INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Acute respiratory distress syndrome (ARDS) is characterized by a low respiratory compliance and a reduced aerated lung volume that result in compromised arterial oxygenation (1). Conventional mechanical ventilation with high tidal volume, high peak airway pressure, and high FI_O2 can have detrimental effects in patients with severe ARDS through ventilation- induced lung volutrauma and barotrauma (5). The concept of a "lung protective approach" in the ventilatory management of ARDS was first proposed by Hickling and colleagues (9). These investigators observed in two nonrandomized studies performed in patients with ARDS a reduction in the actual mortality as compared with the predicted mortality when a pressure- and volume-limited ventilatory strategy was implemented (9, 10). More recently, Amato and colleagues (11) evidenced an improvement in lung function and a decrease in mortality using a similar ventilatory design combined with an active strategy of alveolar recruitment in a series of patients with severe ARDS. However, permissive hypercapnia that invariably results from such a deliberate reduction of tidal volume has some potential drawbacks and may be poorly tolerated (14, 15). There is a consensus for contraindicating its use in patients with brain edema, coronary artery disease, severe metabolic acidosis, and severe hypoxemia (16). The addition of a continuous insufflation of gas in the trachea to conventional mechanical ventilation allows a reduction in Pa_CO2 (17). Insufflation of gas can be performed during the entire respiratory cycle (tracheal gas insufflation) or limited to the expiratory phase (expiratory washout) (24). In the former method, gas insufflation raises airway pressure through an increase in tidal volume and expiratory flow limitation. In expiratory washout (EWO) with volume-control ventilation, tidal volume remains unchanged, but expiratory flow limitation also occurs, generating intrinsic PEEP and reintroducing the risk of lung barotrauma (25). To maintain inspiratory plateau airway pressure constant, one possibility is to reduce the level of extrinsic PEEP of an amount equivalent to EWO-induced intrinsic PEEP. During conventional mechanical ventilation, another simple method for reducing Pa_CO2 is to increase the respiratory rate up to the limit of intrinsic PEEP and to decrease instrumental dead space by removing the respiratory tubing connecting the Y-piece and the proximal tip of the endotracheal tube. The goal of this prospective study was to compare the efficiency of these techniques used alone or in combination for decreasing Pa_CO2 in patients with severe ARDS treated by a pressure- and volume-limited ventilatory strategy.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

During a 12-mo period, six consecutive patients with ARDS diagnosed at the time of or after admission to the Surgical Intensive Care Unit (Department of Anesthesiology) of La Pitié-Salpêtrière Hospital (Paris, France) were prospectively included in the study. Written informed consent was obtained from each patient's next of kin. The study was approved by the Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale of la Pitié-Salpêtrière Hospital. Inclusion criteria were: (1) bilateral infiltrates on a bedside chest radiograph and bilateral and extensive hyperdensities on a high-resolution spiral thoracic computed tomographic (CT) scan; (2) pulmonary capillary wedge pressure less than 18 mm Hg and lack of left ventricular dysfunction as assessed by transesophageal echocardiography; (3) Pa_CO2  50 mm Hg using the following ventilatory settings with the patient sedated and paralyzed: volume-controlled mode with limited tidal volume to obtain a plateau airway pressure corresponding to the upper inflection point of the pressure-volume curve minus 2 cm H₂O, respiratory rate (RR) of 18 breaths/min, and PEEP equal to the lower infection point determined on the pressure-volume curve plus 2 cm H₂O. These inclusion criteria were intended to select patients with severe ARDS at risk for mechanical-ventilation-induced lung volutrauma in whom a pressure- and volume-limited ventilatory strategy was indicated. Exclusion criteria were: (1) head injury, (2) severe hypertensive disease, (3) acute coronary insufficiency, and (4) severe metabolic acidosis. These exclusion criteria were intended to exclude patients in whom permissive hypercapnia could have been harmful.

