American Journal of Respiratory and Critical Care Medicine

Mechanical Ventilation

  Conventional Approaches (6)

  Nonconventional Modes (8)

  Permissive Hypercapnia (3)

  Liquid Ventilation (5)

  Ventilator-induced Lung Injury (9)

  Weaning (4)

  Patient Posture (3)

  Noninvasive Ventilation (6)

Acute Lung Injury and Acute Respiratory Distress Syndrome

  Diagnostic Tests (2)

  Animal Models (4)

  Cellular and Molecular Mechanisms (8)

  Water Transport (1)

  Treatment (2)


  Endotoxemia in Animals (8)

  Sepsis in Animals (10)

  Sepsis in Vitro (4)

Ventilator-associated Pneumonia

  Incidence (1)

  Diagnosis (2)

  Treatment (2)

Nosocomial Infections (2)


  Pressure–Volume Curves (3)

  Intrinsic PEEP and Respiratory Muscles (4)

  Hemodynamic Monitoring (3)

  Gas Exchange (1)

  Acid–Base Disorders (1)

Inhaled Nitric Oxide (2)

Toxicology (1)

Ethical Issues (1)

Nonpulmonary Critical Care

  Renal Disorders (4)

  Gastroenterological Disorders (2)

  Cardiac Disorders (1)

  Hematological Disorders (1)

Conventional Approaches

To decide whether some modification of ventilator management represents a true advance, it is necessary first to define what is meant by conventional ventilation. Esteban and coworkers (1) undertook a 1-day point prevalence study of the use of mechanical ventilation. The study included 1,638 patients in 412 intensive care units (ICUs) in North America, South America, and Europe. The indications and ventilator modes were remarkably similar across countries: the commonest indication was acute respiratory failure (66%); and assist-control was used as frequently as a combination of pressure support and intermittent mandatory ventilation. Median tidal volume was 9 ml/kg and positive end-expiratory pressure (PEEP) was not used in 31% of patients. Methods of weaning varied considerably from country to country. The authors conclude that the primary indications for mechanical ventilation and ventilator settings were remarkably similar across countries, but the selection of ventilator modes and weaning techniques varied considerably from country to country.

It has been shown that heat and moisture exchangers provide satisfactory humidification and airway heating for up to 48 hours without adverse effects. To determine the safety and efficiency of changing heat and moisture exchangers only once a week, Ricard and coworkers (2) prospectively studied 60 exchangers used over 377 days of mechanical ventilation. Without changing an exchanger sooner than 7 days, there was no change in frequency of suctioning, peak pressures, absolute humidity, or bacterial colonization; resistance did not change over the 7 days (2.4 and 2.7 cm H2O per liter per second). The authors conclude that a policy based on changing heat and moisture exchangers only once a week should lead to substantial cost savings.

Because of the fundamental importance of neural inspiratory time in research on patient–ventilator interaction, Parthasarathy and coworkers (3) compared the accuracy of indirect estimates of neural inspiratory time against a direct measurement from the diaphragmatic electromyogram. In nine ventilator-support patients, the agreement between indirect estimates of neural inspiratory time, based on recordings of airflow, esophageal pressure, and transdiaphragmatic pressure, and the direct measurement was poor during both mechanical ventilation and spontaneous breathing. Estimating the onset of inspiratory muscle activity from the flow tracing was systematically delayed because of intrinsic PEEP. The authors conclude that the observed errors in indirect estimates of neural inspiratory time are of sufficient magnitude to produce erroneous conclusions about patient–ventilator interactions and cause problems in patient care.

Although nocturnal continuous positive airway pressure (CPAP) is used in patients with congestive heart failure and central sleep apnea, its effects on the cardiac volumes of patients with congestive heart failure have not been studied. Mehta and coworkers (4) addressed this issue in 13 patients with ischemic cardiomyopathy and 9 patients with idiopathic dilated cardiomyopathy. CPAP did not affect ventricular volumes in patients with ischemic cardiomyopathy, but caused global reductions in patients with idiopathic dilated cardiomyopathy. In patients with dilated cardiomyopathy, the decreases in end-systolic and end-diastolic volumes were greater for the right ventricle than for the left ventricle. The authors conclude that CPAP causes greater reductions in ventricular volume in patients with dilated cardiomyopathy, presumably because they have more compliant ventricles, and it causes greater reductions of the right than of the left ventricle, presumably because the wall is thinner.

