Intrinsic positive end-expiratory pressure (PEEPi)

Intensive Care Med (1995) 21:522-536 9 Springer-Verlag 1995 A. G. G. G. Rossi Polese Brandi Conti Intrinsic positive end-expiratory pressure (PEEPi) Received: 16 April 1993 Accepted: 25 April 1994 Supported by the National Research Council and the Ministry of Education (Special Project on Respiratory Pathophysiology), Rome, Italy, and by Grant no. 407 from Theleton, Rome, Italy A. Rossi (E~) Fisiopatologia Respiratoria, Divisione di Pneumologia, Ospedale Maggiore di Borgo Trento, Piazzale Stefani 1, 1-37126 Verona, Italy G. Polese Servizio Pneumofisiologico, Verona, Italy G. Brandi Institute of Human Physiology, University of Padova, Padova, Italy G. Conti Institute of Anaesthesia and Intensive Care, University "La Sapienza", Rome, Italy Introduction Positive end-expiratory pressure (PEEP) is one of the most frequently discussed topics in critical care medicine. However, alveolar pressure can remain positive throughout expiration without P E E P set by the ventilator whenever the time available to breathe out is shorter than the time required to decompress the lungs to the elastic equilibrium volume of the total respiratory system (Vr). The end-expiratory elastic recoil (Pel,rs) due to incomplete expiration has been termed auto PEEP, occult P E E P [1], inadvertent P E E P [2], endogenous PEEP, in- ternal P E E P and intrinsic P E E P [3, 4] owing to its similarity and contrast with PEEP set by the ventilator. The purpose of this article is to provide a comprehensive review of the studies on this interesting aspect of critical care medicine, from the underlying physiological mechanism(s) to the clinical and therapeutic implications, through measurement and monitoring in the intensive care setting. Physiology The physiological mechanisms leading to intrinsic P E E P are closely related to factors determining the end-expiratory lung volume and the rate of lung emptying. The functional residual capacity (FRC) is defined as the amount of gas in the lungs and the airways at the end of expiration [5]. In normal subjects, during tidal breathing, the elastic energy stored in the respiratory system during the preceding inspiration is sufficient for an expiration and FRC is determined by the opposing elastic forces of the lungs and the chest wall, Under such circumstances, expiratory flow becomes nil, and remains nil for an appreciable time (i. e. the end-expiratory pause) before the onset of the next inspiration: FRC corresponds to the elastic equilibrium volume of the total respiratory system ( F r). Pulmonary hyperinflation defines an increase in FRC above predicted values. Pulmonary hyperinflation can be due, for example, to loss of lung elastic recoil and airway closure at higher volumes. Dynamic pulmonary hyperinflation (DPH) defines an increase in FRC above F r due to the presence of dynamic forces at the end of expiration: (a) increased flow resistance (and expiratory flow limitation); (b) short expiratory duration; and (c) increased post-inspiratory inspiratory muscle activity [6, 7]. Abnormally increased flow resistance, often associated with expiratory flow limitation, is by far the major mechanism leading to D P H in mechanically ventilated patients and in patients with acute respiratory failure (ARF) in 523 general. In addition, short expiratory time (TE) can be an important mechanism of D P H in conditions such as inverse ratio ventilation and spontaneous rapid shallow breathing. Intrinsic P E E P (PEEPi) is a systematic corollary of D P H , reflecting the end-expiratory Pel,rs (Fig. 1) [2, 71. 1.5 1.0 0.5 ~ ~ ~_ o -o.5 -1.0 -1.5 Internal and external factors causing D P H and P E E P i Basically, three factors cause P E E P i and determine its magnitude in mechanically ventilated patients and during weaning: (a) the patient's respiratory mechanics (respiratory system resistance and compliance); (b) added resistance (endotracheal tube and ventilator tubing and devices); and (c) ventilatory pattern (due to the ventilator setting, the patient ventilatory demand or a combination of both). For the purpose of this article, patient's respiratory mechanics and breathing pattern are classified as internal factors, whereas added resistance and ventilator settings are referred to as external factors (Table 1). In most instances, internal and external factors are related and synergetic in causing PEEPi and determining its magnitude, Another factor that promotes PEEPi is the end-inspiratory pause (in general, < 0.4 s) c o m m o n l y used in the ICU to improve gas exchange [8]. It should be noted, however, that the end-inspiratory pause causes not only an increase in inspiratory time, and hence a decrease in TE at a given frequency, but also a decrease in the pressure available to produce expiratory flow. During the pause there is a decrease in Pel,rs, the driving pressure for expiration [9]. F l o w resistance The mechanism by which abnormally increased flow resistance causes D P H is self-evident. Excessive flow resistance actually delays the rate of lung emptying such that expiration cannot be completed within the time actually available for breathing out. Under such circumstances, inspiration (or mechanical lung inflation) starts before full decompression of the lungs, and FRC stabilises above V r. The usual end-expiratory pause is then replaced by a change in flow direction, from expiration to inspiration (Fig. 2). In mechanically ventilated patients and during weaning, total flow resistance is given by: (a) airway and respiratory tissue resistance, (b) added resistance of fine-bore endotracheal tube (ETT), and (c) ventilator tubing, circuits and devices. It has been shown that E T T resistance in vivo is further increased by compression and kinking of the tube as well as by secretions in the lumen [10, 11, 121. 0 2 4 6 0 2 4 6 0 2 4 6 10 12 14 16 8 10 12 14 16 8 10 12 14 16 1.0 o.8 ~0~ o.6 E 0.4 0.2 0 6o 5o 40' ~ z E & ~ a_ 30 20 10 0 Time (s) Fig. 1 Representative record with measurement of PEEP i by endexpiratory airway occlusion EEO in a mechanically ventilated patient with acute exacerbation of COPD during controlled ventilation with constant inspiratory flow. Top to bottom: records of flow, volume, and pressure at the airway opening Pao' Inspiration is upward. The continuous line in the upper panel indicates zero flow. The first mechanical inflation is regular, namely without EEO: endexpiratory pressure is apparently atmospheric. In contrast, at the end of the second tidal expiration, the expiratory circuit of the 900 C Siemens ventilator is occluded using the end-expiratory hold button of the ventilator and Pao becomes positive, reflecting the end-expiratory elastic recoil of the respiratory system due to incomplete expiration. The value of PEEP i is provided by the difference between the EEO Pao plateau and atmospheric pressure. Visual detection of the plateau on Pao provides direct evidence of absence of leaks in the circuit, respiratory muscle relaxation, and equilibration between alveolar and tracheal pressure Table 1 Determinants of DPH and PEEP (T1 inspiratory time, Tro T total time) Internal External Respiratory mechanics Added flow resistance Flow resistance Expiratory flow limitation Respiratory system compliance Fine bore endotracheal tube Ventilator tubing and devices Breathing pattern Frequency I:E Inflation volume End-inspiratory pause Frequency of breathing TI/TroT Tidal volume Ventilator setting 524 F,ow (I/s) V insp, I Volume muscles. Finally, as will be further discussed, application of moderate levels of PEEP will not increase lung volume until a "critical" value, somewhat lower than PEEP i has been exceeded [17, 22]. Respiratory compliance (I) Pao (cm H20 ) Time (s) ' is I Fig. 2 Top to bottom, records of flow, volume, and Pao in a mechanically ventilated patient during inverse ratio ventilation. There is no end-expiratory pause, but the end-expiratory flow is suddenly cut by the onset of the mechanical lung inflation, indicating that expiration was not completed and that the end-expiratory tidal volume was above the elastic equilibrium volume (dynamic pulmonary hyperinflation, DPH). Inspiratory flow is preceded by the onset of the positive pressure swing delivered by the ventilator. At zero flow, i.e. the beginning of inspiration, Pao is already positive, i.e. 3 cmH20. This pressure was needed to counterbalance part of the end-expiratory elastic recoil in order to start inspiration and corresponds to dynamic intrinsic PEEP (PEEPi,dyn) Increased airway resistance and expiratory flow limitation are not the only 'intrinsic' mechanisms causing 'occult' PEEP. Respiratory compliance must also be taken into account. In patients well adapted to the ventilator, the respiratory muscles are relaxed throughout most of the ventilatory cycle and the elastic recoil stored in the respiratory system (Pel,rs) during the preceding inspiration provides the pressure driving expiratory flow. Total flow resistance is the opposing force. Low compliance increases the expiratory driving pressure and hence the rate of lung emptying. This explains why mechanically ventilated patients with pulmonary oedema or adult respiratory distress syndrome (ARDS) have low levels of DPH and PEEPi in general, although total flow resistance has been found to be abnormally high, particularly in the early stages [11, 23, 24]. It has not yet been documented or sufficiently clarified whether