Bedside Selection of Positive End-Expiratory Pressure in Mild, Moderate, and Severe Acute Respiratory Distress Syndrome* Davide Chiumello, MD1,2; Massimo Cressoni, MD2; Eleonora Carlesso, MSc2; Maria L. Caspani, MD1; Antonella Marino, MD2; Elisabetta Gallazzi, MD2; Pietro Caironi, MD1,2; Marco Lazzerini, MD3; Onnen Moerer, MD4; Michael Quintel, MD4; Luciano Gattinoni, MD, FRCP1,2
Objective: Positive end-expiratory pressure exerts its effects keeping open at end-expiration previously collapsed areas of the lung; consequently, higher positive end-expiratory pressure should be limited to patients with high recruitability. We aimed to determine which bedside method would provide positive end-expiratory pressure better related to lung recruitability. Design: Prospective study performed between 2008 and 2011. Setting: Two university hospitals (Italy and Germany). Patients: Fifty-one patients with acute respiratory distress syndrome. *See also p. 448. 1 Dipartimento di Anestesia, Rianimazione (Intensiva e Subintensiva) e Terapia del Dolore, Fondazione IRCCS Ca’ Granda–Ospedale Maggiore Policlinico, Milan, Italy. 2 Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Università degli Studi di Milano, Milan, Italy. 3 Dipartimento di Radiologia, Fondazione IRCCS Ca’ Granda–Ospedale Maggiore Policlinico, Milan, Italy. 4 Department of Anaesthesiology, Emergency and Intensive Care Medicine, Georg-August University of Göttingen, Göttingen, Germany. http://www.clinicaltrial.gov number: NCT00682942. Gattinoni, Chiumello, Moerer, and Quintel contributed to conception and design. Caspani, Chiumello, Marino, Gallazzi, Lazzerini, Moerer, and Quintel contributed to acquisition of data. Gattinoni, Cressoni, Carlesso, Caspani, Marino, Gallazzi, Lazzerini, and Caironi contributed to analysis and interpretation of data. Gattinoni, Cressoni, Carlesso, and Chiumello contributed to drafting of the article. Caironi, Caspani, Marino, Gallazzi, Lazzerini, Moerer, and Quintel contributed to critical revision of the article for important intellectual content. Cressoni and Carlesso contributed to statistical analysis. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/ccmjournal). Supported, in part, by institutional funding. Dr. Gattinoni received an Italian grant (PR-0062) from Fondazione Fiera di Milano for Translational and Competitive Research (2007). The remaining authors have disclosed that they do not have any potential conflicts of interest. Address requests for reprints to: Luciano Gattinoni, MD, FRCP, Dipartimento di Anestesia, Rianimazione (Intensiva e Subintensiva) e Terapia del Dolore, Fondazione IRCCS Ca’ Granda–Ospedale Maggiore Policlinico, via Francesco Sforza 35, 20122 Milan, Italy. E-mail:
[email protected] Copyright © 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/CCM.0b013e3182a6384f
252
www.ccmjournal.org
Interventions: Whole lung CT scans were taken in static conditions at 5 and 45 cm H2O during an end-expiratory/end-inspiratory pause to measure lung recruitability. To select individual positive end-expiratory pressure, we applied bedside methods based on lung mechanics (ExPress, stress index), esophageal pressure, and oxygenation (higher positive end-expiratory pressure table of lung open ventilation study). Measurements and Main Results: Patients were classified in mild, moderate and severe acute respiratory distress syndrome. Positive end-expiratory pressure levels selected by the ExPress, stress index, and absolute esophageal pressures methods were unrelated with lung recruitability, whereas positive end-expiratory pressure levels selected by the lung open ventilation method showed a weak relationship with lung recruitability (r2 = 0.29; p < 0.0001). When patients were classified according to the acute respiratory distress syndrome Berlin definition, the lung open ventilation method was the only one which gave lower positive end-expiratory pressure levels in mild and moderate acute respiratory distress syndrome compared with severe acute respiratory distress syndrome (8 ± 2 and 11 ± 3 cm H2O vs 15 ± 3 cm H2O; p < 0.05), whereas ExPress, stress index, and esophageal pressure methods gave similar positive end-expiratory pressure values in mild, moderate, and severe acute respiratory distress syndrome. The positive end-expiratory pressure selected by the different methods were unrelated to each other with the exception of the two methods based on lung mechanics (ExPress and stress index). Conclusions: Bedside positive end-expiratory pressure selection methods based on lung mechanics or absolute esophageal pressures provide positive end-expiratory pressure levels unrelated to lung recruitability and similar in mild, moderate, and severe acute respiratory distress syndrome, whereas the oxygenation-based method provided positive end-expiratory pressure levels related with lung recruitability progressively increasing from mild to moderate and severe acute respiratory distress syndrome. (Crit Care Med 2014; 42:252–264) Key Words: acute respiratory distress syndrome; lung; lung collapse; positive end-expiratory pressure; positive-pressure respiration; respiratory mechanics
February 2014 • Volume 42 • Number 2
Feature Articles
A
cute respiratory distress syndrome (ARDS) is characterized by dishomogeneity of the lung parenchyma and presents not inflated regions. These may be consolidated and not openable or collapsed and recruitable to aeration during inspiration; the recruitable regions may vary from 0% to 50% of the lung parenchyma (1, 2). It is largely accepted that the positive end-expiratory pressure (PEEP) to be ideally applied to a given patient should be the PEEP that best compromises between its beneficial putative effects (keeping open the recruited lung) and its negative effects (overdistention of already open lung) (3, 4). Over the decades, this problem has been largely investigated and different methods for selecting this ideal PEEP have been proposed based on respiratory system compliance (5, 6), oxygenation and shunt values (7–9), and volume-pressure curve either during inflation or during deflation (10–12). Large randomized trials comparing higher and lower levels of PEEP, in which the higher PEEP was selected according to oxygenation values (Assement of Low Tidal Volume and elevated End-expiratory volume to Obviate Lung Injury [7] and lung open ventilation [LOV] study [8]) or lung mechanics (ExPress trial [6]), failed to show significant advantages of higher PEEP. The main concern related to these studies was that the lung recruitability was ignored. As it has been shown that recruitability is directly associated with the amount of lung edema (i.e., with the ARDS severity) (2), it was postulated that higher PEEP in patients with lower recruitability could enhance its negative effects (overdistention), whereas lower PEEP in patients with higher recruitability cannot exert its beneficial effects (4). Accordingly, lower PEEP should be reserved to patients with lower recruitability like a mild ARDS and higher PEEP to patients with greater recruitability like a severe ARDS. Actually, two meta-analyses suggested outcome benefit of higher PEEP in more severe ARDS and possible harm in patients with mild ARDS (3, 13). Therefore, we examined 51 patients with ARDS of different severity to assess which bedside method selected the PEEP better related to lung recruitability and ARDS severity. Two methods were based on lung mechanics (ExPress study [6] and stress index [5]), one on esophageal pressure (14) and one on oxygenation (LOV study [8]).
