INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During the early phase of acute lung injury (ALI), opening the recruitable lung regions and keeping them open has been advocated as a lung protective ventilatory strategy to minimize ventilator-associated lung injury (1). Volume recruitment maneuvers (RM) have been proposed as ventilatory techniques to maximize lung recruitment (2, 3). Although the concept of a RM was first described in 1952 (4), such maneuvers were only recently employed as an integral part of a lung protective ventilatory strategy in acute respiratory distress syndrome (ARDS) (5). Large tidal volumes (VT) used either during mechanical ventilation (6) or applied as sighs (7) have been shown to recruit lung volume and improve oxygenation of patients with ARDS. Other investigators have used sustained inflations (SI) coupled with conventional or high-frequency ventilation to recruit the lung (8). Previous work in our laboratory (11) demonstrated a sustained (15-min) improvement in Pa_O2 and venous admixture (/T) after a RM during ventilation with a relatively large VT (15 ml/kg) in a lung lavage but not in an oleic acid-induced lung injury (OAI) model in dogs.

Prone positioning is another adjunct to a lung protective ventilatory strategy that may promote lung recruitment by decreasing regional pleural pressure gradients (12). In animal models, the effect of prone positioning on oxygenation is similar to that of positive end-expiratory pressure (PEEP). In a canine lung lavage model, Lim and colleagues (15) recently demonstrated that similar levels of Pa_O2 can be achieved at a lower PEEP during prone as compared with supine position. Consequently, reduction in pleural pressure gradients during prone positioning may achieve a greater degree of lung expansion for a given PEEP level and a RM.

We examined the interactions between position and RM in dogs after OAI. We explored two hypotheses regarding the oxygenation response to a sustained inflation employed as a RM: (1) the increment in Pa_O2 after a RM would be greater in the prone position, and (2) less PEEP would be required to maintain the oxygenation benefits of a RM in the prone compared with the supine position.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal Preparation and Measurements

Twelve adult mixed-breed dogs of either sex (25.7 ± 3.3 kg; mean ± SD), were included in this study using a protocol approved by the Animal Care and Use Committee of Regions Hospital. Each dog was anesthetized with an intravenous bolus of 30 mg/kg of sodium pentobarbital and tracheotomized and intubated with an 8 mm interior diameter endotracheal tube (Mallinckrodt, Argyle, NY). Mechanical ventilation was initiated in constant-flow, volume-cycled mode (7200; Mallinckrodt, Carlsbad, CA), with a VT of 8 ml/kg, a frequency of 15 breaths/min, a PEEP of 3 cm H₂O, an inspiratory to expiratory time ratio of 1:2, and an inspired oxygen fraction of 0.60. The dogs were immobilized by an intravenous injection of pancuronium (0.1 mg/kg). Muscle relaxation and anesthesia were maintained by continuous infusions of pentobarbital (2 to 4 mg/kg/h) and pancuronium (0.1 mg/kg/h) throughout the experimental protocol.

Femoral artery and vein catheters were inserted via an incision and sutured in place. Systemic mean arterial pressure () and heart rate (HR) were monitored via a femoral artery catheter. Arterial pH, PCO₂, and PO₂ were measured continuously using an intravascular blood gas monitoring system (Paratrend 7; Diametrics, St. Paul, MN) inserted via the femoral artery catheter. A pulmonary artery (PA) catheter was introduced via the right external jugular vein, and its tip was positioned in the PA for the measurement of pulmonary artery pressures including balloon-occlusion pressure (Ppao). Pressure tracings were used to verify the positions of the proximal and distal ports of the PA catheter in the right atrium and PA, respectively. All intravascular pressures were measured with low displacement transducers (Model PX600F; Allegiance, Irvine, CA) referenced and zeroed to midchest level. The zero reference position was not changed when the dog was turned prone. Thermodilution cardiac output (T) was measured by a bolus injection of cold 5% dextrose delivered during expiration (0° C; Model 9520A, Allegiance, Irvine CA). Arterial and mixed venous blood gases were analyzed at 37° C (Model 248; Chiron, Medfield, MA) and corrected to core temperature.