expiratory flow limitation can also exist in critically ill patients without COPD. Expiratory flow limitation In patients with advanced chronic obstructive pulmonary disease (COPD), increased flow resistance due to the inflammatory process in the bronchial wall, mucus in the lumen and bronchospasm is not the only cause of retarded expiration. The destruction of lung parenchyma causes loss of alveolar septa (attachments) which support the small airways [13], while excessive pulmonary hyperinflation can determine positive pleural pressure throughout expiration [2, 10, 14]. Under such circumstances, poorly supported small airways are dynamically compressed during expiration, giving rise to expiratory flow limitation [14-191. At its extreme, dynamic compression of the airways becomes airway closure. Air trapping, a well recognised event in advanced COPD, is the result of both the closure of small airways and the abnormally low expiratory flows through compressed airways. DPH is mainly determined by expiratory flow limitation in COPD patients, which has important implications for treatment. First, after the very beginning of expiration, expiratory flow is extremely slow and the time required for a complete expiration can be abnormally long. For example, in some patients with acute exacerbation of COPD, 2 0 - 30 s of relaxed expiration was not sufficient to reach an end-expiratory pause [14, 20, 21]. Second, expiratory flow cannot be increased by application of positive pressure on the chest wall or of negative pressure at the mouth, or by contraction of abdominal (expiratory) Ventilator setting While the terms 'occult' and 'inadvertent' PEEP stress clinical implications due to the lack of recognition of PEEP i, the therms 'intrinsic' and 'auto' PEEP stress the role of patient respiratory mechanics in determining DPH and PEEP i. However, the role played by 'extrinsic' factors should also be emphasised. The magnitude of PEEP i also depends on: (a) the amount of volume to be exhaled, and (b) expiratory time, determined by (i) the ventilator frequency setting and (ii) the I : E ratio (Table 1). Ventilatory variables become internal factors when the patient is controlling the breathing pattern, as may be the case not only during weaning, but also during assisted ventilation. PEEPi and ventilatory modes To some extent, the presence and magnitude of DPH and PEEP i are influenced by the ventilatory mode. In particular, the difference between controlled and assisted modes can be important. 525 P E E P i a n d controlled ventilation Most measurements of PEEPi reported in the literature were obtained during controlled ventilation. In Table 2, we report the range and prevalence of PEEPi in a nonselected consecutive series of patients during controlled ventilation with the same ventilator setting (e. g. VT = 1 2 - 1 0 ml/kg, f = 1 2 - 1 5 breaths/min, I : E = 1 : 2 - 3 ) from some of our previous publications [10, 20, 23, 25]. As predicted, the highest values of P E E P i were found in patients with airway diseases, particularly COPD patients, in whom D P H and P E E P i were essentially caused by expiratory flow limitation. P E E P i can be significantly reduced by changing the ventilatory setting, but will never totally disappear. P E E P i has also been systematically found, albeit at low levels, in stable COPD [26-30] and cystic fibrosis (CF) [31] patients. In patients without chronic airway disease, external factors such as endotracheal and ventilator tubes, exhalation valves and the ventilator setting can play an important role in determining both the presence and the magnitude of P E E P i [23, 25]. In this connection, the high prevalence of PEEPi reported in patients with polytrauma, namely 28/50 patients (56%), in whom intrinsic factors such as high ventilatory demand were also probably involved, is noteworthy [32]. P E E P i a n d inverse ratio ventilation A clear example of D P H and PEEPi entirely caused by the ventilator setting may be seen during inverse ratio ventilation (IRV). This is a particular form of controlled ventilation, which is generally pressure controlled (PC-IRV), with a longer inspiratory phase than usual followed by a short expiration [33-37]. IRV is sometimes used in patients with severe hypoxaemia and respiratory failure, often after conventional CV has failed to maintain a satisfactory PaO2 [33]. P E E P i is present during IRV as an "occult P E E P " because of incomplete expiration (Fig. 2). The presence of P E E P i can explain the improvement in PaO2 observed in patients on IRV and the sometimes unexpected drop in cardiac output (CO) [34]. Not surprisingly, less P E E P than during CV is required to increase PaO2 during IRV since a certain amount of P E E P i is already present [37]. In addition, the conclusion that IRV decreases the risk of barotrauma because of the lower peak cycling pressure is questionable. It is true that alveolar pressure and mean airway pressure (MAP) may be much higher than expected due to DPH, and hence the risk of barotrauma is unchanged if not increased. The lower peak cycling pressure observed during IRV than during CV is the result of the lower resistive pressure drop due to the generally slower end-inspiratory flow, while the end-inspiratory alveolar pressure will approximate the level occurring with a similar amount of P E E P [36]. Table 2 Magnitude and prevalence of PEEP in mechanically ventilated patients (PEEP i intrinsic positive end-expiratory pressure (cmH20), COPD chronic obstructive pulmonary disease, CF cystic fibrosis, ARDS adult respiratory distress syndrome, CPO cardiogenic pulmonary oedema, Other patients with ARF of extrapulmonary origin Diagnosis No. PEEP i (range) Prevalence COPD CF Asthma ARDS CPE Other 45 1 3 28 10 10 2.6 - 22 11 13.5-20 1.0-8.1 1.0-6.0 1.0-4.1 45/45 1/1 3/3 t5/28 8/10 5/10 (100%) (100%) (100%) (58%) (80%) (50%) Another particular form of controlled ventilation is high-frequency ventilation (HFV). PEEPi reaching approximately 9 cmH20 was found in COPD patients ventilated with high-frequency jet ventilation [38]. P E E P i a n d assisted ventilation During assisted ventilation, i.e. assist/control ventilation (ACV), synchronised intermittent mandatory ventilation (SIMV), and pressure support ventilation (PSV), the inspiratory muscles are active at the beginning of inspiration, and even throughout inspiration, so as trigger the mechanical breath [39, 40]. Expiration is mostly passive and hence is controlled by the mechanical properties of the respiratory system, as is the case with CV, together with the added resistance of ETTs and ventilator circuits, and by the ventilator setting. Therefore, in general, the mechanisms leading to D P H and P E E P i during ACV, SIMV, and PSV do not differ from those during CV. However, with assisted ventilatory modes, random and erratic inspiratory efforts can either trigger undesired ventilatory cycles or interrupt expiration in the early phase in such a way that there is a sudden reduction in the expiratory phase leading to a dramatic rise in the magnitude of P E E P i. This has been described by Braschi et al. during SIMV [41], and Fernandez et al. during ACV [42]. P E E P i a n d s p o n t a n e o u s breathing It has been recognised that rapid shallow breathing (RSB) is a common characteristic of patients with ARF from respiratory causes. RSB can occur during the early stage of ARF and is also a common sign of unsuccessful weaning [43]. Although D P H and P E E P i have not been measured on wide scale in patients breathing spontaneously, there can be little doubt that the short TE determined by RSB may be associated with high levels of PEEPi particularly in COPD patients, as illustrated in Fig. 3. In addition, with the shortening of the expiratory phase, inertial fac- 526 tions of PEEP i in mechanically ventilated patients partly depend on the mode of ventilation. Flow (I/s) Controlled ventilation (CV) Volume (~) APpl (cmH20) I PEEP i [1] has three major consequences during controlled ventilation: (a) a drop in cardiac output due to the impediment in cardiac filling determined by the positive intrathoracic pressure; (b) a potentially greater risk of barotrauma due to the high alveolar pressure; and (c) an error in the computation of respiratory compliance. Time = 13.84 s Fig. 3 Representative record (flow, volume, and pleural pressure, APN, from top to bottom) in an intubated, spontaneously breathing patient with acute exacerbation of COPD. The expiratory flow is abruptly cut at the end of expiration, whilst Ppl swing (namely the inspiratory effort) had already begun. The difference between the point corresponding to the onset of the change in Ppl and the point of zero flow on the Pvl tracing, represents PF,EPi,dyn) which had to be counterbalanced by the contracting inspiratory muscles in order to start inspiration. In this patient, the inspiratory muscles had to develop 13 cmH20 before the inspiratory flow could start. Tidal volume amounted to 0.31 frequency was 35 breaths/rain, minute ventilation was 10.5 l/rain tors due to the moving mass of the respiratory system could also be a cause of hyperinflation. It is not possible to provide a quantitative estimation, because inertial factors are generally disregarded. Nevertheless, these may become relevant as regards very rapid breathing. In spontaneously breathing patients, respiratory muscle activity could affect DPH and PEEP i in terms of