All patients were orally intubated with a Hi-Lo Jet no. 8 Mallinckrodt tube (Mallinckrodt Inc., Argyle, NY), which incorporates two internal side ports, one ending at the distal tip of the endotracheal tube, which was used for airway pressure monitoring, and a more proximal port ending 6 cm above, which was used for expiratory washout administration. Patients were sedated and paralyzed with a continuous intravenous infusion of fentanyl 250 µg/h, flunitrazepam 1 mg/h and vecuronium 4 mg/h and ventilated by a César ventilator (Taema, Antony, France) in a conventional volume-controlled mode. In all patients, hemodynamics were monitored using a fiberoptic thermodilution pulmonary artery catheter (CCO/Sv_O2/VIP TD catheter; Baxter Healthcare Corp., Irvine, CA) and a radial or femoral arterial catheter.

High-resolution and Spiral Thoracic Computed Tomographic Scan

In order to assess the severity of ARDS, extension of lung consolidation was quantified by a thoracic computed tomographic scan. Each anesthetized and paralyzed patient was transported to the Department of Radiology (Thoracic Division) and scanning was performed from the apex to the diaphragm (Tomoscan SR 7000; Philips, Eindhoven, The Netherlands). All images were observed and photographed at a window width of 1,600 Hounsfield units (HU) and a level of 700 HU. A spiral CT consisting of contiguous axial 10-mm-thick sections reconstructed from the volumetric data obtained during a 15-s apnea, the tracheal tube being disconnected from the ventilator, was acquired. A quantitative CT assessment of parenchyma consolidation at zero end-expiratory pressure (ZEEP) was performed according to a previously described technique (26). Briefly, the radiologist manually traced the right and left lung outlines with the roller ball on each spiral CT section from the apex to the diaphragm. Lung areas and mean lung density values were determined by using the region of interest function. Frequency histograms of the densities in HU were subsequently generated for each region of interest by using the analysis function. The frequency distribution of CT was computed for 50 compartments, from 1,000 HU to +100 HU, examining a 22-HU segment for each compartment. The frequency distribution of CT numbers of the entire lung was then calculated by adding the absolute values of each compartment. The lung volume of each compartment was calculated by multiplying the following: number of lung pixels times square pixel size times section thickness. Total lung volume was obtained by adding the lung volume of each compartment. In order to differentiate the lung zones with different degrees of ventilation, the entire lung was divided into three zones: lung zones between 1,000 and 500 HU were considered as normally aerated, those between 500 and 100 HU as poorly aerated and those between 100 and +100 HU as nonaerated (3).

Measurements

Systemic arterial pressure and pulmonary arterial pressure were measured simultaneously using the arterial cannula and the fiberoptic pulmonary artery catheter connected to two calibrated pressure transducers (91 DPT-308; Mallinckrodt) positioned at the midaxillary line. Airway pressure was measured proximally to the endotracheal tube connected to a third calibrated pressure transducer (91 DPT-308; Mallinckrodt). At the end of each phase, signals were recorded at a high sample rate of 100 Hz on a data acquisition and analysis system, including MP100 WS data acquisition system (Biopac Systems Inc., Goleta, CA) and a Quadra 610 Macintosh computer (Apple Computer Inc., Cupertino, CA) connected to the analog port of the hemodynamic monitor Merlin (Hewlett-Packard, Palo Alto, CA). Using the software AcqKnowledge included in the MP100 WS system, heart rate (HR), mean arterial pressure (), mean pulmonary arterial pressure (), pulmonary capillary wedge pressure (Ppcw), right atrial pressure (PRA), and peak inspiratory pressure (Pmax) were measured. In addition, after activating the corresponding knobs of the ventilator, inspiratory plateau airway pressure (Pplat) was measured at the end of an inspiratory pause of 3 s, and intrinsic PEEP was measured at the end of an expiratory pause of 5 s. Cardiac output was measured using the semicontinuous thermodilution technique (CCO/ Sv_O2/VIP TD catheter). Systemic and pulmonary arterial blood samples were simultaneously withdrawn and arterial pH, Pa_O2, Pv_O2, and Pa_CO2 were measured using an IL-BGE blood gas analyzer (Instrumentation Laboratories, Paris, France). Hemoglobin concentration, arterial and mixed venous oxygen saturations (Sa_O2 and Sv_O2) were measured using a calibrated OSM3 hemoximeter (Radiometer Copenhagen, Neuilly-Plaisance, France). Standard formulas were used to calculate cardiac index (CI), pulmonary vascular resistance index (PVRI), systemic vascular resistance index (SVRI), right ventricular stroke work index (RSWI), true pulmonary shunt (S/T), arteriovenous oxygen difference (C[a-v]O₂), oxygen delivery (DO₂), oxygen consumption (O₂), and oxygen extraction ratio (Ea_O2).