In mechanically ventilated sheep, Lu and coworkers (5) used computed tomography to study the effect of endotracheal suctioning on bronchial cross-sectional area. At inspired oxygen of 30%, suctioning produced atelectasis, a 29% decrease in bronchial cross-sectional area, oxygen desaturation (from 95 to 87%), an increase in venous admixture (19 to 31%), and an increase in lung tissue resistance. At 100% oxygen, suctioning caused extension of the atelectasis, an increase in venous admixture (from 19 to 36%), but bronchial cross-sectional area decreased by only 7% and oxygen desaturation was prevented. A recruitment maneuver after suctioning completely reversed the increase in lung tissue resistance and atelectasis. The authors conclude that atelectasis caused by endotracheal suctioning is associated with a decrease in the airway cross-sectional area.

Otis (6) recalls a study of the effect of positive-pressure breathing on venous pressures.

Nonconventional Modes

Inverse ratio ventilation has been recommended in the management of patients with the acute respiratory distress syndrome (ARDS) because it may improve gas exchange. Neumann and coworkers (7) compared the effect of inverse ratio ventilation (inspiratory-to-expiratory time ratio of 3:1) with a PEEP of 20 cm H2O in an oleic acid lung injury model in pigs, using the multiple inert gas elimination technique, computed tomography, and perfusion scintigraphy. At similar mean airway pressure, oxygenation improved and shunt decreased more with PEEP than with inverse ratio ventilation; the improvement was caused by greater and more even aeration of the lung, whereas the distribution of perfusion was the same with both. The authors conclude that inverse ratio ventilation did not offer any advantage in this lung injury model.

To compare the responses of two ventilator modes to an increase in respiratory impedance, Grasso and coworkers (8) bound the chest wall of 10 ventilator-supported patients. When ventilated with pressure support, binding caused a 29% decrease in tidal volume, a 58% increase in respiratory frequency, and no change in minute ventilation. Proportional-assist ventilation achieved a minute ventilation equivalent to that with pressure support, but with a smaller decrease in tidal volume (10%) and a lesser increase in frequency (14%). Patient effort was 22% greater with pressure support than with proportional- assist ventilation. The authors conclude that proportional-assist ventilation is more effective than pressure support in defending tidal volume against an increase in respiratory load.

Dojat and coworkers (9) examined the ability of a computerized closed-loop system to achieve a target level of ventilation in patients receiving pressure-support ventilation. The target was achieved 93% of the study time with the closed loop system, and 66% of the time when physicians set pressure support. An indirect measure of patient effort, P0.1 above 4 cm H2O, was met 11% of the study time with closed loop ventilation, and 34% of the time when physicians adjusted pressure support. The authors conclude that closed-loop ventilation increased the time spent at a targeted level of ventilation and reduced periods of excessive work.

In eight patients with ARDS receiving pressure-control ventilation (airway pressure, 40 cm H2O), Hoffman and coworkers (10) assessed the usefulness of tracheal gas insufflation at 10 liters per minute. After commencing gas insufflation, pressure control was decreased by 5 cm H2O: peak airway pressure fell by 15%, tidal volume fell by 16%, but Pco 2 and Po 2 were not different from baseline. A decrease in pressure control of 10 cm H2O caused a 13% fall in Po 2 and a 19% increase in Pco 2. The authors conclude that use of tracheal gas insufflation permits a 5-cm H2O decrease in set ventilator pressure, and the benefit is evident within 30 minutes.

Rossi and coworkers (11) examined the acute response to a modified form of tracheal gas insufflation, using a reverse thrust catheter that delivers gas close to the carina. Switching 11 patients with ARDS from conventional ventilation to gas insufflation caused a 15% fall in Pco 2, despite no change in minute ventilation. Because dynamic hyperinflation is a known complication with other forms of gas insufflation, they monitored end-expiratory volume and found that it fell with use of their reverse-thrust catheter. The authors conclude that a reverse-thrust catheter makes it possible to improve gas exchange in patients with ARDS, without causing dynamic hyperinflation.

During normal breathing, healthy subjects show considerable variability in tidal volume and frequency, but this variability is reduced to a negligible value by conventional mechanical ventilation. The imposition of greater than usual variability of tidal volume and frequency during mechanical ventilation has been reported to improve gas exchange and lung function in pigs experiencing acute lung injury. Nam and coworkers (12) attempted to reproduce these findings. They used conventional ventilation and ventilation with imposed variability to ventilate dogs subjected to an oleic acid lung injury; they matched mean tidal volume and frequency, and hemodynamic variables. Ventilation with imposed variability did not achieve better oxygenation or static compliance than did conventional ventilation. The authors conclude that the generation of greater than usual variability during mechanical ventilation is not superior to conventional ventilation when hemodynamic and ventilatory variables are well matched.