MATERIALS AND METHODS See supplemental data (Supplemental Digital Content 1, http://links.lww.com/CCM/A730) for details. Patients Fifty-one patients with ARDS (i.e., Pao2/Fio2 < 300) (15) were examined between May 2008 and June 2011 in two University Hospitals (Fondazione IRCCS Ca’ Granda–Ospedale Maggiore Policlinico, Milan, Italy and Georg-August University of Göttingen, Germany). The study was approved by the institutional review board of each hospital and written consent was obtained according to the regulations applicable in each institution (consent was delayed in Italy until the patients had recovered from the sedation and obtained from a legal representative in Germany). The study protocol is summarized in Figure 1. Critical Care Medicine
Classification According to the Berlin Definition Patients were retrospectively classified in mild, moderate, and severe ARDS (16, 17) according to the Pao2/Fio2 ratio obtained at a standard PEEP level (PEEP 5 cm H2O): patients with a ratio 200 < Pao2/Fio2 ≤ 300 were classified as having “mild” ARDS, patients with a ratio 100 < Pao2/Fio2 ≤ 200 were classified as having “moderate” ARDS, while patients with a Pao2/Fio2 ratio less than or equal to 100 were classified as “severe” ARDS. Patient Management Patients were deeply sedated and curarized. The anesthesia was maintained with infusion of midazolam (0.05–0.1 mg/ kg) and fentanyl (2–3 μg/kg) and paralysis with vecuronium (0.05–0.1 mg/kg). A recruitment maneuver was performed in pressure-control mode with PEEP 5 cm H2O, pressure above PEEP 40 cm H2O, respiratory rate 10 breaths/min, I/E ratio 1:1, and Fio2 0.7 for 2 minutes. Thereafter, patients were ventilated with a tidal volume of 6–8 mL/kg of ideal body weight (18). A PEEP level of 10 cm H2O was arbitrarily set at the beginning for all patients during the stabilization period (60 min). Bedside PEEP Selection Tests While keeping constant tidal volume (6–8 mL/kgIBW), respiratory rate, and I:E ratio (1:2), five PEEP levels were randomly tested for 20 minutes each. The total length of bedside protocol was approximately 160 minutes. PEEP Setting According to the Increased Recruitment Strategy of the ExPress Study. The PEEP was progressively increased until the airway plateau pressure reached 28–30 cm H2O at constant tidal volume of 6 mL/kgIBW (6). This was the selected PEEP. If the inspiratory airway plateau pressure did not reach 28 cm H2O despite a PEEP of 20 cm H2O, this level of PEEP was chosen as “selected PEEP” and no further increase was performed. PEEP Setting According to the Stress Index. Stress index method defines the slope of the time-pressure curve of the respiratory system with a dimensionless coefficient (5). A value lower than 1 indicates continuous recruitment, whereas a value greater than 1 indicates hyperinflation. The PEEP selected with this method could be the one at which the time-pressure curve becomes linear (minimal PEEP) or at which the time-pressure curve loses its linearity (maximal PEEP). In this study, in the framework of open lung strategy, we chose the “maximal PEEP” approach. PEEP Setting According to Esophageal Pressure. The esophageal pressure was measured at functional residual capacity during a no flow expiratory pause; the results of at least four measurements were averaged. The rational of esophageal pressure method is to provide at end-expiration a positive transpulmonary pressure to prevent lung collapse. Therefore, assuming that pleural pressure is equal to the absolute esophageal pressure, PEEP was set equal to the absolute value of esophageal pressure. This setting is based on the same conceptual framework proposed by Talmor et al (14). PEEP Setting According to the LOV Study. PEEP was selected according to the Lung Open Ventilation Strategy arm of the www.ccmjournal.org
253
Chiumello et al
PEEP 5 cm H2O. PEEP was set at 5 cm H2O to collect all the physiological variables at the same pressure values at which CT images were taken to allow a more consistent analysis of structure/function relationship.
Figure 1. Study protocol: flowchart. Fifty-one patients were enrolled. After a recruitment maneuver, patients underwent a stabilization period of 20 min at positive end-expiratory pressure (PEEP) 10 cm H2O. Five different PEEP selection methods were applied in random order: 1) PEEP 5 cm H2O; 2) according to a PEEP/Fio2 table as in the Lung Open Ventilation (LOV) Strategy arm of the LOV study; 3) to maintain the inspiratory plateau pressure between 28 and 30 cm H2O, according to the “increased recruitment strategy” arm of the ExPress protocol (ExPress); 4) to obtain a stress index coefficient ~1; 5) PEEP was set equal to absolute esophageal pressure at end-expiration. Patients were then transferred to the CT scan facility where, after a recruitment maneuver, whole lung CT scans were taken at 5 and 45 cm H2O during an end-expiratory and an end-inspiratory pause, respectively. Figure shows sample CT images of an higher (right side) and a lower recruiter (left side) at nearly the same lung level. ALI = acute lung injury, ARDS = acute respiratory distress syndrome, I/E = inspiratory/expiratory ratio.
LOV study (8). PEEP was selected according to a predefined PEEP/Fio2 table, targeting to arterial oxygen saturation between 88-93%; since the protocol changed during the study, we chose the PEEP/Fio2 table used after the protocol change (reported below) (Table 1).
Positive End-Expiratory Pressure/ Fio2 Table Table 1.