Pressure-volume curves were obtained in ZEEP using an inspiratory occlusion technique derived from the one described by Levy and colleagues (27). Briefly, Pplat was measured at different randomized tidal volumes (100-ml increments) using the end-inspiratory pause hold functions of the César ventilator, while respiratory rate remained constant at 18 breaths/min. Before administering a given tidal volume, intrinsic PEEP was determined by activating the end-expiratory pause knob. The maximal plateau pressure authorized was 35 cm H₂O. After tracing the pressure-volume curve, static respiratory compliance (Crs) was calculated as the slope of the pressure-volume curve in its linear part, and lower and upper inflection points were visually determined. Quasi-static respiratory compliance (Cqs) was determined by dividing the tidal volume of the patient by the corresponding end- inspiratory pressure minus intrinsic PEEP.

In each patient, expired CO₂ was measured using a nonaspirative calibrated 47210A infrared capnometer (Hewlett-Packard, Andover, MA) positioned between the proximal end of the endotracheal tube and the Y-piece of the ventilator. Expiratory CO₂ curves were recorded on the MP100 WS data acquisition system. After simultaneous withdrawal of an arterial blood sample, the ratio of alveolar dead space (VDA) to VT was calculated according to the following equation:

where PET_CO2 is end-tidal CO₂ measured at the plateau of the expired CO₂ curve. Because expiratory washout markedly interfered with the expired CO₂ curve, VDA/VT was calculated only during conventional and optimized mechanical ventilation.

In Patient 1, 10 ml of arterial blood were withdrawn at the end of each phase in order to determine epinephrine and norepinephrine plasma concentrations using a radioenzymatic assay based on the enzymatic methylation of epinephrine and norepinephrine by catecho- O-methyltransferase in the presence of triated S-adenosyl-L-methionine (radiolabeled SAMe³ H; Amersham). Normal values were 200 to 300 pg/ml for norepinephrine (intra-assay variability, 4.2%; interassay variability, 7.5%) and 25 to 55 pg/ml for epinephrine (intra-assay variability, 6%; interassay variability, 10%). In the five patients receiving exogenous catecholamines for the treatment of their septic shock, accurate measurements were impossible because of the high concentrations of catecholamines found in the plasma.

Expiratory Washout Administration

The ventilator used for the study was a César ventilator (Taema) equipped with a Fisher-Paykel humidifier positioned at the proximal tip of the inspiratory limb. The expiratory washout flow generator (Taema) was synchronized with the expiratory phase of the ventilator by means of a flow sensor connected to the inspiratory limb of the ventilator giving the signal to interrupt the expiratory washout flow when the inspiratory cycle started. An expiratory washout flow of 15 L/min was delivered throughout the entire expiration via the proximal channel of the endotracheal tube, itself connected by a 3-mm internal diameter tube (Argyle; Sherwood Medical, Tullamore, Ireland) to the EWO module. Because the internal volume of this circuit was 15.5 ml and the driving pressure 1.3 bar, 4.7 ml of gas were decompressed when EWO was interrupted and participated to the next inspiration.

Protocol

The study was performed over a 2-d period. On the first day, the protocol consisted of checking inclusion and exclusion criteria, obtaining initial hemodynamic and respiratory measurements in normocapnia, and performing the CT scan. At the end of the first day, permissive hypercapnia was implemented according to the following rationale: using conventional volume-controlled mechanical ventilation, a PEEP equal to the lower inflection point plus 2 cm H₂O was administered and respiratory rate and inspiratory/expiratory (I/E) ratio were set at 18 breaths/min and 33%. Then, VT was reduced in order to obtain a Pplat equal to the upper inflection point minus 2 cm H₂O or to 25 cm H₂O in the absence of an upper inflection point.