Airway opening pressure (Pao) was measured at the lateral pressure tap attached to the proximal end of the endotracheal tube (ETT) using a differential pressure transducer (± 100 cm H₂O, MP-45; Validyne Corp., Northridge, CA). Total PEEP (PEEPtotal = PEEP + auto-PEEP) was measured by occlusion of the airway at end-expiration using a Braschi valve (16).

Lung volumes were measured continuously with a respiratory inductive plethysmograph (RIP) (Respigraph; NIMS, Miami Beach, FL). The RIP was calibrated after initiation of muscle relaxation (17). The bands of the RIP were repositioned and the RIP was recalibrated after stabilization of OAI and each position change. FRC at zero cm H₂O of PEEP (ZEEP) was measured using the helium dilution technique after 20 s of apnea (18).

After recording baseline hemodynamic, gas exchange, and respiratory mechanics measurements in the supine position at the ventilatory settings described previously (pre-OAI stage), oleic acid (0.09 ml/kg) was slowly injected into the right atrium via the proximal port of the PA catheter. This was followed by a 90-min injury stabilization period before initiating the experimental protocol. After OAI, all dogs received a continuous intravenous saline infusion at a rate of 100 ml/h.

Experimental Protocol

The time line of the experimental protocol is schematically depicted in Figure 1. After stabilization of OAI, defined as a less than 5% change in the continuously monitored Pa_O2 over a 15-min period, the hemodynamic, gas exchange, and respiratory mechanics measurements were repeated (post-OAI first baseline period, B1) using the same ventilatory settings as in the pre-OAI stage. Subsequently, the dogs were randomized to one of the two groups.

View larger version (24K): [in this window] [in a new window]   Figure 1.   Schematic diagram of the experimental protocol. At each stage, respiratory mechanics, hemodynamic and gas exchange parameters were measured. Arterial and mixed venous blood gases were obtained before a RM (t = 2 min) and at 0, 2, 4, 6, 10, and 15 min after the RM. After completion of the baseline stage after oleic acid injury (post-OAI, B1), the dogs were randomized to one of two groups (S-P and P-S groups). The sequence of position at each stage is also shown in the figure (supine or prone).

SP group. In this group the dogs were kept supine and ventilated with a PEEP of 8 and 15 cm H₂O sequentially (without changing any of the other ventilatory settings). At each PEEP level, the dogs were ventilated for 15 min and a set of gas exchange, hemodynamic, and lung mechanics measurements were made (pre-RM stage, t = 2 min). We then disconnected the ETT from the ventilator (i.e., applied ZEEP) followed by 20 s of apnea while continuously monitoring lung volume changes by the inductive plethysmograph. In this manner, we were able to measure end-expiratory lung volume above ZEEP (V). Subsequently, we measured FRC at ZEEP (ZEEP-FRC) by the He-dilution technique. We then applied a RM consisting of a single 30 s SI at 60 cm H₂O airway pressure and continued mechanical ventilation at the previous PEEP level for 15 min while sampling arterial and mixed venous blood gases at 0, 2, 4, 6, 10, and 15 min. Time zero was defined as 30 s of ventilation after the completion of the RM and the 15-min point was designated as the post-RM stage for that position and PEEP level. We then dropped PEEP to zero cm H₂O, applied 20 s of apnea, and repeated V and ZEEP-FRC measurements. The dogs were then turned to the prone position and the same stages that were performed during the supine position were duplicated.

PS group. In this group, identical stages as in the SP group were performed, but the order of positioning was first prone then supine. To verify the stability of the experimental model, in both groups, after completion of the supine and prone stages the animals were turned supine, PEEP of 3 cm H₂O (identical to post-OAI B1 stage) was applied, and hemodynamic, gas exchange, and respiratory mechanics were measured after 15 min (post-OAI second baseline stage, B2). Subsequently, a pressure-volume (PV) curve of the respiratory system was constructed from ZEEP using the super-syringe technique and the lower inflection point of the PV curve (LIP) was measured (19).

Data Acquisition and Analysis

Pao and lung volume changes measured by the RIP were simultaneously recorded on a chart recorder (95000; Astro-Med, West Warwick, RI) and stored on a personal computer. Three randomly chosen, consecutive breaths were used for data analysis in each experimental stage. End-inspiratory plateau pressure (Pplat) was measured at the airway opening after a 2-s end-inspiratory pause and used to calculate total respiratory system compliance (Crs) (19). Data analysis was performed using a data acquisition and analysis software program (LabVIEW; National Instruments, Austin, TX). /T was calculated as described previously (19).