Positive End-Expiratory Pressure (cm H2O)
Fio2
0.3
5–10
0.4
10–18
0.5
18–20
0.6
20
0.7
20
0.8
20–22
0.9
22
1 254
22–24 www.ccmjournal.org
CT Scan Acquisition Patients recruited in Germany underwent CT scan acquisition with the GE Medical Systems LightSpeed VCT (Schenectady, NY), while patients enrolled in Italy with Siemens Medical Solution Somaton Definition Flash Syngo CT 2011 (Munich, Germany). At the end of the recruitment maneuver, the patient was ventilated for 5 minutes at PEEP 5 cm H2O and a whole lung CT scan was performed during an endexpiratory pause (ranging from 15 to 25 s). A second whole lung CT scan was taken at inspiratory plateau pressure of 45 cm H2O.
CT Scan Analysis The outline of the lungs was manually drawn in each CT section excluding the hilar vessels and the main bronchi. Segmented images were analyzed with a custom dedicated software (Soft-E-Film, http://www. elekton.it, Milan, Italy). Assuming that lung parenchyma has a density equal to the water and that the lung is composed of two compartments (air with a CT number of –1,000 and lung parenchyma with a CT number of 0), it is possible to compute in each voxel the relative proportion of air and tissue (19): CT / −1000 = Vgas / (Vgas + Vtissue )
where Vgas is gas volume (mL) and Vtissue is the tissue volume (g or mL). As the gas fraction equals CT/–1,000, the tissue fraction will be 1 – CT/–1,000. Whole Lung Recruitability Recruitability was measured according to the following equation: Recruitabilty = (not aerated lung tissue peep 5 cm H2 O – not aeraed tissue PEEP 45cm H2 O) / (total lung tissue PEEP 5cm H2O Lung recruitability should represent the maximal amount lung parenchyma which is possible to recruit. In our definition, we assume that full recruitment (100%) occurs at 45 cm H2O end-inspiratory pressure, which corresponded in this February 2014 • Volume 42 • Number 2
Feature Articles
Table 2.
Patients’ Main Characteristics According to Berlin Classification
Physiological/CT Scan Variables
Pao2/Fio2
Mild ARDS (n = 7)
234 ± 18a
Moderate ARDS (n = 33)
148 ± 26b
Severe ARDS (n = 11)
72 ± 12
p
< 0.0001
Sex (males/females)
4/3
25/8
9/2
0.51
Days elapsed before study
3 ± 3
6 ± 8
4 ± 4
0.31
ICU survival, n (%)
6 (86)
21 (64)
3 (27)
0.04
Lung recruitability (%)
12 ± 9
14 ± 10b
22 ± 11
0.03
Pao2 (mm Hg)
85 ± 12a
68 ± 11b
57 ± 7
< 0.0001
Fio2 (%)
b
36 ± 5
47 ± 11
80 ± 15
< 0.0001
Paco2 (mm Hg)
42 ± 4
b
44 ± 6
54 ± 7
< 0.0001
0.56 ± 0.10
0.65 ± 0.22
0.04
Dead space fraction
b
b
0.48 ± 0.11
b
Minute ventilation (L/min)
9.0 ± 1.4
9.2 ± 1.7
9.8 ± 2.4
0.64
Svo2 (%)
81 ± 3
70 ± 7
71 ± 5
0.001
Respiratory system elastance (cm H2O/L)
29.7 ± 9.0
25.6 ± 8.6
27.9 ± 10.5
0.48
Lung elastance (cm H2O/L)
22.9 ± 7.5
19.4 ± 8.4
21.1 ± 8.1
0.57
6.8 ± 5.2
0.99
Chest wall elastance (cm H2O/L) Total lung weight (g)
a
6.8 ± 2.3 1,252 ± 207
6.7 ± 4.4 b
1,380 ± 388
b
1,947 ± 607
< 0.001
Not inflated tissue (%)
38 ± 19
42 ± 15
47 ± 16
0.46
Poorly inflated tissue (%)
28 ± 10
28 ± 11
33 ± 11
0.31
Well inflated tissue (%)
34 ± 14
30 ± 13
19 ± 13
0.03
Over inflated tissue (%)
0 ± 0
0 ± 1
0 ± 0
0.10
Over inflated tissue positive endexpiratory pressure 45 cm H2O (%)
3 ± 5
4 ± 4
2 ± 3
0.17
ARDS = acute respiratory distress syndrome. a p < 0.05 versus moderate and p < 0.05 versus severe ARDS. b p < 0.05 versus severe ARDS. The table summarizes the main physiological and CT scan data standardized at positive end-expiratory pressure 5 cm H2O. Plus-minus values are means ± sd. Because of rounding, percentages may not total 100. p-values were obtained with one-way analysis of variance. Multiple comparisons were performed with the Bonferroni correction. Chi-square test or Fisher exact test was used for qualitative variables. Elapsed days were counted from ICU admission to the CT scan acquisition. Physiological dead space was available in 47 patients (31 moderate ARDS and 9 severe ARDS). Lung and chest wall elastance were available in 50 patients (32 in moderate ARDS).
population to a median transpulmonary pressure of 34.6 cm H2O (interquartile range, 33.1–36.5 cm H2O) well in the range of total lung capacity (20). Higher recruitment pressure (60 cm H2O) were avoided as the recruitability between 45 and 60 cm H2O is negligible at cost of possible barotrauma, severe acidosis, and volume load for hemodynamic support (21, 22). Data Collection Twenty minutes after the application of the PEEP, selected blood gases, lung mechanics, and hemodynamic variables were measured. Statistical Methods The relationship between the PEEP selected by each of the bedside method and lung recruitability was assessed with linear regression. In Table 2, one-way analysis of variance was used for normally distributed variables and Kruskal-Wallis test for variables that did not appear normally distributed on graphic inspection and multiple comparisons were performed with the Critical Care Medicine
Bonferroni correction; the chi-square test or Fisher exact test was used for qualitative variables. Tables 3 and 4 compare the influence of PEEP selection methods and of the Berlin classification on physiological variables. The test was performed with mixed methods and multiple comparisons were performed with the Bonferroni correction. The statistical analysis was performed with SAS 9.1 (SAS Institute, Cary, NC) and the R software (R Development Core Team, 2010; R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/).