The study itself was performed on the second day approximately 12 h after the implementation of permissive hypercapnia. As shown in Figure 1, three ventilatory modes were tested in a randomized order: optimized mechanical ventilation using a respiratory frequency of 30 ± 4 breaths/min, expiratory washout, and the combination of the two. Each mode was separated from the preceding by a 1-h period of conventional mechanical ventilation with permissive hypercapnia (control). For each phase, a 1-h steady state was observed before measurements. Optimized mechanical ventilation consisted of increasing the respiratory rate to the maximal rate possible without development of intrinsic PEEP and removing the 15-cm-long tubing connecting the proximal tip of the endotracheal tube and the Y-piece. When increasing respiratory frequency, the occurrence of intrinsic PEEP was detected by the visual analysis of the expiratory flow displayed on the screen of the César ventilator. During expiratory washout and the combination of both modes, extrinsic PEEP was reduced in order to maintain the Pplat at its control value. As a consequence, all ventilatory conditions were compared at the same Pplat.

View larger version (26K): [in this window] [in a new window]   Figure 1.   Representation of the three different ventilatory modes. Each technique is represented at end-expiration in a separate quadrant with the ventilator César (V), inspiratory and expiratory limbs, the Y-piece, the tubing connecting the Y-piece to the endotracheal tube, the Mallinckrodt endotracheal tube and its two additional channels, and the tracheal carina. The dotted area represents respiratory tubing and upper airways with CO2-laden gas. For each ventilatory technique, the corresponding flow is represented in the inferior part of the quadrant. Expiratory washout (EWO) is indicated by an additional flow superimposed on the expiratory flow. During optimized mechanical ventilation (OPTIMV), respiratory frequency is increased to the limit of intrinsic PEEP, and the tubing connecting the Y-piece to the endotracheal tube is removed.

Statistical Analysis

All data are presented as mean ± SD in the text and the tables and as mean ± SEM in the figures. Control conditions (conventional mechanical ventilation) and the three experimental conditions (optimized mechanical ventilation, expiratory washout, and the combination of these) were compared by a one-way analysis of variance followed by a Bonferroni multiple comparison test. Each condition was compared with its previous control by Student's paired t test. The interaction between optimized mechanical ventilation and expiratory washout was tested by a two-way analysis of variance for two within factors. Level of significance was considered as 5%.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Individual characteristics of the patients are shown in Table 1. Five of the six patients received a continuous infusion of catecholamines for an associated septic shock. Survival rate was 50%. According to the density histogram analysis, 41 ± 16% of the lung parenchyma was nonaerated, 41 ± 17% was poorly aerated, and only 18 ± 21% was normally aerated. The individual pressure-volume curves of the patients are shown in Figure 2. Five patients had a Pplat above the upper inflection point determined on the individual pressure-volume curve when ventilated using conventional mechanical ventilation with PEEP. Individual hemodynamic and respiratory parameters measured the first day of the study with a FI_O2 of 100% and using conventional mechanical ventilation with PEEP are summarized in Table 2.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Individual pressure-volume curves of the total respiratory system obtained with the occlusion technique. All patients exhibited a lower inflection point, and all but one (Patient 4) exhibited an upper inflection point.

Cardiorespiratory Effects of Reducing Tidal Volume during Conventional Mechanical Ventilation

Tidal volume was reduced by 43 ± 17% in order to reduce the Pplat to a mean value of 24 ± 3 cm H₂O, i.e., 2 cm H₂O below the mean upper inflection point. The decrease in VT resulted in respiratory acidosis: pH decreased from 7.39 ± 0.06 to 7.21 ± 0.11 and Pa_CO2 increased from 47 ± 6 to 84 ± 24 mm Hg (p < 0.05). Simultaneously, increased from 24 ± 5 to 29 ± 4 mm Hg (p < 0.05) and VDA/VT decreased from 35 ± 10 to 25 ± 8%, corresponding to a decrease in alveolar dead space from 236 ± 63 to 100 ± 47 ml. All the other parameters measured did not change significantly.