Statistical Analysis

In each group, paired t tests were used to compare hemodynamic, respiratory mechanics, and gas exchange parameters before and after OAI (pre-OAI versus post-OAI B1 stage) and to compare the two baseline stages after OAI (post-OAI B1 versus B2 stages). The effects of group, position, and PEEP on the measured parameters were examined by a three-way analysis of variance (ANOVA). Tukey's compromise post hoc analysis was performed if the global F-test indicated statistical significance (p < 0.05). The change in Pa_O2 and /T over time after a RM at each stage was analyzed using a repeated ANOVA and Tukey's post hoc analysis to compare the value immediately before the RM (pre-RM, t = 2 min) with each of the subsequent seven time points. At each stage, the change in V and ZEEP-FRC before and 15 min after a RM was tested by a paired t test. A p value of < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of OAI

There were six dogs in each group with similar weights (SP: 25.6 ± 3.2 kg and PS: 24.7 ± 3.5 kg). As expected, peak, mean and end- inspiratory plateau airway pressures increased (p < 0.03) after OAI except the increase in mean Pao in the PS group did not reach statistical significance (Table 1). Total Crs decreased (p < 0.01) in both groups after OAI (Table 1). OAI did not affect hemodynamic parameters significantly except for Ppao (Table 2), whereas gas exchange deteriorated as evidenced by a fall in Pa_O2 (p < 0.001) and pH (p < 0.01), and a rise in /T (p < 0.001) and Pa_CO2 (p < 0.001, Table 3). The LIP of the inflation PV curves measured at the end of the experimental protocol was 13 ± 2 cm H₂O.

Comparison of Post-OAI Baseline Stages (B1 and B2)

In both groups, hemodynamic, respiratory mechanics, and gas exchange parameters were similar between the two baseline stages that bracketed the experimental stages after OAI (B1 versus B2; Tables 1, 2, and 3), except for central venous pressure (Pcv) which increased slightly by the end of the experiment in the PS group (p = 0.042, Table 2). At the same time, mean pulmonary arterial pressure () and pulmonary vascular resistance (Rpv) increased slightly in the SP group (p < 0.05, Table 2). The relative constancy of these parameters indicates that the experimental model was stable in both groups throughout the experimental protocol.

Effect of Group, Position, and PEEP

In both groups, the measured Pao increased significantly when PEEP was increased (p < 0.001), but Crs was not affected by PEEP (Table 1). Pulmonary artery pressures (mean and Ppao) and Pcv increased significantly with PEEP (p < 0.04), but the other measured hemodynamic parameters did not change with PEEP (Table 2). In both groups before a RM (pre-RM, t = 2 min), increasing PEEP raised Pa_CO2 (p < 0.02) with a concomitant fall in pH (p < 0.03), improved Pa_O2 (p < 0.01), and decreased /T (p < 0.01, Table 3). A similar pattern was observed for the gas exchange parameters in both groups after a RM (15 min post-RM, Table 4).

Position did not influence the respiratory mechanics and hemodynamic parameters except for T and Pcv which tended to increase slightly during prone as compared with supine position (p < 0.03, Table 2). The effect of position was significant in the pre-RM oxygenation parameters (p < 0.01) except for pH and Pa_CO2 (Table 3). Before application of a RM, Pa_O2, (Pv_O2), and /T demonstrated a statistically significant interaction in the ANOVA analysis between group and position (p < 0.01, Table 3). Similar observations were made for the post-RM conditions, except group and position for Pv_O2 lacked a significant interaction (Table 4).

At each stage, respiratory mechanics, hemodynamic, and gas exchange parameters were similar between the groups, except Ppao was slightly lower in the PS group relative to the SP group (p = 0.043, Table 2).

In all of the stages, lung volume above ZEEP (V) increased significantly after a RM (p < 0.02, Figure 2). Before and after a RM, V was affected significantly only by PEEP (p < 0.03), and not by group or position (Figure 2).