RESULTS Main Physiologic and CT Scan Characteristics in Mild, Moderate, and Severe ARDS In Table 2, we summarized the main clinical characteristics of the patient population. Seven patients presented with mild ARDS, 33 with moderate ARDS, and 11 with severe ARDS. As www.ccmjournal.org
255
Chiumello et al
Table 3. Physiologic Response to Positive End-Expiratory Pressure Application in Mild, Moderate, and Severe Acute Respiratory Distress Syndrome Mild ARDS (n = 7)
Respiratory Mechanics Variables
Moderate ARDS (n = 33)
Severe ARDS (n = 11)
Positive end-expiratory pressure (cm H2O) 8 ± 2a
11 ± 3a
15 ± 3
ExPress
14 ± 2
14 ± 3
16 ± 3
Stress index
11 ± 2
14 ± 3
14 ± 3
Esophageal pressure
13 ± 3
12 ± 4
13 ± 4
LOV
22 ± 4a
23 ± 5a
29 ± 5
ExPress
28 ± 2
b
27 ± 2
28 ± 2
Stress index
25 ± 3
26 ± 3b
26 ± 3
Esophageal pressure
27 ± 5b
25 ± 5
25 ± 5
7.3 ± 2.3
8.5 ± 3.1
11.3 ± 4.8
10.8 ± 1.6b
10.5 ± 3.7b,c
11.3 ± 3.8
9.8 ± 1.3
10.9 ± 4.2b,c
10.7 ± 3.4
10.8 ± 3.1
9.5 ± 5.1
9.6 ± 4.6
LOV
17.6 ± 3.9
17.9 ± 5.1
22.6 ± 6.2
ExPress
21.5 ± 3.0
Stress index Esophageal pressure
LOV
p (method)
b
b b
b
< 0.001
p (classification according to Berlin definition)
0.02
p (interaction)
0.01
Plateau pressure (cm H2O) b
p (method)
0.0001
p (classification according to Berlin definition)
0.19
p (interaction)
0.002
End-expiratory transpulmonary pressure (cm H2O) LOV ExPress Stress index Esophageal pressure
b
p (method)
0.001
p (classification according to Berlin definition)
0.68
p (interaction)
0.06
Plateau transpulmonary pressure (cm H2O)
p (method)
0.004
p (classification according to Berlin definition)
0.48
p (interaction)
0.03
20.5 ± 4.0
21.7 ± 5.4
20.2 ± 2.5
20.4 ± 4.6
20.4 ± 3.6
21.0 ± 4.6
18.9 ± 6.4
20.1 ± 6.9
36.2 ± 8.9
44.4 ± 11.4
39.9 ± 14.0
b
b,c
b
Respiratory system compliance (mL/cm H2O) LOV
(Continued) 256
www.ccmjournal.org
February 2014 • Volume 42 • Number 2
Feature Articles
Table 3. (Continued). Physiologic Response to Positive End-Expiratory Pressure Application in Mild, Moderate, and Severe Acute Respiratory Distress Syndrome Respiratory Mechanics Variables
Mild ARDS (n = 7)
Moderate ARDS (n = 33)
ExPress
32.9 ± 8.2
41.0 ± 9.5b
44.0 ± 15.0
Stress index
34.5 ± 10.3
43.0 ± 10.1
49.3 ± 24.4b
Esophageal pressure
33.8 ± 10.7
41.8 ± 10.6
42.3 ± 14.5
LOV
46.7 ± 13.1
63.3 ± 25.7
50.7 ± 20.3
ExPress
45.5 ± 14.1
65.2 ± 43.8
58.6 ± 25.0
Stress index
43.8 ± 15.2
62.4 ± 26.1
60.1 ± 24.3
Esophageal pressure
45.9 ± 17.2
60.1 ± 23.4
56.4 ± 25.2
p (method)
0.23
p (classification according to Berlin definition)
0.16
p (interaction)
0.01
Severe ARDS (n = 11)
Lung compliance (mL/cm H2O)
p (method)
1.00
p (classification according to Berlin definition)
0.84
p (interaction)
1.00
ARDS = acute respiratory distress syndrome, LOV = lung open ventilation. Plus-minus values are means ± sd. p-values were obtained by mixed-method analysis for positive end-expiratory pressure (PEEP) selection method (p [method]), classification according to the Berlin definition (p [classification according to Berlin definition]), and their interaction (p [interaction]). We compared the physiological values obtained with each bedside PEEP selection method among the three classes according to the Berlin definition. a p < 0.05 vs severe ARDS) and within each PEEP selection method. b p < 0.05 vs LOV selection method. c p < 0.05 vs esophageal pressure selection method). Oxygen extraction ratio was calculated as (Cao2 – Cvo2)/Cao2 (see supplemental data, Supplemental Digital Content, http://links.lww.com/CCM/A730). Multiple comparisons were performed with the Bonferroni t test. Step based on the absolute esophageal pressure was available in 48 patients (30 moderate ARDS).