Cardiorespiratory Effects of the Three Ventilatory Modes

The hemodynamic and ventilatory parameters measured during the six phases of the study are shown in Table 3. The three control phases (conventional mechanical ventilation) were similar for all the parameters tested. As shown in Figure 3, optimized mechanical ventilation and expiratory washout had similar effect on Pa_CO2 (Pa_CO2 = 28 ± 11 versus 27 ± 12%), pH (pH = +0.15 ± 0.04 versus +0.14 ± 0.03) and ( = 4 ± 3 versus 5 ± 4 mm Hg). No significant changes in Pa_O2 and S/T were observed. Extrinsic PEEP remained unchanged during optimized mechanical ventilation, but it had to be reduced by 44% during expiratory washout in order to maintain the Pplat constant. This result suggests that expiratory washout implemented at a respiratory rate of 18 breaths/min and at a flow of 15 L/min induced an intrinsic PEEP of 5 ± 2 cm H₂O. The addition of expiratory washout to optimized mechanical ventilation further decreased Pa_CO2 by 18 ± 5%. As a consequence, the combination of optimized mechanical ventilation and expiratory washout had additive effects and decreased Pa_CO2 by 46 ± 7% as compared with conventional mechanical ventilation (Figure 3). Pa_CO2 was below 50 mm Hg in four of the six patients studied when optimized mechanical ventilation was combined with expiratory washout. In order to maintain Pplat constant, extrinsic PEEP had to be further reduced to 4 ± 2 cm H₂O (61 ± 11%), and a slight but nonsignificant deterioration in arterial oxygenation was observed.

View larger version (14K): [in this window] [in a new window]   Figure 3.   Changes in PaCO2, inspiratory plateau airway pressure (Pplat), PEEP, and PaO2 (FIO2-1) induced by optimized mechanical ventilation (OPTIMV), expiratory washout (EWO), and the combination of OPTIMV and EWO in six patients with severe ARDS. Control values (control) were obtained during conventional mechanical ventilation using a respiratory frequency of 18 breaths/min and the presence of a 15-cm-long tubing connecting the Y-piece to the endotracheal tube.

The left panel of Figure 4 shows the profile of variations of PET_CO2 when switching from expiratory washout, optimized mechanical ventilation, or the combination of both to conventional mechanical ventilation. A steady state in PET_CO2 values was observed after 30 min. The right panel of the figure shows the decrease in PET_CO2 when switching the ventilatory mode from conventional mechanical ventilation to optimized mechanical ventilation. A steady state was also obtained after 30 min. In Patient 1, who did not receive exogenous catecholamines, norepinephrine blood concentration decreased from 776 to 198 pg/ml (75%), from 667 to 273 pg/ml (71%), and from 956 to 118 pg/ml (82%) with optimized mechanical ventilation, expiratory washout, and the combination of the two, respectively. Simultaneously, epinephrine blood concentration decreased from 390 to 76 pg/ml (81%), from 363 to 103 pg/ml (77%), and from 449 to 38 pg/ml (89%).

View larger version (18K): [in this window] [in a new window]   Figure 4.   (Left panel ) Profiles of increase in end-tidal CO2 (PETCO2) when switching the ventilatory mode from expiratory washout, optimized mechanical ventilation, and the combination of both modes to conventional mechanical ventilation. For each patient, PETCO2 was measured each minute after the interruption of the mode, and the mean value of PaCO2 corresponding to the three ventilatory modes was calculated. The mean value of the six patients is represented in the figure. (Right panel ) Profile of the decrease in PETCO2 when switching from conventional mechanical ventilation to optimized mechanical ventilation in the six patients of the study.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study, performed in patients with severe ARDS, demonstrates that Pa_CO2 levels resulting from a pressure- and volume-limited ventilatory strategy can be maintained within safe limits when combining expiratory washout with an optimization of mechanical ventilation. As a consequence many of the deleterious effects resulting from severe hypercapnia such as pulmonary hypertension, tachycardia, high cardiac output, and catecholamine release can be avoided in these patients.