View larger version (32K): [in this window] [in a new window]   Figure 2.   Change in end-expiratory lung volume from a PEEP of 8 cm H2O and 15 cm H2O to ZEEP (V) measured by the RIP after disconnecting the endotracheal tube from the ventilator before (dark bars) and 15 min after (gray bars) a RM. Position (supine or prone) and group (S-P or P-S groups) are specified above each figure. Asterisk denotes statistically significant difference as compared with pre-RM conditions; bars represent SD.

Effect of a RM on Gas Exchange

The temporal profiles of Pa_O2 and /T after a RM at each stage are shown in Figures 3 and 4, respectively. Under all conditions there was a transient increase in Pa_O2 with a corresponding decrease in /T. However, in both groups the improvement in oxygenation was sustained after 15 min in prone position only at a PEEP of 8 cm H₂O (p < 0.02), whereas sustained improvement required a PEEP of 15 cm H₂O in the supine position (p < 0.04, Figure 3). In the PS group in prone position at a PEEP of 8 cm H₂O and 15 cm H₂O a RM increased Pa_O2 after 15 min by 62 ± 34 mm Hg (p < 0.02) and 43 ± 74 mm Hg, respectively. In the same group, in the supine position the corresponding increments in Pa_O2 were 28 ± 118 mm Hg and 40 ± 54 mm Hg, respectively. In the SP group, in supine position at a PEEP of 8 cm H₂O and 15 cm H₂O a RM increased Pa_O2 after 15 min by 27 ± 21 mm Hg and 69 ± 42 mm Hg (p < 0.05), respectively. In prone position the corresponding increments in Pa_O2 were 71 ± 52 mm Hg and 14 ± 14 mm Hg, respectively. We could not find a consistent relationship between the changes in V and the changes in Pa_O2 and /T after a RM.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Temporal profile of PaO2 at each experimental stage before and after a RM at the two PEEP levels (8 cm H2O, solid circles and 15 cm H2O, open circles). Position (supine or prone) and group (S-P or P-S groups) are specified above each figure. Asterisk denotes statistically significant difference as compared with pre-RM conditions; bars represent SD.

View larger version (24K): [in this window] [in a new window]   Figure 4.   Temporal profile of -v/ T at each experimental stage before and after a RM at the two PEEP levels (8 cm H2O, solid circles and 15 cm H2O, open circles). Position (supine or prone) and group (S-P or P-S groups) are specified above each figure. Asterisk denotes statistically significant difference as compared with pre-RM conditions; bars represent SD.

He-dilution ZEEP-FRC Measurements

He-dilution end-expiratory lung volume measurements (ZEEP-FRC) were performed after 20 s of apnea at ZEEP in six dogs (three each in PS and SP groups). The main purpose of ZEEP-FRC measurements was to determine whether a RM changed absolute lung volume at ZEEP. Because of the small number of measurements, the data in both groups were pooled for each position irrespective of group. The effect of a RM was tested using a paired t test for each position. ZEEP-FRC in the supine position did not change significantly with a RM (364 ± 205 ml and 389 ± 222 ml before and after a RM, respectively). In the prone position, ZEEP-FRC increased slightly but significantly (p < 0.02) after a RM from 399 ± 232 ml to 444 ± 273 ml.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The main findings of this study are: (1) prone position improved oxygenation, especially if the supine position had preceded the prone position; (2) a RM was more effective at improving oxygenation in the prone position; (3) preservation of the oxygenation benefit of a RM required a lower PEEP in the prone as compared with the supine position (8 cm H₂O versus 15 cm H₂O); and (4) the oxygenation benefit of prone position was carried over to the supine position, i.e., deterioration in Pa_O2 was less when the animals were turned from prone from supine position as compared with the increment in Pa_O2 when the dogs were turned from supine to prone position.

Effect of Position

In experimental ALI models, prone position has been shown to improve oxygenation by homogenizing the ventral-dorsal distribution of ventilation, thereby increasing ventilation to dorsal regions (12, 13, 20, 21). During prone position pulmonary blood flow is preferentially distributed to dorsal (nondependent) lung regions, irrespective of the effects of gravity (20, 22, 23). Consequently, the prone position improved the / relationships of the dorsal regions where lung collapse and / mismatch predominate in the supine position (21). Because of these observations, Chatte and colleagues (24) postulated that relatively less PEEP would be required to reach the targeted oxygenation in prone as compared with supine position. Indeed, in a lung lavage model, Lim and colleagues (15) demonstrated that prone position augmented the effect of PEEP on oxygenation. Our data confirm these findings and extend the benefits of prone positioning to a lung recruitment strategy.