shown, physiological variables, as well as CT-derived variables, worsened from mild to severe ARDS, in particular oxygenation, dead space (48% ± 11%, 56% ± 10%, and 65% ± 22%; p = 0.04), total lung weight (1,252 ± 207 g, 1,380 ± 388 g, and 1,947 ± 607 g; p < 0.001), and well inflated tissue (34% ± 14%, 30% ± 13%, and 19% ± 13%; p = 0.03); in contrast, the lung recruitability increased from mild to severe ARDS (12% ± 9%, 14% ± 10%, and 22% ± 11%; p = 0.03). Bedside PEEP Selection Methods in Mild, Moderate, and Severe ARDS The PEEP selected with the different methods are reported in Table 3; as shown, the oxygenation-based method (LOV) provided significantly lower PEEP in mild ARDS (8 ± 2 cm H2O) compared with the severe ARDS (15 ± 3 cm H2O; p < 0.05); accordingly, plateau pressure and end-expiratory transpulmonary pressure increased from mild to severe ARDS. In contrast, the methods based on lung mechanics and esophageal pressure provided similar high PEEP, plateau pressure, and end-expiratory transpulmonary pressure in mild, moderate, and severe ARDS. At the selected PEEP, all the methods provided similar Pao2/Fio2 ratio (Table 4). Critical Care Medicine
Relationship Between Bedside Selected PEEP and Lung Recruitability In Figure 2, we reported the relationship between the PEEP (upper panel) and end-expiratory transpulmonary pressure (lower panel) and the lung recruitability (Figs. E1–E4, Supplemental Digital Content 1, http://links.lww.com/CCM/A730). As shown in Figure E3 (Supplemental Digital Content 1, http:// links.lww.com/CCM/A730), the relationship was significant only for the oxygenation-based method. In Figure 3, we report the PEEP obtained with the different methods as a function of quartiles of hyperinflation; as shown, only the PEEP selected by the ExPress method was weakly but significantly associated with hyperinflation (r2 = 0.09, p = 0.04). Relationship Between Bedside PEEP Selection Methods The PEEP selected by the different methods were unrelated to each other with the exception of the two methods based on lung mechanics (ExPress and stress index, stress index-derived PEEP [cm H2O] = 1.98 + 0.81 × ExPress protocol-derived PEEP, r2 = 0.51, p < 0.0001; Figs. E5–E10, Supplemental Digital Content 1, http://links.lww.com/CCM/A730). It is worth www.ccmjournal.org
257
Chiumello et al
Table 4. Physiologic Gas-Exchange Response to Positive End-Expiratory Pressure Application in Mild, Moderate, and Severe Acute Respiratory Distress Syndrome Mild ARDS (n = 7)
Gas Exchange and Hemodynamics Variables
Moderate ARDS (n = 33)
Severe ARDS (n = 11)
Fio2 (%) LOV
30 ± 0a,b
37 ± 6b
56 ± 19a
ExPress
36 ± 5a,b,c
47 ± 11b,c
71 ± 25a,c
Stress index
36 ± 5a,b,c
47 ± 11b,c
74 ± 23a,c
Esophageal pressure
36 ± 5a,b,c
48 ± 11b,c
66 ± 26a,c
LOV
70 ± 4
66 ± 9
67 ± 12
ExPress
p (method)
< 0.001
p (classification according to Berlin definition)
< 0.001
p (interaction)
0.38
Pao2 (mm Hg) 91 ± 16
c
80 ± 22
83 ± 24c
Stress index
88 ± 15c
87 ± 44c
78 ± 25c
Esophageal pressure
88 ± 15c
87 ± 39c
87 ± 32c
p (method)
c
< 0.001
p (classification according to Berlin definition)
0.24
p (interaction)
0.80
Pao2/Fio2 LOV
234 ± 13a,b
183 ± 33b
129 ± 29
ExPress
253 ± 45a,b
174 ± 43b
126 ± 41
Stress index
244 ± 42a,b
186 ± 56b
123 ± 34
Esophageal pressure
241 ± 34
179 ± 52
132 ± 49
LOV
42 ± 3b
44 ± 6b
53 ± 9
ExPress
42 ± 4
45 ± 6
54 ± 9
Stress index
42 ± 3
45 ± 6
53 ± 9
Esophageal pressure
42 ± 4
45 ± 6
54 ± 9
76.6 ± 3.4a
68.9 ± 9.7
72.6 ± 6.9
p (method) p (classification according to Berlin definition) p (interaction)
a,b
b
0.73 < 0.0001 0.51
Paco2 (mm Hg)
p (method) p (classification according to Berlin definition) p (interaction)
b
b b
b b
b
0.06 < 0.001 0.76
Svo2 (%) LOV
(Continued) 258
www.ccmjournal.org
February 2014 • Volume 42 • Number 2
Feature Articles
Table 4. (Continued). Physiologic Gas-Exchange Response to Positive End-Expiratory Pressure Application in Mild, Moderate, and Severe Acute Respiratory Distress Syndrome Gas Exchange and Hemodynamics Variables
Mild ARDS (n = 7)
Moderate ARDS (n = 33)
Severe ARDS (n = 11)
ExPress
79.1 ± 3.4a
71.7 ± 8.0c
74.0 ± 6.4
Stress index
78.9 ± 4.0
71.3 ± 7.6
76.5 ± 5.8c
Esophageal pressure
79.9 ± 4.4a
70.6 ± 9.9c
73.3 ± 7.4
LOV
92.3 ± 1.5
91.8 ± 3.2
91.9 ± 3.2
ExPress
95.1 ± 1.7
c
94.2 ± 2.7
94.2 ± 3.5c
Stress index
94.8 ± 1.4c
94.3 ± 3.0c
94.5 ± 3.5c
Esophageal pressure
94.9 ± 1.2c
93.9 ± 2.9c
93.3 ± 3.8
LOV
37 ± 3
36 ± 20
41 ± 11
ExPress
31 ± 5
31 ± 9
40 ± 12c
Stress index
33 ± 6c
30 ± 10c
43 ± 14c
Esophageal pressure
33 ± 7c
31 ± 10c
40 ± 11c
LOV
77 ± 4
80 ± 11
75 ± 11
ExPress
77 ± 4
90 ± 22
79 ± 11
Stress index
77 ± 4
79 ± 12
75 ± 9
Esophageal pressure
79 ± 7
79 ± 12
75 ± 12
LOV
89 ± 22
89 ± 19
84 ± 23
ExPress
90 ± 22
90 ± 17
87 ± 23
Stress index
83 ± 18
87 ± 17
86 ± 24
p (method)
a
c
< 0.001
p (classification according to Berlin definition)
0.04
p (interaction)
0.61
Sao2 (%)
p (method)
c
< 0.0001
p (classification according to Berlin definition)
0.76
p (interaction)
0.90
Shunt (%)
p (method)
c
c
< 0.001
p (classification according to Berlin definition)
0.07
p (interaction)
0.15
Mean arterial pressure (mm Hg)
p (method)
0.94
p (classification according to Berlin definition)
0.69
p (interaction)
0.90
Heart rate (beats/min)
(Continued)
Critical Care Medicine
www.ccmjournal.org
259
Chiumello et al
Table 4. (Continued). Physiologic Gas-Exchange Response to Positive End-Expiratory Pressure Application in Mild, Moderate, and Severe Acute Respiratory Distress Syndrome Gas Exchange and Hemodynamics Variables
Mild ARDS (n = 7)
Esophageal pressure
86 ± 24
86 ± 17
86 ± 23
LOV
10 ± 3
13 ± 4
13 ± 3
ExPress
11 ± 3
13 ± 4
13 ± 3
Stress index
11 ± 3
13 ± 4
13 ± 3
Esophageal pressure
11 ± 3
13 ± 4
12 ± 3
p (method)
0.09
p (classification according to Berlin definition)
0.84
p (interaction)
0.58
Moderate ARDS (n = 33)
Severe ARDS (n = 11)
Central venous pressure (mm Hg) c
c,d
c c
p (method)
0.001
p (classification according to Berlin definition)
0.28
p (interaction)
0.009
Oxygen extraction ratio LOV
0.17 ± 0.03a
0.25 ± 0.10
0.21 ± 0.06
ExPress
0.17 ± 0.02
0.24 ± 0.08
0.21 ± 0.06
Stress index
0.17 ± 0.04
0.24 ± 0.08
0.19 ± 0.06
Esophageal pressure
0.16 ± 0.04
0.25 ± 0.10
0.22 ± 0.06
a a
a
p (method)
0.63
p (classification according to Berlin definition)
0.04
p (interaction)
0.72
ARDS = acute respiratory distress syndrome, LOV = lung open ventilation. Plus-minus values are means ± sd. p-values were obtained by mixed-method analysis for positive end-expiratory pressure (PEEP) selection method (p [method]), classification according to the Berlin definition (p [classification according to Berlin definition]), and their interaction (p [interaction]). We compared the physiological values obtained with each bedside PEEP selection method among the three classes according to the Berlin definition. a p < 0.05 vs moderate ARDS. b p < 0.05 vs severe ARDS) and within each PEEP selection method. c p < 0.05 vs LOV selection method. d p < 0.05 vs esophageal pressure selection method). Oxygen extraction ratio was calculated as (Cao2 – Cvo2)/Cao2 (see supplemental data, Supplemental Digital Content, http://links.lww.com/CCM/A730). Multiple comparisons were performed with the Bonferroni t test. Step based on the absolute esophageal pressure was available in 48 patients (30 moderate ARDS).