Permissive Hypercapnia: Rational and Cardiopulmonary Effects

The patients included in the present study had severe ARDS as attested by the CT scan analysis and the alteration of their respiratory mechanics. Less than 20% of the lung parenchyma was normally aerated, whereas about 80% was either poorly aerated or nonaerated on the CT scan performed at ZEEP. In five of the six patients included, an upper inflection point was identified on the pressure-volume curve, with values ranging between 24 and 33 cm H₂O, as previously reported (28). It is very likely that many of these patients were at risk of ventilator-induced lung barotrauma since, at the end of each respiratory cycle, the Pplat was above the upper inflection point when ventilatory parameters were set to provide normocapnia. When VT was reduced to prevent lung barotrauma, alveolar dead space decreased significantly by an amount of 136 ± 81 ml, strongly suggesting that mechanical ventilation set up for providing normocapnia was associated with an overdistension of some parts of the lung parenchyma. This result is in accordance with the work by Kiiski and coworkers (29) who demonstrated that VDA/VT remains unchanged when VT is varied in patients with ARDS, suggesting that alveolar dead space increases with high tidal volumes and decreases with low tidal volumes. This dependency of dead space upon tidal volume seems specific to ARDS since it was not observed in healthy volunteers (30) or in patients with chronic obstructive pulmonary disease (31).

Permissive hypercapnia implemented over a 12-h period resulted in an increase in Ppa secondary to a nonsignificant increase in cardiac output and capillary wedge pressure. Neither Pa_O2 nor shunt changed significantly. These hemodynamic and respiratory changes were less pronounced than those occurring after an acute reduction in VT where a marked rise in pulmonary vascular resistance and shunt fraction is classically observed (14). Differences between the hemodynamic effects of acute and chronic permissive hypercapnia are likely related to differences in blood pH, the kidney allowing a time-dependant increase in bicarbonate plasma level with a concomitant increase in blood pH (12, 32). It is also clear from the present study that a steady state in terms of Pa_CO2 levels was obtained 30 min after acute changes in Pa_CO2 resulting from changing the ventilatory mode.

Cardiopulmonary Effects of Optimized Mechanical Ventilation and Expiratory Washout

Optimized mechanical ventilation improved alveolar ventilation by increasing minute ventilation (E) through the increase in respiratory rate and by the reduction of instrumental dead space. In intubated and mechanically ventilated patients, instrumental dead space is composed of the endotracheal tube, the head exchanger filter, and the tubing connecting the Y-piece to the endotracheal tube. In the present study, the head exchanger filter was replaced by a humidification chamber, and the tubing connecting the Y-piece to the endotracheal tube was removed during optimized mechanical ventilation. This modification reduced the total instrumental dead space by 40 ml, which decreased from 106 ml during conventional mechanical ventilation to 66 ml during optimized mechanical ventilation. It has to be pointed out that the pneumotachograph and the capnometer had a total internal volume of 27 ml and were also an integral part of the instrumental dead space during conventional mechanical ventilation and the three methods tested. In clinical practice, these devices are not routinely used, and it can be speculated that their removal would have led to an additional reduction in Pa_CO2. In this category of patients with severe ARDS and reduced lung compliance, the respiratory frequency could be increased to values around 30 breaths/min without generating intrinsic PEEP because of the stiffness of the respiratory system, allowing the lungs to completely deflate in an expiratory time less than 1.5 s.

As previously reported, expiratory washout decreased Pa_CO2 by removing the CO₂-laden gas occupying the endotracheal tube and the tubing connecting the proximal tip of the endotracheal tube and the Y-piece (25, 33, 34). As a consequence, the interface between fresh inspiratory gas and CO₂-laden gas is shifted to the distal part of the trachea, resulting in a reduction of dead space to tidal volume ratio. There is also a superimposed jet effect that produces projections beyond the site of administration and removes further CO₂-laden gas from the distal portion of the tracheobronchial tree (18, 25, 35, 36). The efficacy of expiratory washout observed in our study was similar to the one described by Nahum and colleagues in dogs (34) and by Kalfon and colleagues (25) in humans. Although the combination of optimized mechanical ventilation and expiratory washout resulted in a spectacular reduction in Pa_CO2 that even normalized in four patients, expiratory washout was less efficient during optimized than during conventional mechanical ventilation. It is highly likely that this difference was related to the lower end-tidal fraction of CO₂ during optimized mechanical ventilation. Nahum and colleagues (34) observed in hypercapnic dogs with oleic-acid-induced lung injury that expiratory washout decreased Pa_CO2 by 29 ± 5% when administered with a VT of 10 ml/kg, whereas Pa_CO2 decreased by only 19 ± 3% when administered with a VT of 15 ml/kg (34). This result is explained by the particular relationship between Pa_CO2 and the dead space to tidal volume ratio, which is curvilinear. Assuming a constant metabolic CO₂ production, at high level of Pa_CO2 resulting from a small VT and a low E, a given EWO-induced decrease in dead space, provides a significant drop in Pa_CO2 because the system operates on the rectilinear part of the curve. If alveolar ventilation is increased by increasing respiratory rate, the Pa_CO2 level decreases and the same EWO-induced decrease in dead space will be associated with a less important decrease in Pa_CO2 because the system operates on the flat part of the curve (34). In other words, the greater the end-tidal fraction of CO₂, the more efficient is the "washout effect" on CO₂ elimination.