Prone position stabilizes the dorsal lung units by decreasing the pleural pressure surrounding them (12, 20), thereby increasing the transpulmonary pressure and decreasing the compressive forces acting on the dorsal region. This property of prone positioning may facilitate recruitment of collapsed dorsal lung units during the application of a RM at a lower pressure cost. Our results suggest that prone positioning may improve the efficacy of a RM and preserve its oxygenation benefits at a lower PEEP level. In OAI, the oxygenation benefits were sustained at a PEEP of 8 cm H₂O in the prone position, whereas the same response required 15 cm H₂O PEEP in supine dogs. In other words, in our study prone position was equivalent to an additional PEEP of approximately 7 cm H₂O. The elevated Pa_O2 and lack of response to a RM in prone position at a PEEP of 15 cm H₂O suggest that most of the recruitable lung regions were already open at that level of PEEP. This finding suggests that a RM may not offer any physiological benefit during prone positioning at relatively high PEEP levels.

In our ALI model, the increment in oxygenation gained by prone position was relatively conserved even after returning to supine position. This phenomenon was also described in patients with ARDS, where a portion of the oxygenation benefits of prone position was preserved after patients were returned to the supine position (24). These observations suggest that some of the beneficial effects of prone positioning are sustained independent of position and possibly lung volume history.

Recent computed tomography studies in patients with ARDS demonstrated that in the supine position lung collapse occurred not only along the ventral-dorsal axis but also the along the cephalo-caudal axis, with lung collapse predominating in the dorsal and caudal regions near the diaphragm (28). Therefore, an effective recruitment maneuver should preferentially expand the caudal-dorsal lung regions. The alveolar recruitment and stabilizing potential of PEEP, however, is directed globally with a tendency to overdistend already open normally compliant lung units (28, 29). Consequently, rather large levels of PEEP (20 cm H₂O) were required to achieve full recruitment both in human (29) and animal studies (30) in the supine position. Collapse and possibly absorption atelectasis of caudal-dorsal lung units would tend to diminish juxtadiaphragmatic transpulmonary pressures, thus increasing the regional opening pressures and decreasing the recruiting potential of a RM (28). These effects are probably partially abolished in the prone position, boosting the recruitment potential of a RM and allowing the recruited alveoli to stabilize at a lower PEEP level.

Effect of a RM on Lung Volume

In a pig model of OAI, Neumann and colleagues (30, 31) demonstrated that lung collapse was nearly complete after 20 s of apnea at ZEEP. Consequently, it was not surprising that in our study RM had a minimal effect on ZEEP-FRC, as it was measured after 20 s of apnea at ZEEP. We chose to use a 20-s apnea period at ZEEP before application of a RM to standardize the volume history of the lung.

At each experimental stage, V consistently increased when PEEP was increased and after application of a RM. However, the observed increment in V after a RM did not always translate into an oxygenation benefit. Examination of the mechanisms of the oxygenation response to a RM requires determination of regional perfusion/ventilation relationships of the lung. We did not measure the effect of a RM on the distribution of pulmonary perfusion within the lungs. Moreover, as V measures the global changes in lung volume, it is influenced by the expansion of already aerated lung regions and does not fully reflect the extent of alveolar recruitment. In supine patients with ALI, only 18% to 25% of the increment in V induced by increased PEEP resulted from lung recruitment (32, 33). Similarly, in patients with ALI who exhibited a concave PV relationship, lung recruitment (defined as the decrement in nonaerated volume by thoracic chest tomography) accounted for approximately 20% of the increment in the lung volume when PEEP was increased from ZEEP to LIP + 2 cm H₂O (34). Consequently, the major portion of PEEP-induced increase in V may have been caused by expansion of already aerated lung regions. These considerations may explain the lack of a consistent correlation between changes in V and Pa_O2 after a RM in our study.

Limitations of the Current Study

By design, our study was of short duration (15 min after each RM) and did not explore the long-term effects of a RM and prone position. Conceivably, longer application of prone position may allow a less aggressive RM (in terms of pressure and duration of SI) to accomplish similar results in terms of oxygenation.

