 1984;85:21-28 ventilation. Chest 1989;96:1380-1389 72. Mathru M, Rao TL, El-Etr AA, Pifarre R. Hemodynamic 88. Abraham E, Yoshihara G. Cardio-respiratory effects of response to changes in ventilatory patterns in patients pressure controlled inverse ratio ventilation in severe with normal and poor ventricular reserve. Crit Care Med respiratory failure. Chest 1989;96:1356-1359. 1982;10:423-426. 89. Lessard MR, Guerot E, Lorino H, Lemaire F, Brochard 73. Mathru M. Editorial. Mechanical breath: L. Effects of pressure controlled with different ratios non-pharmacological support for the failing heart. Chest versus volume controleed ventilation on respiratory 1984;85:1. mechanics, gas exchange, and in patients 74. Pinsky MR, Summer WR. Cardiac augmentation by with adult respiratory distress syndrome. Anesthesiology phasic intrathoracic pressure support in man. Chest 1994;80:983-991. 1983;84:370-375. 90. Gattinoni L, Agostini A, Damia G. et al. Hemodynamics 75. Aubier M, Viires N, Syllie G, Mozes R, Roussos C. and renal function during low frequency positive Respiratory muscle contribution to lactic acidosis in low pressure ventilation with extracorporeal CO2 removal. cardiac output. Am Rev Respir Dis 1982;126:648-652. Intensive Care Med 1980;6:155-161. 76. Squara P. Dhainaut J-F, Brunet F. Acute right 91. Hirschl RB, Pranikoff T. Wise C, et al. Initial experience ventricular failure. Ch 18. Pathophysiologic with partial liquid ventilation in adult patients with acute Foundations of Critical Care. 1993. Editors: Pinsky MR, respiratory distress syndrome. JAMA 1996;275:383- Dhainaut J-F. Publisher: Williams & Wilkins. 390. 77. Pinsky MR. Determinants of Right Ventricular 92. Rossaint R, Falke KJ, Lopez F, Slama K, Pison U, Zapol Performance. Ch 17. Pathophysiologic Foundations of WM. Inhaled nitric oxide for the adult respiratory Critical Care. 1993. Editors: Pinsky MR, Dhainaut J-F. distress syndrome. N Engl J Med 1993;328:399-405. Publisher: Williams & Wilkins. 93. Bone RC, Slotman G. Maunder R. et al. Randomised 78. Brooks H, Kirk ES, Vokonas PS, Urschel CW, double-blind multicentre study of prostaglandin E1 in Sonnenblick EH. Performance of the right ventricle patients with adult respiratory distress syndrome. Chest under stress: relation to right coronary flow. J Clin 1989;96:114-119. Invest 1971;50:2176-2183. 94. Borelli M, Benini A, Denkewitz T, Acciaro C, Foti G, 79. Vlahakes GJ, Turley K, Hoffman JIE. The Pesenti A. Effects of continuous negative extrathoracic pathophysiology of failure in acute right ventricular pressure versus positive end-expiratory pressure in acute hypertension: hemodynamic and biochemical lung injury patients. Crit Care Med 1998;26:1025-1031. correlations. Circulation 1981;63:87-95 95. Peters J, Hecker B, Neuser D, Schaden W. Regional 80. Ghiogne M, Girling L, Prewitt RM. Volume expansion blood volume distribution during positive and negative versus norepinephrine in treatment of a low cardiac airway pressure breathing in supine humans. J Appl output complicating an acute increase in right Physiol 1993;75:1740-1747. ventricular afterload in dogs. Anesthesiology 96. Bersten AD, Holt AW. Acute cardiogenic pulmonary 1984;60:132-135. edema. Curr Opinion Crit Care 1995;1:410-419. 81. Ducas J, Duval D, Dasilva H, Boiteau P, Prewitt RM. 97. Fellahi JL, Valtier B, Beauchet A, Bourdarias JP, Jardin Treatment of canine pulmonary hypertension: effects of F. Does positive end-expiratory pressure ventilation norepinephrine and isoproterenol on pulmonary vascular improve left ventricular function? Chest 1998;114:556- pressure flow characteristics. Circulation 1987;75:235- 562. 242. 98. Bersten AD, Holt AW, Vedig AE, Skowronski GA, Baggoley CJ. Treatment of severe cardiogenic

398 Critical Care and Resuscitation 1999; 1: 388-399 G. J. DUKE

pulmonary edema with continuous positive airway from mechanical ventilation. Anesthesiology pressure delivered by face mask. New Engl J Med. 1988;69:171-179. 1991;325:1825-1830. 101. Naughton MT, Liu PP, Bernard DC, Goldstein RS, 99. Richard C, Teboul JL, Archambaud F, Hebert JL, Bradley TD. Treatment of congestive heart failure and Michaut P, Auzepy P. Left ventricular function during Cheynes-Stokes respiration during sleep by continuous weaning of patients with chronic obstructive pulmonary positive airway pressure. Am J Respir Crit Care Med disease. Intensive Care Med 1994;20:181-186. 1995;151:92-97. 100. Lemaire F, Teboul J-L, Cinotti L, et al. Acute left ventricular dysfunction during unsuccessful weaning

399