both increased post-inspiratory muscle activity (PIIA) and expiratory muscle contraction. The former has been found in normal subjects and asthmatics during induced bronchoconstriction [44], whereas the latter has been observed during stable COPD [29, 45] and during application of continuous positive airway pressure [46]. In critically ill patients, there has only been one recent report of abdominal (expiratory) muscle activity [47]. Implications Just over 10 years ago, Pepe and Marini [1] made the brilliant observation that in a few mechanically ventilated patients who were well adapted to the machine cycle and for whom no PEEP was set by the ventilator, airway occlusion at the end of tidal expiration revealed positive endexpiratory alveolar pressure. This was termed auto PEEP and occult PEEP [1], i.e. PEEP i (Fig. 1). Subsequent research has shown that PEEPi is quite common in mechanically ventilated patients [25] (Table 2). The implica- PEEP i and haemodynamics Although it is widely accepted that PEEP i can unduly increase alveolar and intrathoracic pressure and hence reduce cardiac output, few measurements of haemodynamics in patients with significant levels of PEEP~ [3, 36, 48], have been reported in the literature since the original observation by Pepe and Marini [1]. PEEPi reduces cardiac output [36], whereas a reduction in PEEP i increases cardiac output [48]. Unrecognised PEEP i, i.e. occult PEEP, may lead to misinterpretation of haemodynamic data, generating erroneous interpretation of the patient's volemic status [1]. High levels of DPH and PEEP i can be associated with reduced cardiac filling and also significant cardiac arrhythmia, which are reversed by a reduction in DPH and PEEP i. P E E P i a n d barotrauma Pulmonary barotrauma, defined by the presence of extraalveolar air in locations where it is not normally found, is a well-recognised and teared event in mechanically ventilated patients [50, 51, 52]. In patients with high levels of DPH and PEEP i, conventional tidal volume (i.e. 10-15 ml/kg) may shift ventilation toward the upper flat portion of the volume-pressure curve of the lungs, thus exposing terminal airspaces to the risk of overdistension and rupture. However, to our knowledge, no studies have been conducted to date showing a direct relationship between PEEP i and increased incidence of pulmonary barotrauma. It has been suggested that PEEP and PEEP i may help in preventing ventilatory induced lung injury [52, 54]. PEEPi prevents end-expiratory alveolar collapse and allows a smaller than usual tidal volume to be used to ventilate the lungs, thus reducing the shear forces between lung units with different regional degrees of inflation [52, 54]. However, this interesting suggestion, which puts PEEP i in a new perspective, requires further clinical investigation. 527 PEEPi and measurement of respiratory compliance As shown by Jonson et al. [55] and our work [3], if PEEP i is not recognised and measured, a significant error is introduced in the calculation of static respiratory compliance (Cst, rs), a variable which is commonly used not only to assess the status and progress of ARF in mechanically ventilated patients, but also to set PEEP at the ventilator [56]. Traditionally, Cst, r s is computed from the following equation: Cst, rs = V T / ( P p l a t - P E E P ) (i) but the correct equation is as follows Cst,r s = VT/(Pplat - P E E P - P E E P i ) (2) where VT is the tidal volume from the end-expiratory position and Pp:at is the plateau pressure measured during end-inspiratory airway occlusion [1]. It has been shown that true compliance may be underestimated by up to 100~ and 30~ in COPD [25] and ARDS [23] patients, respectively, if PEEP i is not taken into account (i. e. Eq. 1 vs Eq. 2). In this context, it should be noted that due to the interdependence of several non-constant factors (Table 1), the magnitude of PEEPi, and hence the extent of the error in the computation of compliance, is difficult to predict. It has to be stressed that most modern microprocessor-equipped mechanical ventilators that can display the on-line computation of respiratory compliance do not take PEEP i into account and thus provide incorrect and therefore useless, if not misleading, information. Assisted ventilation and weaning In patients with ARF during assisted ventilation and weaning, the implications of DPH and PEEP i are important in terms of the energetics of breathing [57]. First, dynamic hyperinflation profoundly alters the capacity of the inspiratory muscles to sustain a load due to lengthtension considerations as well as changes in the geometrical arrangement [58]. Second, PEEP: is an inspiratory threshold load [59] which must be counterbalanced by the contracting inspiratory muscles in order to create sub-atmospheric pressure in the central airways and hence either trigger the mechanical breath or generate inspiratory flow. In these circumstances, the conventional equation of motion becomes: Pappl(t) = PEEP i + [ VT(t)/Cdyn] + [RToT" V:(t)] the contribution required from the patient's respiratory muscles is negligible. However, Marini et al. [39, 40] have shown that the patient's inspiratory muscles can perform a significant amount of work, which can equal, in some conditions, the work performed during unassisted ventilation. This situation may be even worse in the presence of PEEP i. Indeed, PEEP i adds to the triggering pressure and the total amount of the effort required by the patient's respiratory muscles to produce the pressure boost may be seriously increased. With high levels of PEEP i (Fig. 3), the patient's inspiratory muscles continue to contract under load, even during ventilatory assistence, and cannot recover from fatigue, so that weaning may be delayed or become impossible [60]. Because of its important clinical implications [61], measurement of PEEP i should be a mandatory part of the routine assessment of respiratory function in critically ill patients. Measurement First of all, PEEP i must be suspected from the shape of the expired flow versus time (Fig. 2) or volume (Fig. 4) relationship, whenever low flow, the final portion of the expiration, is suddenly cut by the onset of inspiration. We believe that continuous and adequate display of flow and pressure versus time records should be an essential monitoring facility in modern microprocessor-equipped ventilators. In general, measurement of PEEP i and DPH is much easier during controlled ventilation than during assisted ventilation and weaning, since respiratory muscles are more active in the latter conditions. When the respiratory muscles are relaxed and sufficient time for equilibration between regional units with different time constants has been allowed, the pressure at the airway opening during airway occlusion reflects mean alveolar pressure for the definition of still fluid (Pascal's law). Simple and commonly available equipment such as a pneumotachograph for measuring flow, and a differential pressure transducer for sampling airway pressure can be used to measure PEEPi, DPH and respiratory system mechanics [62, 63]. Since measurement of PEEPi may present more problems during assisted ventilation and weaning than during CV, the two modes of ventilation will be discussed separately. (3) where Pappl is the total pressure applied to inflate the respiratory system, Vx is tidal volume, Cdy n is dynamic respiratory compliance, RTOT is total flow resistance, and VI is the inspiratory flow rate, at any instant (t). During assisted ventilation, the triggering pressure is set at very low levels, e.g. 1 - 2 cm H20, so that the ventilator does most of the work of inflating the lungs and Controlled ventilation During CV, PEEP i can be easily measured either by endexpiratory airway occlusion (EEO) [2, 10, 23] (Fig. 1) or by simultaneous recording of flow (~'i) and pressure at the airway opening (Pao) (Fig. 2) [31. Although in one study [3] the two methods yielded similar results, a more 528 with fast time constants start filling such that the change in Pao preceding flow reflects the amount of pressure required to counterbalance P E E P i in the fast time constant units, i.e. the lowest P E E P i , to start inspiratory flow. P E E P i obtained from continuous recording of Pao and Vi is referred to herein after dynamic P E E P i (PEEPi, dyn). 1.0 0,8 18 ............ 0,6 16 0.4 14 0.2 PEEP i 0 12 --~" -0.2 -0.4 "5 o -0.6 8 o3 6 {3- <::1 -0.8 -1.0 -1.2 -1.4 ./" 4 ~i'.' 