We measured the PV curve of the respiratory system at the end of the experimental protocol to avoid changing the volume history of the lungs during the experimental protocol and altering the oxygenation response of the respiratory system to a RM. Similar to our previous experiments with OAI in supine dogs (11), we obtained a value of 13 ± 2 cm H₂O for the LIP of the inflation PV curve constructed from ZEEP. Consequently, our chosen PEEP levels (8 and 15 cm H₂O) bracketed the LIP by approximately 1 to 3 cm H₂O on either side and should have favored lung collapse and recruitment, respectively. Nevertheless, defining the exact volume history of the lung under dynamic conditions (during tidal ventilation with a specific PEEP and VT combination) requires measurement of both static and dynamic PV relationships.

In two species (pig and dog), different experimental models of ALI displayed variable lung collapse and recruitment patterns. After a RM, the rise in Pa_O2 (11) and the increment in lung volume (30, 31) were lost faster than the other models of ALI indicating that OAI had a tendency to be more unstable than the other models. Consequently, our choice of an OAI model may have minimized the potential beneficial effects of a RM. The results of our study should be interpreted with caution as it is primarily applicable to the setting of experimental OAI in dogs ventilated with relatively small VT (8 ml/kg) and moderate levels of PEEP and should not be extrapolated directly to the ARDS/ALI setting.

Comparison with Previous Studies

In a previous study in supine dogs (11), we showed that OAI was not responsive to a RM at a PEEP of 10 or 20 cm H₂O despite a more aggressive RM consisting of three SI of 30-s duration each that pressurized the airway opening sequentially to 40, 60, and 60 cm H₂O, respectively. In contrast, in the current study in the supine position at a PEEP of 15 cm H₂O, we observed an oxygenation response after a single SI using 60 cm H₂O pressure for 30 s. The major difference between our two studies was that we employed a much smaller VT in the current study (8 versus 15 ml/kg), possibly enhancing alveolar collapse during tidal ventilation. Consequently, in the current study the dogs may have had a larger volume of collapsed alveoli that could be potentially recruited by a RM. This explanation is supported by experimental studies that demonstrated that SI recruited lung volume during high-frequency ventilation but failed to do so during conventional mechanical ventilation with a large VT (8, 10). Alternatively, in our current study we may have stabilized the lung units recruited by a RM with PEEP, as we also used a higher level of PEEP (15 cm H₂O) than our previous study (10 cm H₂O) (11).

Using the parenchymal marker technique in dogs with OAI, Martynowicz and colleagues (35) demonstrated that dependent lung regions were not collapsed but remained expanded, owing to the presence of blood and/or edema fluid within the alveoli. As a result, in the absence of PEEP (ZEEP conditions) most of the VT was delivered to the nondependent lung regions, and PEEP homogenized the distribution of tidal ventilation (35). Our previous results (11) are consistent with these findings. Both studies (11, 35) used large VT (20 ml/kg and 25 ml/kg, respectively), most likely preventing the collapse of unstable lung units. In contrast, in the current study we purposefully employed a small VT (8 ml/kg) to promote lung collapse after OAI. The improvement in oxygenation in the current study, in contrast to the lack of improvement in our previous study (11) after a RM, suggests that lung collapse occurred when the dogs were ventilated with a low VT after OAI.

SI, large VT, sighs, and prone position have all been advocated as ventilatory techniques that can be used as methods of recruitment (2, 5, 6, 8). To our knowledge, there are no studies that have directly compared different methods of lung recruitment. Moreover, once recruited, the minimal level of PEEP required for a given VT and ventilatory pattern to maintain the long-term patency of the recruited lung regions remains unknown. Conceivably, considerably less pressure for a shorter duration may be required during a RM in the prone position compared with the supine position. Because the clinical role of various RM in the ventilator management of patients with ARDS remains to be elucidated, future studies investigating the effects of different recruitment maneuvers should examine them in conditions including the prone position.

In summary, in an OAI model of ALI in dogs, a SI of 60 cm H₂O applied for 30 s was more effective in improving oxygenation in prone as compared with the supine position. Moreover, less PEEP was required in the prone position to preserve the oxygenation benefits of a RM.