observing, however, that with these two methods the PEEP selected tended to decrease instead of increasing with ARDS severity and recruitability.
DISCUSSION In this study, within the methods proposed over the decades to clinically select the PEEP, we choose the ones which we believe representative of different conceptual approaches: the LOV (8), which is the evolution of ARDS Network (18) and the ALVEOLI (7) approaches, was chosen as representative of oxygenation-based methods. The ExPress and the stress index methods, which are based on the Volume-Pressure curve of 260
www.ccmjournal.org
the respiratory system, were chosen as representative of the lung mechanics-based methods, conceptually similar to the ones proposed by Suter et al (9), Lemaire and coworkers (23), Lachmann (24), and Hickling (25). The esophageal pressure method, although not tested in large clinical trials, was chosen as representative of the transpulmonary pressure approach, never attempted before the study of Talmor et al (14). Due to the time required for the whole study, other interesting methods, as the decremental PEEP (21, 26), were not tested. The main finding of this study is that, within the methods we selected, only the LOV method, based on oxygenation and developed on “expert-opinion basis” (Fio2/PEEP scale), provided PEEP levels that increased from mild to severe ARDS, as defined in Berlin, February 2014 • Volume 42 • Number 2
Feature Articles
Figure 2. Upper Panel: Relationship between the average positive end-expiratory pressure (PEEP) and end-expiratory transpulmonary pressure levels selected with the four bedside methods and lung recruitment. Upper Panel summarizes the behavior of the PEEP levels selected with the four bedside methods as a function of lung recruitability. For the sake of clarity, we report lung recruitability for each PEEP selection method divided in quartiles on the x-axis (the average value within each quartile is shown) and on the y-axis the corresponding average PEEP selected.
Critical Care Medicine
www.ccmjournal.org
261
Chiumello et al
the time-pressure curve loses its linearity (stress index [5]) or 30 cm H2O plateau pressure (ExPress [6]) are reached. These phenomena occur in the region of total lung capacity where the lung parenchyma is fully distended. Therefore, with both methods, greater is the respiratory system compliance, higher the PEEP will be (Figs. E11 and E12, Supplemental Digital Content 1, http://links. lww.com/CCM/A730). In fact, in healthy lung, with a tidal volume of 6 mL/kgIBW, 30 cm H2O of plateau pressure will never be reached, and, by protocol, with the Express method, the PEEP would be always set at 20 cm H2O. By contrast, in patients with more severe ARDS, lower is the respiratory system compliance, lower is the Figure 3. Relationship between the average positive end-expiratory pressure (PEEP) selected with the four PEEP required to reach 30 cm bedside methods and overinflation. Figure 3 summarizes the behavior of the PEEP levels selected with the four H2O plateau pressure. In fact, bedside methods as a function of the fraction of overinflated tissue (CT < –900) measured at 45 cm H2O endinspiratory pressure. For the sake of clarity, we report fraction of overinflated tissue for each PEEP selection the plateau pressure is equal to method divided in quartiles on the x-axis (the average value within each quartile is shown) and on the y-axis the ratio of respiratory system the corresponding average PEEP. Horizontal and vertical bars represent sds. The statistical analysis was linear compliance to the tidal volregression performed on individual data points; only PEEP by the ExPress method was significantly related to the fraction of overinflated tissue (ExPress study protocol-derived PEEP [cm H2O] = –0.02 + 0.004 × fraction ume to which the PEEP value of overinflated tissue, r2 = 0.09, p = 0.04), whereas the relationship between PEEP selected by lung open must be added. The inspiratory ventilation (LOV) (LOV study-derived PEEP [cm H2O] = 0.04 – 0.0002 × fraction of overinflated tissue, r2 = 0, 2 recruitment may in part alter p = 0.92), stress index (PEEP [cm H2O] = 0.008 + 0.002 × fraction of overinflated tissue, r = 0.03, p = 0.21), and absolute value of esophageal pressure (esophageal pressure-derived PEEP [cm H2O] = 0.06 – 0.002 × this relationship; however, as fraction of overinflated tissue, r2 = 0.05, p = 0.14) and fraction of overinflated tissue were not significant. the recruitability is remarkable only in few ARDS patients, the and that were significantly associated, although weakly, with general pattern holds. Actually, in our study, the ExPress and stress index methods selected higher PEEP in healthier patients lung recruitability (2, 27). The PEEP levels obtained by the different methods, although with a larger amount of normally aerated tissue (Figs. E13 and similar on average, were not related in individual patients with E14, Supplemental Digital Content 1, http://links.lww.com/ the exception of the ExPress (6) and stress index (5) methods. CCM/A730). The amount of the overinflated tissue at 45 cm These two approaches are based on the upper inflection region of H2O was small (2–4%) but in line with the literature. It must be the volume-pressure curve (stress index) or its surrogate (30 cm noted that CT scan is not the ideal tool to measure overinflation H2O airway plateau pressure, ExPress [6]). In practice, being the because it is defined by the gas/tissue ratio and, in ARDS, the tidal volume constant, the PEEP is progressively increased until tissue mass is largely increased due to edema. Figure 2. (Continued). Horizontal and vertical bars represent sds. The statistical analysis was linear regression performed on individual data points; only the PEEP selected by the lung open ventilation (LOV) method was significantly related to lung recruitability (LOV study protocol-derived PEEP [cm H2O] = 8.8 + 17.9 × lung recruitability, r2 = 0.29, p < 0.0001), whereas the relationship between PEEP selected by ExPress (ExPress studyderived PEEP [cm H2O] = 15.0 – 3.9 × lung recruitability, r2 = 0.02, p = 0.33), stress index (stress index-derived PEEP [cm H2O] = 14.5 – 5.5 × lung recruitability, r2 = 0.03, p = 0.22), and absolute value of esophageal pressure (esophageal pressure-derived PEEP [cm H2O] = 11.4 + 7.4 × lung recruitability, r2 = 0.04, p = 0.19) and lung recruitability were not significant. Lower Panel summarizes the behavior of the end-expiratory transpulmonary pressure levels selected with the four bedside methods as a function of lung recruitability. For the sake of clarity, we report lung recruitability for each PEEP selection method divided in quartiles on the x-axis (the average value within each quartile is shown) and on the y-axis the corresponding average end-expiratory transpulmonary pressure selected. Horizontal and vertical bars represent sds. The statistical analysis was linear regression performed on individual data points; only the end-expiratory transpulmonary pressure selected by the LOV method was significantly related to lung recruitability (LOV study protocol-derived end-expiratory transpulmonary pressure [cm H2O] = 6.5 + 16.3 × lung recruitability, r2 = 0.22, p = 0.0006), whereas the relationship between end-expiratory transpulmonary pressure selected by ExPress (ExPress study-derived end-expiratory transpulmonary pressure [cm H2O] = 10.6 + 0.8 × lung recruitability, r2 = 0.001, p = 0.86), stress index (stress index-derived transpulmonary pressure [cm H2O] = 11 – 1.8 × lung recruitability, r2 = 0.003, p = 0.72), and absolute value of esophageal pressure (esophageal pressure-derived transpulmonary pressure [cm H2O] = 8.4 + 8.0 × lung recruitability, r2 = 0.03, p = 0.23) and lung recruitability were not significant.
262
www.ccmjournal.org
February 2014 • Volume 42 • Number 2
Feature Articles
The rationale behind the esophageal pressure method is to provide at end-expiration a positive transpulmonary pressure to prevent lung collapse. Therefore, assuming that the absolute value of the esophageal pressure is equal to the pleural pressure, the PEEP must be at least equal to the esophageal pressure. This assumption, on which the method is based, however, is questionable: while it is generally accepted that esophageal pressure may record regional pressure variations accurately, its absolute value is unrelated with pleural pressure (28, 29), and possibly influenced by several factors, as the presence of pleural effusion or ascites (30). Therefore, and not surprising, the PEEP values provided by esophageal pressure method were similar in mild, moderate, and severe ARDS and unrelated with lung recruitability (Fig. E15, Supplemental Digital Content 1, http://links.lww.com/CCM/ A730). The method we applied is similar but not equal to the one proposed by Talmor et al (14), as the present study began before his publication. To obtain PEEP values more comparable to the ones obtained by Talmor et al, the values we obtained should be multiplied for the ratio of total respiratory system elastance to the lung elastance. The final results are PEEP values 2–3 cm H2O higher than reported in this study: higher, but similar in mild, moderate, and severe ARDS and unrelated to recruitability. The LOV method (8) does not aim specifically to full lung opening, but to acceptable oxygenation that may be reached leaving part of the lung unrecruited. Actually the same level of oxygenation may be reached with a not inflated tissue ranging from 20% to 40% of the entire lung (27, 31). The degree of oxygenation depends on the size of the baby lung (32), which, in turn, is inversely related with the lung edema and recruitability. In a previous study, we found that the degree of oxygenation at 5 cm H2O PEEP was the variable most associated with lung recruitability (2); therefore, it is not surprising that in patients with lower recruitability (higher amount of well inflated tissue), the PEEP necessary to reach the oxygenation target is lower than the one necessary in ARDS patients with a small “baby lung” and higher recruitability (Fig. E16, Supplemental Digital Content 1, http://links.lww.com/CCM/A730). Two meta-analyses concluded that higher PEEP levels seem beneficial in severe ARDS, while in mild ARDS could be potentially harmful (3, 13). This implies that recruitability could be a key factor to exploit the PEEP benefits as the severe ARDS is characterized by higher recruitability compared with the moderate and mild ARDS (Table 2). If we consider our findings in this framework, we may conclude that: 1) higher PEEP should be used in severe ARDS patients, with higher recruitability, in whom all the methods, based on mechanics, esophageal pressure, or oxygenation, provided similar higher PEEP as explicitly suggested by Briel et al (3) in his meta-analysis; 2) lower PEEP, as indicated by LOV study, should be reserved to mild ARDS patients with lower recruitability. In fact, although direct evidences are lacking, it does not appear reasonable to expose these patients to the side effects of higher PEEP, as indicated by ExPress, stress index, and esophageal pressure methods, to keep open only few grams of lung tissue. Critical Care Medicine
Therefore, we believe that the advantage of the LOV method is that it minimizes the risk of higher PEEP in lower “recruiters,” while providing in the higher “recruiters” similar PEEP indicated by the other methods. We realize that a randomized study in which higher and lower PEEP are tested in higher and lower recruiter patients is necessary. However, independent on the possible ethical problems potentially implied in such a study (we would exitate to provide 5 cm H2O PEEP in very severe ARDS patients), it would take hundreds of patients in whom lung recruitability should be assessed by CT scan and several years to be completed. In the meantime, it is possible that data provided by our study may help to provide a more rational use of PEEP.
CONCLUSIONS Most methods provided similar PEEP in mild, moderate, and severe ARDS, unrelated to the lung recruitability. However, as it seems unreasonable to expose lower recruiters (mild ARDS patients) to higher PEEP to keep open few grams of lung, the LOV method appears the only one which selects a PEEP somehow related to lung recruitability and to the degree of severity of Berlin definition, avoiding higher PEEP in patients with lower recruitability.