2 -1,6 -1.8 I 0.1 i I t I 02 0.3 Flow (I/s) i I 0.4 ' 0 0.5 Fig. 4 Relaxed expiratory volume-flow curves in a mechanically ventilated patient with acute exacerbation of COPD during controlled ventilation. The continuous line represents tidal expiration, whereas the dashed line represents the attempt of complete expiration to the relaxation volume, which, however, was not achieved. At the end of the tidal expiration, flow is suddenly cut by the next mechanical lung inflation, indicating DPH. The volume axis (extrapolated to zero) was transformed into a pressure axis (right y axis) by dividing volume by respiratory compliance according to the formula: Pez rs = V~ where Pel rs is the elastic recoil pressure of the respiratory system, and Crs is 'the respiratory compliance calculated by the interrupter technique (Fig. 5). Dynamic hyperinflation, i.e. the difference between the relaxed volume and the tidal expired volume, amounted to > 1.51 with a corresponding PEEP i amounting to > 12 cmH20 recent report revealed that P E E P i measured by EEO was markedly greater than P E E P i measured by simultaneous ~ri of flow and Pao recording [64]. P E E P i measured by EEO actually reflects the end-expiratory elastic recoil of the respiratory system under static conditions, whereas P E E P i measured from the change in Pao preceding Vi reflects the minimum "dynamic" P E E P i. Since the lung units time constants are unequal, a well-recognised phenomenon in patients with airway and lung parenchyma diseases [10, 20, 25, 65, 66], P E E P i cannot be distributed homogeneously in the lungs, but will be greater in units with long time constants and a slow rate of emptying than in units with short time constants and rapid expiration. During EEO, there is time for equilibration (Pendelluft) of lung units with different regional P E E P i such that P E E P i measured as the plateau pressure at airway opening during occlusion (Fig. 1) reflects the mean value after equilibration. In contrast, P E E P i measured from the change in Pao preceding inspiratory flow (Fig. 2) is the minimum P E E P i. With incomplete expiration, the units with long time constants are still emptying while those End-expiratory airway occlusion End-expiratory airway occlusion can be performed manually at the expiratory port of the ventilator during the last 0.5 s of an expiration [1, 67]. Alternatively, a pneumatic valve can be used for rapid occlusion at the airway opening [3, 16, 25]. The second technique has the advantage of excluding the compliance of gas compression in the ventilator tubing (around 0.7 ml/cmH20 [20]), though some variability may be due to occlusion timing. However, near the end of expiration the volume expelled per unit of time is very small due to extreme airway compression, so that small changes in TE would not be expected to affect the magnitude of PEEPi substantially, at least for TE around 3 s [3, 67]. With shorter TE, the difference is not negligible and P E E P i can increase significantly with decreasing TE [3]. P E E P i can be measured with ventilators which are already equipped with the end-expiratory occlusion hold, i.e., a button for the rapid occlusion of the expiratory port exactly at the end of the tidal expiration (e.g. Servo 900C, Siemens). With other ventilators, which do not include this option, airway occlusion can be achieved by means of manually operated valves [68, 69]. In our experience (Table2), a plateau in airway pressure was constantly observed within 1.0 s from the onset of occlusion. However, D'Angelo et al. [9] observed that a longer time was required for equilibration (up to 5 s) even in normal anaesthetised subjects during airway occlusion at different lung volumes so as to observe a plateau in airway pressure. Since airway occlusion lasting a few seconds does not create any discomfort to the patient, occlusion longer than 1 s (up to 5 s) may also be used in clinical practice. With EEO, P E E P i can also be directly read on the ventilator's analog pressure display when recording facilities are not available. However, a significant advantage of recording signals on paper or screen is the possibility of observing the plateau in airway pressure, thus providing direct evidence of respiratory muscle relaxation and the absence of leaks in the circuit as well as equilibration between alveolar and airway opening pressure. With the airway pressure analogic signal recorded either on paper or on screen P E E P i is measured as the difference between the value of plateau pressure during airway occlusion and atmosphere, as illustrated in Fig. 1. If P E E P was set by the ventilator, P E E P i is measured as the difference between end-expiratory plateau pressure, which represents total P E E P (PEEPt) and the pre-interruption level, which reflects the level of P E E P set by the 529 ventilator. Although this technique can lead to slight underestimation of P E E P i and overestimation of P E E P because of a small amount of resistive pressure dur to end-expiratory flow, the error is likely to be negligible. This may be particularly useful in calculating changes in P E E P i during a stepwise application of external P E E R P E E P i measured by EEO reflects the end-expiratory elastic recoil of the total respiratory system (PEEP~,r~). We recently divided PEEPi,rs into its lung (PEEPi,1) and chest wall component (PEEPi,w) using EEO together with the oesophageal balloon technique [10, 19, 70]. In our COPD and ARDS patients, PEEPi,rs substantially reflected PEEPi,I, with a small chest wall contribution (about 16~ on average). PEEPi can be divided not only between the lung and the chest wall, but also between the two lungs. In some instances, P E E P i is unevenly distributed between the two lungs. For example, P E E P i measured during differential lung ventilation amounted to 12 cmH20 in the more severely diseased lung, whereas 4 cmH20 of PEEP~ was found in the other lung [71]. In certain patients, such as those with coexisting chronic airway obstruction and fibrotic lung process, unilateral pulmonary hyperinflation and unilateral P E E P i up to 15 cmH20 can be observed [72]. and P E E P i measured using the supersyringe in 16 patients (r = 0.96, slope close to 1). A new computer-controlled occlusion method (SCASS = static compliance by automated single steps) has been recently proposed for the determination of P E E P i in mechanically ventilated patients [751. Continuous recording of flow and pressure at airway opening As illustrated in Fig. 2, this is an indirect method for measuring P E E P i. On the basis of the equation of motion (Eq. 3), it assumes that the change in Pao preceding inspiratory flow reflects the amount of pressure required to counterbalance P E E P i. This technique provides the value of PEEPi,dyn and is of particular value for continuous monitoring of P E E P i, for example during changes in the ventilator setting or to follow the effects of bronchodilator therapy. Provided that the difference between E E O - P E E P i and PEEPi,dyn is kept in mind, this technique has the advantage of avoiding any manoeuvre in ventilator setting, allowing monitoring of P E E P i 1.0 Interrupter technique The interrupter technique consists of multiple brief airway occlusions performed during relaxed expiration and has been described in detail elsewhere [16, 20]. As illustrated in Fig. 5, the interrupter technique enables both dynamic hyperinflation, as the difference between Vr (intercept on the volume axis) and the AV = 0, which represents end-expiratory lung volume during tidal ventilation, and P E E P i, as the value of airway pressure at AV = 0, to be measured. Total respiratory system compliance can be obtained from the slope of the regression line calculated from the points through the central (linear) portion of the VP relationship. Static pressure-volume curve P E E P i was recently measured on the static pressure-volume curve obtained by means of the supersyringe technique [73] which also enables respiratory compliance to be determined and the "inflection point" to be detected in patients with ARDS [74]. For measurement of P E E P i, complete expiration to Vr is not possible. Instead, the airway is occluded at the end of the tidal expiration and P E E P i is measured as the amount of pressure which has to be applied with the supersyringe before the respiratory system starts inflating [73]. Fernandez et al. [73] found an excellent correlation between P E E P i measured by EEO 0,5 ,Ira'" 0 I/If" .im.mI"""l =~'-0,5 ~><::1-1.0 I IIII- 'j -1.5 . -2.0 ..... .m 0 PEEPi ~ 5 10 15 Pao(cmH20) 20 25 Fig. 5 Volume-pressure relationship in a mechanically ventilated patient with acute exacerbation of COPD during controlled ventilation. Pressure was measured at the airway opening (Pao)" The interrupter technique was used throughout a complete relaxed expiration. Dots are measurements from multiple interruptions. Complete expiration went well below the tidal expiration and the elastic equilibrium volume of the respiratory system (Vr) is represented by the intercept on the Y axis: the difference between Vr and AVolume = 0 provides the amount of DPH (> 1.51 in this patient). The value for P~o at AVolume = 0 reflects the amount of the PEEP (about 10 cmH20 in this patient) 530 breath-by-breath. Moreover, with some ventilators that do not have the end-expiratory occlusion button, continuous recording of V~ and Pao can be the only means of assessing rapid changes in PEEP i. A modification of this approach has been suggested by Braschi et al. [76]. To summarise, during CV, measurement of PEEP i is important, particularly in patients with obstructive airway disease and also easy to perform with simple, non-invasive, bedside techniques. o Because of its implications in terms of respiratory muscle function and energetics of breathing, measurement of DPH and PEEP i should routinely be performed in the course of the assessment of respiratory function before weaning is attempted, particularly in patients recovering from acute exacerbation of COPD. EEO should be attempted first. However, since the respiratory muscles are active in patients during assisted ventilation and weaning, techniques requiring respiratory muscle relaxation, such as EEO and the interrupter technique, may be unsuccessful, at least in common clinical practice. Nevertheless, some investigators have also obtained reasonable plateaux during assisted