Mutch and coworkers (13) produced right lung collapse in pigs and compared the recruitment achieved by two modes of ventilation: conventional support and imposed variation in delivered volume and rate. Ventilation with imposed variability produced increases in Po 2 (32%) and compliance (45%), and decreases in Pco 2 (37%), shunt (34%), and mean airway pressure (16%). The authors conclude that ventilator support incorporating the normal respiratory variability achieves better recruitment of atelectatic lung.

In three canine models of acute lung injury, Van der Kloot and coworkers (14) investigated the response to a recruitment maneuver, consisting of a sustained inflation at high airway pressures. The response differed among the models: the pneumonia model was least responsive to PEEP; only in the oleic acid model did changes in oxygenation correlate with end- expiratory volume; and only in the saline lavage model did increasing tidal volume improve oxygenation. Recruitment was not seen at a PEEP of 20 cm H2O, probably because the lungs were already fully recruited. The authors conclude that the response to a recruitment maneuver depends on the cause of the underlying lung injury.

Permissive Hypercapnia

The change in oxygenation with permissive hypercapnia is variable because multiple factors are involved. To investigate the influence of permissive hypercapnia on gas exchange, Feihl and coworkers (15) used the multiple inert gas elimination technique in eight patients with ARDS. Permissive hypercapnia, achieved by a 40% decrease in tidal volume, produced a decrease in Po 2 (109 to 92 mm Hg) and increases in cardiac output (by 1.4 liter per minute per m2) and shunt (from 32 to 48%), but no change in the dispersion of ventilation–perfusion ratios. On returning patients to conventional ventilation and adding dobutamine to avoid a fall in cardiac output, the investigators found the shunt still elevated (38%) and Po 2 still depressed (93 mm Hg). The authors conclude that permissive hypercapnia increases shunt and that the deterioration in gas exchange is explained by a decrease in alveolar ventilation and an increase in cardiac output.

Weber and coworkers (16) investigated the usefulness of tromethamine [tris(hydroxymethyl)aminomethane, THAM], a buffer that does not increase CO2 production, in preventing hemodynamic consequences of permissive hypercapnia. After aiming for a target Pco 2 of 80 torr, 12 patients with ARDS were randomized to THAM or pH was left uncorrected. Permissive hypercapnia produced an increase in cardiac output and a fall in systemic vascular resistance. An increase in pulmonary artery pressure occurred only in patients with untreated acidosis. Myocardial contractility, measured by transesophageal echocardiography, fell by 18% in patients with untreated acidosis and by 10% in patients receiving THAM. One hour after stopping permissive hypercapnia, all values returned to baseline. The authors conclude that permissive hypercapnia produces reversible depression of myocardial contractility and hemodynamic abnormalities that were attenuated by THAM buffering.

THAM has greater buffering capacity than bicarbonate, is effective in buffering both metabolic and respiratory acidosis, and is excreted by the kidneys so that CO2 production is not increased. Kallet and coworkers (17) administered THAM to ten patients with acute lung injury and acidosis (pH 7.14), in whom pH could not be controlled despite eucapnic ventilation or permissive hypercapnia. THAM produced improvement in pH and base deficit, and a decrease in Pco 2. The authors conclude that THAM may prove a useful alternative to bicarbonate therapy in patients with acute lung injury.

Liquid Ventilation

Partial liquid ventilation consists of conventional gas ventilation superimposed on lungs partly filled with perfluorocarbon. Among potential beneficial actions is a decrease in pulmonary capillary permeability secondary to tamponade of injured alveoli. In dogs with lung injury induced by oleic acid, Lange and coworkers (18) found similar lung water and pulmonary vascular permeability with both liquid and gas ventilation, but decreased myeloperoxidase in lung tissue with liquid ventilation. The authors conclude that the liquid ventilation does not decrease the pulmonary vascular leak or lung water accumulation in acute lung injury, and its beneficial action may instead be mediated through modifying the inflammatory response.

Perfluorocarbons, which are used with partial liquid ventilation, do not mix with water and are almost biologically inert, yet they may interact mechanically with elements in the lumena of alveoli and airways, and, thus, affect clearance of alveolar liquid. To address these issues, Ricard and coworkers (19) studied the action of perflubron (LiquiVent) on the transport properties of alveolar epithelium. When Ringer solution was instilled into the airspace of rats, perflubron accelerated its resorption. Ringer solution on its own exponentially increased the passive permeability of the alveolar–airway barrier in proportion to the volume instilled; a high dose of perflubron prevented the increase in permeability. The authors conclude that perfluorocarbons may help in pulmonary edema because they cause the redistribution of alveolar liquid over a larger surface area, thus accelerating reabsorption.