Expiratory Washout-induced Intrinsic PEEP

A significant increase in airway pressure and in lung volumes is a well-known side effect of tracheal gas insufflation and correlates with the flow used (17, 18). When insufflation of gas is limited to the expiratory phase (expiratory washout), VT remains virtually unchanged during volume-control ventilation, but airway pressures can still increase through expiratory flow limitation and intrinsic PEEP (25). Kalfon and colleagues (25), using an expiratory washout flow of 15 L/min, observed a 26% increase in Pplat related to expiratory flow limitation. In the present study, extrinsic PEEP was adjusted in order to maintain Pplat constant between the three ventilatory modes. This method allows an indirect evaluation of expiratory washout-induced intrinsic PEEP. As expected, intrinsic PEEP was around 5 cm H₂O during expiratory washout (25) and increased to 8 ± 2 cm H₂O when expiratory washout was combined with optimized mechanical ventilation. The different behavior of OPTIMV+EWO and EWO in terms of intrinsic PEEP might be explained by the following. The EWO module that was used in the present study stops the expiratory gas flow 40 ms after the beginning of inspiration, thereby increasing the VT by 10 ml. Five additional milliliters resulting from the decompression of the gas (see METHODS, EXPIRATORY WASHOUT ADMINISTRATION) are also delivered during early inspiration. Therefore, EWO increased VT by 3.5%. When respiratory frequency was low, this small increase in VT did not result in any detectable intrinsic PEEP because the respiratory system had enough time to empty. During OPTIMV, where respiratory rate was increased to the maximal rate possible without development of intrinsic PEEP, by definition any increase in VT results in the occurrence of intrinsic PEEP. Therefore, it is not surprising that extrinsic PEEP had to be decreased significantly more during OPTIMV+EWO than during OPTIMV alone. This result outlines a critical issue of EWO: the flow generator should interrupt the expiratory flow a few milliseconds before the beginning of the inspiratory phase in order to avoid any increase in VT. In addition, it can be expected that ending the expiratory flow before the inspiratory phase may prevent the occurrence of any intrinsic PEEP. Further studies are needed to verify this hypothesis. Because mean airway pressure was carefully controlled throughout the study, no significant variation in Pa_O2 was found between the three ventilatory modes. This result contrasts with the 44% increase in Pa_O2 described by Kalfon and colleagues (25) when using expiratory washout without controlling mean airway pressure. In this previous study, the increase in Pa_O2 correlated with the increase in mean airway pressure because of expiratory washout-induced intrinsic PEEP, suggesting that the improvement in arterial oxygenation resulted from an additional alveolar recruitment. This hypothesis is confirmed by the fact that arterial oxygenation did not improve in the present study in which a careful control of airway pressure was obtained. As a consequence, these results suggest that expiratory washout per se has no direct effect on alveolar recruitment when intrinsic PEEP does not induce an increase in mean airway pressure.

In conclusion, during conventional mechanical ventilation, increasing respiratory frequency and reducing instrumental dead space is as efficient as expiratory washout for reducing Pa_CO2 in patients with severe ARDS and treated by permissive hypercapnia. If expiratory washout-induced intrinsic PEEP is counterbalanced by a reduction of extrinsic PEEP in order to maintain the same level of mean airway pressure, Pplat and arterial oxygenation remain constant. When expiratory washout is combined with optimized mechanical ventilation, a spectacular decrease in Pa_CO2 is observed, resulting in a quasi-normal CO₂ elimination in a majority of patients and questioning the concept of "permissive hypercapnia."