ventilation [67] and continuous positive airway pressure (CPAP) [18]. Since expiration is mostly relaxed during assisted ventilation, the interrupter technique could also be used, as is the case in experimental animals [77] and in both anaesthetised [78] and awake normal subjects [79]; however, this has yet to be attempted in critically ill patients with ventilator modes other than CV. Hoffmann et al. recently suggested the use of respiratory inductive plethysmography (RIP) to measure PEEP i during assisted ventilation [80]. Using the RIP technique, the magnitude of PEEP i is calculated from the amount of external PEEP which can be applied without changes in the end-expiratory thoracic volume. The technique has the advantage of being non-invasive, but still poses a few problems; for example, the patient has to be relaxed, at least during the procedure, in the same way as during satisfactory end-expiratory airway occlusion. In addition, the value of PEEP~ obtained with the RIP technique is lower than the true PEEP i since application of PEEP equal to the initial PEEP i induces a moderate increase in thoracic volume [8i]; only values of PEEP between 75% [18] and 85~ [19] of the initial PEEP i do not significantly increase the end-expiratory volume. Finally, the RIP calibration procedure may not be easy or straightforward at the bedside. Oesophageal balloon Although it represents an additional invasive procedure in the crowded atmosphere surrounding the patients in the sB 9 CPAP [] PSV 10 PEEP+PSV 9 S ~// d ,~ ^ , " " - jr1 ,, 0.8 o /o/ 0.6 x~ E O v Assisted modes of mechanical ventilation and weaning 1.0 12 6 w I_o a_ 0.4 4 / , - . " &" o .--- 0 0 2 4 6 8 10 12 PEEPi,dy n uncorrected (cmH20) Fig. 6 Relationship between values of PEEPidyn corrected (PEEPidyn) and uncorrected (PEEPidyn uncorrected') for expiratory muscle relaxation in seven patients with COPD and ARE The symbols are data from each patient during different ventilatory modes with (closed symbols) and without (open symbols) application of CPAP or PEEP. Solid line is the line of identity. Dashed lines indicate different ratios of corrected to uncorrected PEEPi,dy n intensive care unit, the oesophageal balloon technique to measure changes in Ppl remains, in most cases, the only means of measuring PEEP i and pulmonary mechanics in actively breathing patients. As illustrated in Fig. 3, PEEP i is measured from the change in APpl preceding flow. PEEP i measured in this way represents dynamic PEEPi, as is the case for PEEPi obtained from continuous recording of V~ and Pao (Fig. 2), and it is significantly lower, compared with static PEEP i [18]. This method of measuring PEEPi,dyn, which has been used in both acutely ill [18, 67, 82, 83] and stable [26-28] COPD patients, is valid provided that the expiratory muscles are relaxed at the end of expiration. If the expiratory muscles actively contract during expiration, the decrease in Pp] in early inspiration could be due to expiratory muscle relaxation rather than to inspiratory muscle contraction [29, 45]. However, in acutely ill COPD patients the gastric component of PEEP i did not exceed a few centimetres of water becoming negligible at the highest values of PEEP i (Fig. 6). Pleural pressure can be measured by a balloon-catheter system incorporated in a nasogastric tube used for nutrition thus rendering the process less invasive [84]. 531 Treatment The problem of reducing DPH and PEEPi has a major clinical impact for the treatment of patients with acute exacerbation of COPD and asthma. In such patients, DPH and PEEP i are systematically present reaching excessive values. This can lead to impaired cardiac function, enhanced risk of barotrauma, reduced inspiratory muscle pressure-generating capacity and abnormally increased energy cost of breathing. Although some levels of PEEP i have also been observed in patients without chronic airway disease [23, 25], the reduction of PEEP i in these patients is likely to be less relevant in the clinical picture and easy to accomplish with small changes in T E. Therefore this section of the article will concentrate mainly on the treatment of PEEP i in patients with acute exacerbation of COPD and asthma. Basically, there are four methods for reducing PEEPi: (1) changes in ventilator setting; (2) reduction in patient ventilatory demand; (3) bronchodilators; and (4) application of external PEEP. Clearly, the latter does not reduce DPH. All methods are suitable for simple application in the clinical setting and can be used in synergy. [85], who used "controlled hypoventilation" to improve the clinical outcome in patients with status asthmaticus, and also by Hickling et al. [86] in ARDS patients. Tuxen et al. [56] showed that the magnitude of pulmonary hyperinflation is a major problem in patients with status asthmaticus, and that the reduction of pulmonary hyperinflation can be associated with a better clinical outcome [55]. In order to improve expiratory flow, the ETT should have the largest possible internal diameter compatible with the patient's characteristics; cleaning of the tube by frequent suctioning is also extremely important in prevening an abnormal increase in flow resistance due to the accumulation of secretions in the lumen. Reduction of ventilatory demand During assisted ventilation techniques such as ACV, SIMV and PSV, and weaning, it is more difficult to control the ventilator setting than during Cu because it depends, at least in part, upon the patient's ventilatory demand and pattern. In such circumstances, it can be important to reduce the patient's ventilatory demand and minute ventilation. For example, overfeeding and/or excessive carbohydrate feeding unduly increase CO 2 production. Correct nutritional support will reduce CO2 and Changes in ventilator setting and apparatus hence the ventilatory drive. It is also important to reduce the dead space as possible by keeping the proximal tip of The ventilatory pattern should be set so as to provide the the ETT as close as possible to the Y-piece of the venlongest expiratory phase compatible with the patient's tilatory tubing. Changes in ventilatory mode may also decomfort and adequate gas exchange. This can be achieved crease PEEPi; for example, Conti et al. [49] reduced during controlled ventilation by decreasing ventilatory PEEP i from 17 crnH20 to 7 cmH20 in a mechanically frequency (e.g. from 15 to 12-10 breaths/min) and the ventilated COPD patient by switching the ventilatory inspiratory time (T~), or increasing inspiratory flow rate mode from SIMV to PSV: the respiratory frequency de(V~). At any given level of minute ventilation (VE) a rise creased from 31 to 13 breaths/m, and cardiac complicain VI will reduce T I and increase T~; reducing the tions due to excessive pulmonary hyperinflation disapTI/TTo7 or the I : E ratio. High values of peak cycling peared. pressure (Ppeak) associated with high inspiratory flow should not be noticeable. The high Ppeak can in fact be due to the resistive pressure required to push inspiratory Bronchodilators flow through the endotracheal tube and the conducting airway, and will not affect alveolar pressure. The use of It has been shown that commonly used bronchodilators high inspiratory flow to lengthen the expiratory duration, such as methylxanthines (doxofylline) and [32-adrenergic although useful, may not be sufficient. In mechanically agonists (e. g. fenoterol and albuterol) are also active in ventilated patients with severe airflow obstruction, a com- mechanically ventilated patients with acute exacerbation plete expiration can require more than 20 s [20, 21]. Such of COPD [87-91]. Both groups of drugs significantly rea long T E cannot be used in clinical ventilator setting. In duce airway resistance because of their relaxant effect on addition, a T E slightly longer than 3 s may give rise to bronchial smooth muscle. The improved rate of lung only a small reduction in PEEP i due to the extremely low emptying decreases DPH and PEEP i. Bernasconi et al. expiratory flow rate. [88] showed that [32-adrenergic agonists could determine Tidal volume can be also reduced so as to decrease the a significant, albeit brief, bronchodilatation in patients amount of air to be exhaled and hence DPH and PEEP i. already receiving a continuous infusion of aminophylline. However, reduction in VT will decrease alveolar ventila- Manthous et al. [91] recently showed that the decrease in tion and hence determine a rise in PaCO2 . Traditionally, resistive pressure was greater with nebulised albuterol a PaCO2 higher than 45 mmHg has been perceived as an than with equal doses delivered by metered dose inhalers. adverse event in mechanically ventilated patients. This The delivery system may be as important as the drug itself traditional view has been challenged by Darioli and Perret in determing the therapeutic response [92]. 