Because partial liquid ventilation produces increases in alveolar and pleural pressure, Fessler and Pearce (20) investigated whether it would cause pulmonary artery wedge pressure to overestimate transmural left arterial pressure (as occurs with PEEP). They compared hemodynamic measurements in sheep receiving conventional and liquid ventilation. Two levels of partial liquid ventilation (10 and 30 ml/kg) had minimal influence on pericardical pressure or cardiac output. At varying levels of PEEP, measurement of pulmonary artery wedge pressure during partial liquid ventilation agreed at least as well with transmural left arterial pressure as during conventional ventilation. The authors conclude that hemodynamic measurements are at least as accurate during partial liquid ventilation as during conventional ventilation.

In sheep experiencing acute lung injury secondary to saline lavage, Fujino and coworkers (21) investigated whether gas exchange would differ between partial liquid ventilation and conventional ventilation if PEEP were set at 1 cm H2O above the lower inflection point on the pressure–volume curve. With PEEP above the lower inflection point, oxygenation improved equally with both ventilator techniques. With conventional ventilation, the use of PEEP produced increases in dead space and Pco 2 that were prevented with liquid ventilation. Histology revealed less injury of the dependent lung with liquid ventilation. The authors conclude that when PEEP exceeds the lower inflection point, conventional ventilation and liquid ventilation achieve similar oxygenation, although lung structure is better preserved with liquid ventilation.

To elucidate the effect of infant respiratory distress syndrome and subsequent partial liquid ventilation on cardiopulmonary function, Lopes Cardozo and coworkers (22) did ventricular pressure–volume tracings in newborn lambs. Induction of the respiratory distress syndrome caused Po 2 to fall, pulmonary artery pressure to increase, and no change in cardiac output or stroke volume. Right ventricular contractility increased (but not when corrected for preload), and left ventricular contractility did not change. Use of partial liquid ventilation improved pulmonary function, pulmonary artery pressure remained high, and right ventricular contractility increased. The authors conclude that, despite increased afterload, the right ventricle of the newborn maintains cardiac output through homeometric autoregulation.

Ventilator-induced Lung Injury

To elucidate the role of cell deformation in causing ventilator-induced lung injury, Tschumperlin and coworkers (23) investigated the effect of tightly controlled deformations on the viability of the rat alveolar epithelial Type II cell in vitro. Cyclic deformations produced more injury than tonically held deformation, and injury was proportional to the frequency of the deformations. Most of the cell death occurred in the first five minutes. Superimposing small cyclic deformations on a tonic deformation and thus reducing the amplitude of cell deformations decreased cell death. The authors conclude that ventilator strategies that limit the amplitude of cell deformation may decrease epithelial injury.

To investigate the mechanism whereby hypercapnia protects against acute lung injury, Laffey and coworkers (24) studied 40 isolated perfused rabbit lung preparations exposed to ischemia–reperfusion injury. Both metabolic and hypercapnic acidosis protected against lung injury, but hypercapnic acidosis provided the greater protection. Buffering of the hypercapnic acidosis lessened the protection. The different degrees of protection are not explained by inhibition of xanthine oxidase, because the enzyme was inhibited to a similar degree by both forms of acidosis. The authors conclude that both hypercapnic acidosis and metabolic acidosis protect against acute lung injury caused by ischemia reperfusion, but hypercapnic acidosis provides the greater protection.

To study the effect of ventilator strategies on the translocation of endotoxin from the lung to systemic circulation, Murphy and coworkers (25) studied rabbits subjected to saline lung lavage. After instilling endotoxin into the trachea, animals receiving conventional ventilation (tidal volume of 12 ml per kg and PEEP of 0 cm H2O) developed increased levels of endotoxin and tumor necrosis factor α in their plasma; this did not occur in animals ventilated by a protective strategy (tidal volume of 5 ml per kg and PEEP of 10–13 cm H2O) or in animals receiving conventional ventilation without endotoxin instilled in the trachea. Nonsurvivors had higher levels of plasma endotoxin. The authors conclude that the translocation of lung- derived endotoxin into the systemic circulation depends on the ventilator strategy.