ACKNOWLEDGMENTS We thank Prof. Peter Suter and Prof. John J. Marini for their invaluable help and suggestions and Dr. Angelo Colombo for statistical advice. We are indebted to Matteo Brioni, Federica Menga, and Irene Cigada for technical assistance with the CT scan analysis.
REFERENCES
1. Ashbaugh DG, Bigelow DB, Petty TL, et al: Acute respiratory distress in adults. Lancet 1967; 2:319–323 2. Gattinoni L, Caironi P, Cressoni M, et al: Lung recruitment in patients with the acute respiratory distress syndrome. N Engl J Med 2006; 354:1775–1786 3. Briel M, Meade M, Mercat A, et al: Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome systematic review and meta-analysis. JAMA 2010; 303:865–873 4. Caironi P, Cressoni M, Chiumello D, et al: Lung opening and closing during ventilation of acute respiratory distress syndrome. Am J Respir Crit Care Med 2010; 181:578–586 5. Grasso S, Terragni P, Mascia L, et al: Airway pressure-time curve profile (stress index) detects tidal recruitment/hyperinflation in experimental acute lung injury. Crit Care Med 2004; 32:1018–1027 6. Mercat A, Richard JC, Vielle B, et al; Expiratory Pressure (Express) Study Group: Positive end-expiratory pressure setting in adults with acute lung injury and acute respiratory distress syndrome: A randomized controlled trial. JAMA 2008; 299:646–655 7. Brower RG, Lanken PN, MacIntyre N, et al; National Heart, Lung, and Blood Institute ARDS Clinical Trials Network: Higher versus lower positive end-expiratory pressures in patients with the acute respiratory distress syndrome. N Engl J Med 2004; 351:327–336 8. Meade MO, Cook DJ, Guyatt GH, et al; Lung Open Ventilation Study Investigators: Ventilation strategy using low tidal volumes, recruitment maneuvers, and high positive end-expiratory pressure for acute lung injury and acute respiratory distress syndrome: A randomized controlled trial. JAMA 2008; 299:637–645 9. Suter PM, Fairley B, Isenberg MD: Optimum end-expiratory airway pressure in patients with acute pulmonary failure. N Engl J Med 1975; 292:284–289 www.ccmjournal.org
263
Chiumello et al 10. Gattinoni L, Pesenti A: ARDS: The non-homogeneous lung; facts and hypothesis. Intensive Critical Care Digest 1987; 6:1–4 11. Grasso S, Stripoli T, Sacchi M, et al: Inhomogeneity of lung parenchyma during the open lung strategy: A computed tomography scan study. Am J Respir Crit Care Med 2009; 180:415–423 12. Matamis D, Lemaire F, Harf A, et al: Total respiratory pressure-volume curves in the adult respiratory distress syndrome. Chest 1984; 86:58–66 13. Phoenix SI, Paravastu S, Columb M, et al: Does a higher positive end expiratory pressure decrease mortality in acute respiratory distress syndrome? A systematic review and meta-analysis. Anesthesiology 2009; 110:1098–1105 14. Talmor D, Sarge T, Malhotra A, et al: Mechanical ventilation guided by esophageal pressure in acute lung injury. N Engl J Med 2008; 359:2095–2104 15. Bernard GR, Artigas A, Brigham KL, et al: The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 1994; 149:818–824 16. ARDS Definition Task Force, Rubenfeld GD, Thompson BT, et al: Acute respiratory distress syndrome: The Berlin Definition. JAMA 2012; 307:2526–2533 17. Camporota L, Ranieri VM: What’s new in the “Berlin” definition of acute respiratory distress syndrome? Minerva Anestesiol 2012; 78:1162–1166 18. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000; 342:1301–1308 19. Gattinoni L, Chiumello D, Cressoni M, et al: Pulmonary computed tomography and adult respiratory distress syndrome. Swiss Med Wkly 2005; 135:169–174 20. Agostoni A, Hyatt RE: Static behavior of the respiratory system. In: Handbook of Physiology, Volume III. Mechanics of Breathing, Part 1; American Physiological Society, Bethesda, Maryland, 1986, pp 113–130
264
www.ccmjournal.org
21. Borges JB, Okamoto VN, Matos GF, et al: Reversibility of lung collapse and hypoxemia in early acute respiratory distress syndrome. Am J Respir Crit Care Med 2006; 174:268–278 22. Pelosi P, Goldner M, McKibben A, et al: Recruitment and derecruitment during acute respiratory failure: An experimental study. Am J Respir Crit Care Med 2001; 164:122–130 23. Beydon L, Svantesson C, Brauer K, et al: Respiratory mechanics in patients ventilated for critical lung disease. Eur Respir J 1996; 9:262–273 24. Lachmann B: Open lung in ARDS. Minerva Anestesiol 2002; 68:637–642; discussion 640, 643 25. Hickling KG: The pressure-volume curve is greatly modified by recruitment. A mathematical model of ARDS lungs. Am J Respir Crit Care Med 1998; 158:194–202 26. de Matos GF, Stanzani F, Passos RH, et al: How large is the lung recruitability in early acute respiratory distress syndrome: A prospective case series of patients monitored by computed tomography. Crit Care 2012; 16:R4 27. Cressoni M, Caironi P, Polli F, et al: Anatomical and functional intrapulmonary shunt in acute respiratory distress syndrome. Crit Care Med 2008; 36:669–675 28. Brander L, Ranieri VM, Slutsky AS: Esophageal and transpulmonary pressure help optimize mechanical ventilation in patients with acute lung injury. Crit Care Med 2006; 34:1556–1558 29. Hedenstierna G: Esophageal pressure: Benefit and limitations. Minerva Anestesiol 2012; 78:959–966 30. Chiumello D, Marino A, Cressoni M, et al: Pleural effusion in patients with acute lung injury: A CT scan study. Crit Care Med 2013; 41:935–944 31. Dantzker DR, Lynch JP, Weg JG: Depression of cardiac output is a mechanism of shunt reduction in the therapy of acute respiratory failure. Chest 1980; 77:636–642 32. Gattinoni L, Pesenti A: The concept of “baby lung”. Intensive Care Med 2005; 31:776–784
February 2014 • Volume 42 • Number 2