532 It has been shown in stable COPD patients that the decrease in lung volume determined by bronchodilators can increase inspiratory muscle strength by improving the operational length of diaphragmatic fibres in their lengthtension relationship [26]. This effect, associated with a reduction in P E E P i and in work of breathing [93] can be very important for weaning. There is little evidence of the effect of steroids in mechanically ventilated COPD patients. It could be predicted that steroids improve airway patency by reducing the amount of secretions in the bronchial lumen as well as the inflammatory process in the bronchial wall. Nava et al. [941 showed that parenteral steroids can reduce D P H and PEEPi in mechanically ventilated COPD patients. A special condition requiring aggressive measures in order to reduce airway resistance is a life-threatening asthma attack [95]; for example, adrenaline can determine an important reduction in both P E E P i and D P H (Fig. 7). However, in some cases, the failure of conventional bronchodilators to reverse the severe airflow obstruction and D P H can be observed. In some patients, supplemental bronchodilatation can be obtained by deep sedation reaching anaesthetic levels, considering that profound sedation can have per sea bronchial smooth muscle relaxant effect [96]. In addition, some volatile and/or intravenous anaesthetics have a bronchodilating action, such as ketamine and, particularly, halothane, a volatile general anaesthetic that is a popular "last resort" powerful bronchodilator [961. Application of PEEP C o n t r a r y to the t r a d i t i o n a l n o t i o n a n d t a k i n g s o m e controversy into a c c o u n t [29, 97], it seems a c c e p t a b l e t h a t m o d e r a t e (judicious) levels o f P E E P can be o f benefit in patients with acute e x a c e r b a t i o n o f C O P D d u r i n g assisted v e n t i l a t i o n a n d weaning [22, 83]. A s m e n t i o n e d above, P E E P i is an i n s p i r a t o r y t h r e s h o l d l o a d which has to be c o u n t e r b a l a n c e d by the p a t i e n t ' s i n s p i r a t o r y muscles either b e g i n n i n g i n s p i r a t i o n or triggering the m e c h a n i c a l breath. P r o v i d e d t h a t tidal e x p i r a t i o n is flow limited, app l i c a t i o n o f P E E P will p a r t l y replace P E E P i, w i t h o u t a d d i n g to it a n d hence w i t h o u t increasing lung volume, until a critical value s o m e w h a t lower t h a n P E E P i has been exceeded [17, 83]. In line with the cascade t h e o r y o f e x p i r a t o r y flow l i m i t a t i o n [98, 99], these results were obt a i n e d in several clinical studies [17 - 19, 47, 80, 100]. Figure 8 shows t h a t a p p l i c a t i o n o f P E E P lower t h a n P E E P i in a C O P D p a t i e n t d u r i n g c o n t r o l l e d v e n t i l a t i o n d i d n o t increase p e a k cycling pressure. D u r i n g assisted ventilation a n d weaning, a p p l i c a t i o n o f P E E P can u n l o a d the p a t i e n t ' s i n s p i r a t o r y muscles. 60 50 40 0 0.8 0.6 30 v v 0.4 0 0.2 13.. d3 E 0 > -0.2 <1 -0.4 -0.6 d~ n 20 10 e d2d ~ Adrenahne -0.8 -1 I I I I 10 20 30 40 I"" 50 Pao (cm H20 ) Fig. 7 Static volume-pressure relationship in a mechanically ventilated patient with acute severe asthma. As in Fig. 5, experimental points were obtained by the interrupter technique. Stars and empty triangles are interrupter measurements before and after 0.1 mg adrenaline. The relaxant effect of adrenaline on bronchial smooth muscle decreased both the amount of DPH and PEEP i, without changes in the slope of the volume-pressure relationship, i.e. respiratory compliance. Though the patients were allowed 30 s for breathing out, expiration was still not completed because of the exterme flow resistance and maybe expiratory flow limitation. Measurement of DPH and PEEP i was performed as illustrated in Fig. 5 0 L__. I I 5 i I 10 15 20 25 30 Time (s) Fig. 8 Pressure at the airway opening (Pao) versus time during controlled ventilation in a given patient with acute exacerbation of COPD. PEEP i was measured by EEO. From the first mechanical inflation, PEEP was set by the ventilator at 5 cmH20 (second inflation) and at 10 cmH20 (third inflation). Then the total PEEP (i.e. PEEPtot = PEEP+PEEP i) was measured again by EEO. Since PEEP partly replaced PEEP i, without adding to it, because of flow limitation, the peak cycling pressure did not change despite increasing PEEP set by the ventilator. This figure illustrates a simple way to assess the effect of PEEP set by the ventilator in patients with PEEP i at least during CV 533 The initial value of P E E P i should not be exceeded [97], and individual differences should also be taken into account [101]. Moreover, the presence of expiratory flow limitation should also be assessed, for example following the procedure illustrated in Fig. 4. It has been shown by Tuxen, that application of P E E P in patients without evidence of expiratory flow limitation, can further increase pulmonary hyperinflation and its adverse consequences [97]. In this context, it should be noted that, while there is evidence of expiratory flow limitation in mechanically ventilated C O P D patients, it is still not known whether expiratory flow limitation exists in mechanically ventilated patients with status asthmaticus. In such circumstances, the rationale for application of P E E P in asthma remains uncertain and further investigation is required. High levels of PEEP, up to 25 cmH20, have been used with the purpose of opening closed airways [102]. An unusual case was recently published [103] concerning a 51-year-old w o m a n with pulmonary oedema and bronchospasm, who managed to produce "auto P E E P " by holding her head out of the window while her husband was driving to the hospital at 80 mph; it was calculated that the effect of the above corresponded to 8 c m H 2 0 PEEP. Apparently, this was needed to keep the patient out of a coma [103]. Although no controlled, randomised clinical trials have shown that weaning of C O P D patients is significantly improved by the use of P E E P (or CPAP), clinical and physiological studies have documented that application of P E E P or CPAP nearly up to the level of initial P E E P i, in C O P D patients, being of benefit during assisted ventilation [67], pressure support ventilation [100] and weaning [18]. As shown by Fernandez et al. [42], moderate levels of P E E P can be adapted to ventilated patients with P E E P i, who were thought to be fighting the ventilator. Low levels of P E E P and CPAP (about 5 c m H 2 0 ) were used for mask (non-invasive) ventilatory support in patients with acute exacerbation of C O P D [47]. CPAP was also applied to asthmatic patients with acute asthma [104] and respiratory failure [105] in the emergency room. Although it is generally agreed that the initial level of PEEPi should not be exceeded, some controversy still remains as to the best level of P E E P to be set. Ranieri et al. [19] suggested a level of external P E E P amounting to 85% of the initial PEEPi. Georgopoulos et al. [101] recently concluded that the individual patient response to P E E P is still unpredictable. Fernandez et al. [81] observed that levels of P E E P equal to initial P E E P i caused a moderate but detectable increase in end-expiratory volume, which was more marked in patients with high pulmonary compliance. During mask ventilatory support, Appendini et al. [47] used a level of CPAP and P E E P amounting to 90~ of the initial P E E P i and did not observe significant changes in end-expiratory volume. Table3 Treatment of PEEP i Changes in ventilator setting 9 Increase expiratory duration 9 Decrease ventilatory frequency 9 Decrease tidal volume Reduction in the ventilatory demand 9 Decrease carbohydrate intake 9 Reduce dead space 9 Reduce anxiety, pain, fever, shivering Reduction of total flow resistance 9 Use of large bore endotracheal tubes 9 Frequent suctioning 9 Bronchodilators Application o f external PEEP nearly up to the level o f initial PEEPi In our practice, the minimum and m a x i m u m level of P E E P are represented by PEEPi, dyn and P E E P i, respectively, measured during end-expiratory airway occlusion. I f PEEPi,ay n is not measured, a level of P E E P amounting to 85O7o of E E O - P E E P i appears safe [19, 47]. Application of moderate levels of PEEP, i.e. lower than P E E P i measured by end-expiratory airway occlusion, should be started almost simultaneously with the institution of assisted modes of ventilation to decrease the ventilatory load [106]. Whether or not CPAP is a choice technique for weaning has yet to be established [107]. Methods for the management of P E E P i are listed in Table 3. Conclusion In this review we have examined the physiological basis, the clinical implications, the measurement techniques and the treatment of P E E P i in critically ill patients during mechanical ventilation and weaning. P E E P i actually occurs much more frequently in the intensive care setting than was originally thought. However, all the published studies address mainly the physiological aspects and concern a small number of patients. More complete, longterm, randomised clinical trials are needed in the near future to investigate the clinical impact of PEEPi as well as the advantage of applying PEEP, and CPAP. Acknowledgements The authors of this article are indebted to Prof. J. Milic-Emili (Montreal, Canada) for all they have been privileged to learn from his lectures on the addressed topic and from many friendly discussions with him on both physiological mechanisms and clinical implications. They also want to express their gratitude to Prof. A. Braschi (Pavia, Italy) and to Prof. S.K. Pingleton (Kansas City, USA) for their valuable contributions to the discussion on the clinical impact of intrinsic PEEP. We would like to thank Miss Anna Ciucci for her help in preparing the manuscript. 534 References i. Pepe PE, Marini JJ (1982) Occult positive end-expiratory pressure in mechanically ventilated patients with airflow obstruction. Am Rev Respir Dis 126: 166-170 2. Simbruner G (1986) Inadvertent positive end-expiratory pressure in mechanically ventilated newborn infants: detection and effect on lung mechanics and gas exchange. J Pediatr 108:589-595 3. Rossi A, Gottfried SB, Zocchi L et al (1985) Measurement of static compliance of the total respiratory system in patients with acute respiratory failure during mechanical ventilation: the effect of "intrinsic" PEEP. Am Rev Respir Dis 131:672-767 4. Eberhard L, Guttmann J, Wolff G, Bertschmann W, Minzer A, Kohl H-J, Zeravik J, Adolph M, Eckart J (1992) Intrinsic PEEP monitored in the ventilated ARDS patient with a mathematical method. J Appl Physiol 73: 479-485 5. Agostoni E, Mead J (1964) Statics of the Respiratory System. In: Fenn WO, Rahn H (eds) Handbook of physiology, section 3: Respiration, vol I. American Physiological Society, Washington DC, pp 387-409 6. Martini JG, De Troyer A (1985) The thorax and control of functional residual capacity. In: Roussos C, Macklem PT (eds) Lung biology in health and disease, the thorax. (vol 29, Part B) Dekker, New York, pp 899-921 7. Rossi A, Polese G, Brandi G (1991) Dynamic hyperinflation. In: Marini J J, Roussos C (eds) Ventilatory failure. (vol 15) Spinger, Berlin, pp 199-218 8. Fuleihan SF, Wilson RS, Pontoppidan H (1976) Effect of mechanical ventilation with end-inspiratory pause on blood-gas exchange. Anesth Analg 55:122-130 9. D'Angelo E, Calderini E, Torri G, Robatto FM, Bono D, Milic-Emili J (1989) Respiratory mechanics in anesthetized paralyzed humans: effects of flow, volume, and time. J Appl Physiol 67:2556-2564 10. Polese G, Rossi A, Brandi Get al (1991) Partitioning of respiratory mechanics in mechanically ventilated patients. J Appl Physiol 71:2425-2433 11. Wright PE, Bernard GR (1989) The role of airflow resistance in patients with the adult respiratory distress syndrome. Am Rev Respir Dis 139:1169-1174 12. Wright PE, Marini JJ, Bernard GR (1989) In vitro versus in vivo comparison on endotracheal tube airflow resistance. Am Rev Respir Dis 140:10-16 13. Saetta M, Ghezzo H, Kim VD et al (1985) Loss of alveolar attachment in smokers. Am Rev Respir Dis 132: 894 - 900 14. Kimball WR, Leith DE, Robins AG (1982) Dynamic hyperinflation and ventilator dependence in chronic obstructive pulmonary disease. Am Rev Respir Dis 126:991-995 15. Collet PW, Roussos C, Macklem PT (1988) Respiratory mechanics. In: Murray F, Nadel J (eds) Textbook of respiratory Medicine, vol 1. Saunders, Philadelphia, pp 85-128 16. Gottfried SB, Rossi A, Higgs BD et al (1985) Non-invasive determination of respiratory system mechanics during mechanical ventilation for acute respiratory failure. Am Rev Respir Dis 131: 414-420 17. Gay CC, Rodarte JR, Hubmayr RD (1989) The effects of positive expiratory pressure on isovolume flow and dynamic hyperinflation in patients receiving mechanical ventilation. Am Rev Respir Dis 139:621-626 18. Petroff BJ, Legare M, Goldberg P, Milic-Emili J, Gottfried SB (1990) Continuous positive airway pressure reduces work of breathing and dyspnea during weaning from mechanical ventilation in severe chronic obstructive pulmonary disease. Am Rev Respir Dis 141: 281-289 19. Ranieri M, Giuliani R, Cinnella Get al (1993) Physiologic effects of positive end-expiratory pressure in patients with chronic obstructive pulmonary disease during acute ventilatory failure and controlled mechanical ventilation. Am Rev Respir Dis 147:5-13 20. Broseghini C, Brandolese R, Poggi R et al (1988) Respiratory mechanics during the first day of mechanical ventilation in patients with pulmonary edema and chronic airway obstruction. Am Rev Respir Dis i38:355-361 21. Tnxen DV, Lane S (1987) The effects of ventilatory pattern on hyperinflation airway pressures, and circulation in mechanical ventilation of patients with severe air-flow obstruction. Am Rev Respir Dis 136:872-879 22. Rossi A, Brandolese R, Milic-Emili J, Gottfried SB (1990) The role of PEEP in patients with chronic obstructive pulmonary disease during assisted ventilation. Eur Respir J 3:818-822 23. Broseghini C, Brandolese R, Poggi R et al (1988) Respiratory resistance and intrinsic positive end-expiratory pressure (PEEP) in patients with the adult respiratory distress syndrome (ARDS). Eur Respir J 1:726-731 24. Pesenti A, Pelosi P, Rossi Net al (1991) The effects of positive end-expiratory resistance in patients with the adult respiratory distress syndrome and in normal anesthetized subjects. Am Rev Respir Dis 144:101-107 25. Bernasconi M, Ploysongsang Y, Gottfried SB, Milic-Emili J, Rossi A (1988) Respiratory compliance and resistance in mechanically ventilated patients with acute respiratory failure. Intensive Care Med i4:547-553 26. Dal Vecchio L, Polese G, Poggi R, Rossi A (1990) "Intrinsic" positive end-expiratory pressure in stable patients with chronic obstructive pulmonary disease. Eur Respir J 3:74-80 27. Begin P, Grassino A (1991) Inspiratory muscle dysfunction and chronic hypercapnia in chronic obstructive pulmonary disease. Am Rev Respir Dis 143: 905-912 28. Haluszka J, Chartrand DA, Grassino A, Milic-Emili J (1990) Intrinsic PEEP and arterial PCO 2 in stable patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 141: 1194-1197 29. Ninane V, Yernault JC, De Troyer A (1993) Intrinsic PEEP in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 148:1037-1042 30. Aldrich TK, Hendler JM, Vizioli LD, Park M, Multz AS, Shapiro SM (1993) Intrinsic positive end-expiratory pressure in ambulatory patients with airway obstruction. Am Rev Respir Dis 147: 845- 849 31. Pradal U, Poggi R, Polese G, Braggion C, Mastella G, Rossi A (1994) Malnutrition not hyperinflation is the major cause of diaphragm weakness in cystic fibrosis. Am J Respir Crit Care Med 150:167- i73 32. Moore FA, Haenel JB, Moore EE, Abernathy CM (1990) Auto PEEP in patients with multisystemic failure: an ambiguous complicance. J Trauma 30:1316-1323 33. Gurevitch M J, Van Dyke JV, Young ES, Jackson K (1986) Improved oxygenation and lower peak airway pressure in severe adult respiratory distress syndrome. Chest 2:211-213 34. Abraham E, Yoshihara G (1989) Cardiorespiratory effects of pressure controlled inverse ratio ventilation in severe respiratory failure. Chest 96:1356-1359 35. Cole AGH, Weller SF, Sykes MK (1984) Inverse ratio ventilation compared with PEEP in adult respiratory failure. Intensive Care Med 10:227-232 535 36. Brandolese R, Broseghini C, Polese G 48. Rossi A, Santos C, Roca J Tortes A, 61. Rossi A, Polese G, De Sandre G (1992) et al (1993) Effects of intrinsic PEEP Felez MA, Rodriguez-Roisin R (1994) Respiration failure in chronic airflow on pulmonary gas exchange in mechanEffects of intrinsic PEEP on ventilaobstruction: recent advances and ically-ventilated patients. Eur Respir J therapeutic implications in the critically tion-perfusion mismatching mechan6:358-363 ically ventilated patients with acute or ill patients. Eur J Med 2:50-58 37. Gattinoni L, Marcolin R, Caspani ML chronic airway obstrcution. Am J 62. Rodriguez-Roisin R, Rossi A (1991) Aset al (1985) Constant mean airway presRespir Crit Care Med 149:1077-1084 sessment of lung function in the critisure with different patterns of positive 49. Conti G, Bunfi M, Antonelli M, Rocco cally ill patient. Clin Intensive Care pressure breathing during the adult re2:97 - 103 M, Gasparetto A (1989) Pressure supspiratory distress syndrome. Bull Eur port ventilation (PSV) reverses hyperin- 63. Marini JJ (1992) What derived variPhysiopathol Respir 21:275-279 flation-induced isorhythmic A-V dissoables should be monitored during me38. Conti G, Bufi M, Rocco Met al (1990) ciation. Intensive Care Med t5: chanical ventilation? Respir Care 37: 319-321 Auto-PEEP and dynamic hyperinflaI097 - 1107 tion in COPD patients during con- 50. Haake R, Schlichtig R, Ulstad DR, 64. Maltais F, Sovilj M, Ranieri VM, trolled mechanical ventilation and high Henschen R (1987) Barotrauma. PathoNavalese P, Gottfried SB (1992) Comfrequency jet ventilation. Intensive Care physiology, risk factors and prevention. parison between static and dynamic Med I6:81-84 Chest 91:608-613 measurement of intrinsic PEEP 39. Marini J J, Rodriguez RM, Lamb VJ 51. Pingleton SK (1988) Complications of (PEEPi) in mechanically ventilated pa(1986) The inspiratory workload of paacute respiratory failure. Am Rev Respir tients. Intensive Care Med [Suppl 2] tient-initiated mechanical ventilation. Dis 137:1463-1493 18:$94 Am Rev Respir Dis 134:902-909 52. Lachmann B (1992) Open the lung and 65. Bates JHT, Rossi A, Milic-Emili J 40. Marini JJ, Smith TC, Lamb VJ (1988) (1985) Analysis of the behaviour of the keep the lung open, Intensive Care Med External work output and force genera18:319-321 respiratory system with constant intion during synchronized intermittent 53. Corbridge TC, Wood LDH, Crawford spiratory flow. J Appl Physiol 58: mechanical ventilation. Am Rev Respir GP, Chudoba M J, Yanos J, Sznajder JI 1840-1848 Dis 138:1169-1179 (1990) Adverse effects of large tidal vol- 66. Rossi A, Gottfried SB, Higgs BD, Zoc41. Braschi A, Iotti G, Rodi G, Emmi V, ume and low PEEP in canine acid aspichi L, Grassino A, Milic-Emili J (1985) Sala Gallini G (1988) Dynamic pulmoration. Am Rev Respir Dis 142: Respiratory mechanics in mechanically nary hyperinflation (DPH) during in311-315 ventilated patients with respiratory failtermittent mandatory ventilation 54. Lichtwarck-Arschoff M, Nielsen JB, ure. J Appl Physiol 58:1849-1858 (IMV). Intensive Care Med 14:89 Sjostrand UH, Edgren EL (1992) An 67. Smith T, Marini JJ (1988) Impact of 42. Fernandez R, Benito S, Blanch LL, Net experimental randomized study of five PEEP on lung mechanics and work of A (1988) Intrinsic PEEP: a cause of indifferent ventilatory modes in a piglet breathing in severe airflow obstruction. spiratory muscle ineffectivity. Intensive model of severe respiratory distress, InJ Appl Physiol 65:i488-1499 Care Med 15:5i-52 tensive Care Med 18:339-347 68. Iotti G, Braschi A (1990) respiratory 43. Yang KL, Tobin M (1991) A prospective 55. Jonson B, Nordstrom L, Olsson SG, mechanics in chronic obstructive pulstudy of indexes predicting the outcome Akerback D (1975) Monitoring of venmonary disease. In: Vincent JL (ed) Inof trials of weaning from mechanical tilation and lung mechanics during autensive care medicine UPDATE 1990. ventilation. N Engl J Med 324: tomatic ventilation. A new device. Bull Springer, Berlin Heidelberg New York, 1445-1450 Physiopathol Respir 11:729-743 pp 223-230 44. Martin JG, Powell R, Shore S, Emrich 56. Suter PM, Fairley B, Isenberg MD 69. Gottfried SB, Reissman H, Ranieri MV J, Engel LA (1980) The role of respira(1975) Optimum end-expiratory in pa(1992) A simple method for the meatory muscles in the hyperinflation of tients with acute pulmonary failure. N surement of intrinsic positive end-expibronchial asthma. Am Rev Respir Dis Engl J Med 292:284-289 ratory pressure during controlled and 121:441 57. Gottfried SB, Rossi A, Milic-Emili J assisted modes of mechanical ventila45. Ninane V, Rypens F, Yernault JC, De (1986) Dynamic hyperinflation, instrintion. Crit Care Med 20:621-629 Troyer A (1992) Abdominal muscle use sic PEEP, and the mechanically ven- 70. Polese G, Rossi A, Brandi G, Ranieri during breathing in patients with tilated patient. Intensive Crit Care Dig VM, Giuliani R (1993) Partitioning of chronic airflow obstruction. Am Rev 5:30-33 intrinsic PEEP (letter) Am Rev Respir Respir Dis 146:16-21 58. Roussos C, Macklem PT (1986) InDis 148:1145- 1146 46. Martin JG, Shore S, Engle LA (1982) spiratory fatigue. In: Handbook of 71. Alberti A, Valenti S, Gallo F, Vincenti Effect of continuous positive airway physiology, sect 3: The respiratory sysE (1992) Differential lung ventilation pressure on respiratory mechanics and tem, vol III, part 2. American physiowith a double-lumen tracheostomy tube pattern of breathing in induced asthma. logical Society, Washington DC pp in unilateral refractory atelectasis. InAm Rev Respir Dis 126:812-817 511-527 tensive Care Med 18:479-484 47. Appendini L, Patessio A, Zanaboni Set 59. Milic-Emili J (1991) Control of ventila- 72. Kollef MH, Turner JF (1992) Intrinsic al (1994) Physiologic effects of positive tion and response to loading. In: PEEP and unilateral lung hyperinflaend-expiratory pressure in patients with Marini J J, ROUSSOSC (eds) Ventilatory tion. Phathophysiology and clinical sigchronic obstructive pulmonary disease failure. Springer, Berlin Heidelberg nificance. Chest I02:1220-1224 during acute exacerbations and non-inNew York, pp 37-48 73. Fernandez R, Mancebo J, Blanch LL, vasive ventilatory assistance. Am J 60. Benito S, Vallverdu I, Mancebo J (1991) Benito S, Claf N, Net A (1990) Intrinsic Respir Crit Care Med 149:I069-1076 Which patients need a weaning techPEEP on static pressure-volume curves. nique? In: Marini J J, ROUSSOSC (eds) Intensive Care Med 16:233-236 Ventilatory failure. Springer, Berlin Heidelberg New York, pp 419-429 536 74. Lemaire F, Beydon L, Jonson B (1991) Lung mechanics in ARDS. Compliance and pressure-volume curves. In: Zapol WM, Lemaire F (eds) Adult respiratory distress syndrome. Dekker, New York, pp 139-161 75. Sydow M, Burchardi H, Zinserling J, Ische H, Crozier TA, Weyland W (1991) Improved determination of static compliance by automated single volume steps in ventilated patients. Intensive Care reed 17:108-114 76. Braschi A, Iotti G, Gallini SA et al (1990) A new method of indirect evaluation for AutoPEEP during controlled ventilation (CMV). Am Rev Respir Dis 141:A577 77. Gottfried SB, Rossi A, Calverly PMA, Zocchi L, Milic-Emili J (1984) Interrupter technique for measurement of respiratory mechanics in anesthetized cats. J Appl Physiol 56:681-690 78. Gottfried SB, Higgs BD, Rossi A et al (1985) Interrupter technique for measurement of respiratory mechanics in anesthetized humans. J Appl Physiol 58:647-651 79. Shee CD, Ploysongsang 5(, Milic-Emili J (1985) Decay of inspiratory muscle pressure during expiration in conscious humans. J Appl Physiol 58:1859-1865 80. Hoffman RA, Ershowsky P, Krieger BP (1989) Determination of auto-PEEP during spontaneous and controlled ventilation by monitoring changes in endexpiratory thoracic gas volume. Chest 96:613 -616 81. Fernandez Mond6jar E, Vazquez Mata G, Navarrete Navarro P, Rivera Fernandez R, Torres Ruiz JM, Carazo E (1992) Increase in lung volume originated by extrinsic PEEP in patients with auto-PEEP. Intensive Care Med 18: 269-273 82. Brochard L, Isabey D, Piquet J et al (1990) Reversal of acute exacerbations of chronic obstructive lung disease by inspiratory assistance with a face mask. N Engl J Med 323:1523-1530 83. Gottfried SB (1991) The role of PEEP in the mechanically ventilated COPD patient. In: Marini J J, Roussos C (eds) Ventilatory failure. Springer, Berlin Heidelberg New York, pp 392-418 84. Gillespie DJ (1982) Comparison of intraesophageal balloon pressure measurements with a nasogastric esophageal balloon system in volunteers. Am Rev Respir Dis 126:583-585 85. Darioli R, Perret C (1984) Mechanically controlled hypoventilation in status asthmaticus. Am Rev Respir Dis 129: 385-387 86. Hickling KG, Henderson SJ, Jackson R (1990) Low mortality associated with low volume pressure limited ventilation with permissive hypercapnia in severe adult respiratory distress syndrome. Intensive Care Med 16:372-377 87. Poggi R, Brandolese R, Bernasconi M, Manzin E, Rossi A (1989) Doxofylline and respiratory mechanics. Chest 96: 772-778 88. Bernasconi M, Brandolese R, Poggi R, Manzin E, Rossi A (1990) Dose-response and time course effects of inhaled fenoterol on respiratory mechanics and arterial oxygen tension in mechanically ventilated patients with chronic airflow obstruction. Intensive Care Med 16:108-114 89. Gay PC, Rodarte JR, Tayyab M, Hubmayr RD (1987) Evaluation of bronchodilator responsiveness in mechanically ventilated patients. Am Rev Respir Dis 136:880-885 90. Navalesi P, Maltais F, Gursahaney A, Hernandez P, Sovilj M, Gottfried SB (1992) Cumulative dose-response relationship for salbutamol delivered by metered dose inhaler in mechanically ventilated patients with severe airflow obstruction. Intensive Care Med 18: $169 91. Manthous CA, Jesse BH, Schmidt GA, Wood LDH (1993) Metered-dose inhaler versus nebulized albuterol in mechanically ventilated patients. Am Rev Respir Dis 148:1567-1570 92. Newhouse MT, Fuller HD (1993) Rose is a rose is a rose? Am Rev Respir Dis 148:1444-1446 93. Mancebo J, Amaro P, Lorino H, Lemaire F, Harf A, Brochard L (1991) Effects of albuterol inhalation on the work of breathing during weaning from mechanical ventilation. Am Rev Respir Dis 144:95- 100 94. Nava S, Rubini F, Zanotti E, Fracchia C, Rampulla C (1992) Effects of steroids on respiratory mechanics in mechanically ventilated patients with COPD and respiratory failure. Intensive Care Med 18:$102 95. Sydow M, Burchardi H (1991) Intensive care management of life-threatening status asthmaticus. In: Vincent JL (ed) Intensive care and emergency medicine 14. Update 1991. Springer, Berlin Heidelberg New York, 313 -323 96. Maltais F, Sovilj M, Goldberg P, Gottfried SB (1992) Effect of inhalation anesthesia on respiratory mechanics in status asthmaticus. Am Rev Respir Dis 145:A77 97. 1hxen DV (1989) Detrimental effects of positive end-expiratory pressure during controlled mechanical ventilation of patients with severe airflow obstruction. Am Rev Respir Dis 140:5-9 98. Tobin MJ, Lodato FR (1986) PEEP, auto-PEEP and waterfalls. Chest 96: 449 - 451 99. Marini JJ (1989) Should PEEP be used in airflow obstruction? (editorial) Am Rev Respir Dis 140:1-3 100. Calderini E, Petrof BJ, Gottfried SB (1989) Continuous positive airway pressure (CPAP) improves the efficacy of pressure support (PS) ventilation in severe chronic obstructive pulmonary disease (COPD). Am Rev Respir Dis 139:A155 101. Georgopoulos D, Giannoulli E, Patakas D (1993) Effects of extrinsic positive end-expiratory pressure on mechanically ventilated patients with chronic obstructive pulmonary disease and dynamic hyperinflation. Intensive Care Med 19:197-203 102. Qvist J, Pemberton M, Bennike KA (1982) High-level PEEP in severe asthma. N Engl J Med 307:1347. 1348 103. McKee JD, Spevetz A, Simony P (1993) Auto PEEP (letter) N Engl J Med 329:212-213 104. Schivaram U, Miro AM, Cash ME et al (1993) Cardiopulmonary responses to continuous positive airway pressure in acute asthma. J Crit Care 8:87-92 105. Mansel JK, Stogner SW, Norman JR (1989) Face-mask CPAP and sodium bicarbonate infusion in acute, severe asthma and metabolic acidosis. Chest 96:943 - 944 106. Fleury B, Murciano D, Talamo D, Aubier C, Pariente R, Milic-Emili J (1985) Work of breathing in patients with chronic obstructive pulmonary disease in acute respiratory failure. Am Rev Respir Dis 132:822-827 107. Milic-Emili (1986) Is weaning an art or a science? Am Rev Respir Dis 134:1107-1108