To determine whether hypocapnic alkalosis causes lung injury independent of alterations in ventilator settings, Laffey and coworkers (26) used an isolated perfused lung model. After prolonged ventilation, hypocapnia caused greater increases in airway pressure, pulmonary artery pressure, lung weight, and microvascular permeability than in control animals. Injury following ischemia–reperfusion was also worse in the hypocapnia group. The authors conclude that hypocapnic alkalosis is directly injurious to the lung, and the degree of injury is proportional to the degree of hypocapnic alkalosis.

The benefit of permissive hypercapnia is usually attributed to decreased stretch of ventilated lung, but it may have other protective effects. This possibility was investigated by Laffey and coworkers (27) in an ischemia–reperfusion model. Rabbit lungs were ventilated with an inspired carbon dioxide concentration of 0 or 12%; the latter group had a Pco 2 of 101 mm Hg and pH of 7.10. Compared with normocapnia, hypercapnia reduced pulmonary edema and protein leakage, preserved mechanics, and improved oxygenation in the ischemia–reperfusion model. Lavage levels of tumor necrosis factor α in the hypercapnic rabbits were 4% of those in normocapnic rabbits; levels of 8-isoprostane were 30% of control levels. Tissue nitration was attenuated in the hypercapnic animals. The authors conclude that hypercapnia protects against ischemia–reperfusion injury by attenuating inflammation and free radical–mediated injury. An editorial commentary by Hickling (28) accompanies this article.

In a study of isolated, perfused rabbit lungs, Hotchkiss and coworkers (29) asked, “At matched pulmonary artery and airway pressures, does a decrease in ventilator frequency decrease lung damage?” When tidal volumes and peak pulmonary artery pressures were matched, a ventilator frequency of 20 breaths per minute induced three times more perivascular hemorrhage and 4.5 times greater lung weight than a frequency of 3 breaths per minute. An increase in peak pulmonary artery pressure from 20 to 35 mm Hg also caused more alveolar hemorrhage. The authors conclude that decreasing ventilator frequency can diminish lung damage.

Keratinocyte growth factor is secreted by fibroblasts and causes hyperplasia of alveolar Type II cells, and it has been shown to reduce lung injury caused by several agents. In an ex vivo rat model of ventilator-induced injury, Welsh and coworkers (30) found that pretreatment with keratinocyte growth factor decreased the accumulation of lung water and alveolar protein, and decreased the accumulation of fluorescein-labeled dextran that had been administered intravenously. The authors conclude that pretreatment with keratinocyte growth factor attenuates ventilator-induced lung injury by a mechanism that prevents increases in permeability.

Deslee and coworkers (31) described 10 consecutive patients developing a fibrinous pseudomembrane in the trachea after a brief period of endotracheal intubation. The pseudomembrane consisted of a thick tube adhering to the tracheal wall at the site of the cuff. It was white and rubbery in consistency and molded to the tracheal wall. Shortly after extubation, patients developed acute respiratory failure that was positional in nature, with the partly detached membrane causing a valvelike action. One patient died, and mechanical ablation was successful in the remaining nine patients. The authors conclude that the pathological findings and location of the lesion are consistent with an ischemic injury of the tracheal wall caused by pressure in the cuff.


Gastric mucosal acidosis is associated with a poor prognosis in critically ill patients. To estimate mesenteric perfusion and oxygenation in patients being weaned from mechanical ventilation, Maldonado and coworkers (32) used capnographic recirculation gas tonometry. Patients failing a weaning trial developed a 12% increase in regional intramucosal Pco 2 over time, whereas it fell by 9% in successfully weaned patients—the absolute values, however, did not differ between the groups. Gas tonometry was able to predict weaning outcome, but regional intramucosal Pco 2 was less accurate than frequency-to-tidal volume ratio. The authors conclude that the changes in regional intramucosal Pco 2 differ between weaning success and failure patients, but the measurement is a less reliable predictor of weaning outcome than some pre-existing tests.

When a patient with impaired consciousness performs satisfactorily on weaning assessment, many physicians commonly delay extubation out of concern that the patient will not maintain a satisfactory airway. Coplin and coworkers (33) prospectively studied the effect of delaying extubation of 136 brain- injured patients receiving mechanical ventilation. When patients met readiness criteria, 73% were extubated within 48 hours. When extubation was delayed, patients developed almost twice as much pneumonia and stay in the ICU was doubled. Severity of coma did not account for delayed extubation, and the rate of reintubation was the same for both groups. The authors conclude that extubation should not be delayed solely because of impaired neurologic status.

Accidental extubation occurs in 3–16% of mechanically ventilated patients. In a case-control study, Epstein and coworkers (34) examined the effect of accidental extubation on clinical outcome. Of 75 patients experiencing accidental extubation, 42 (56%) required reintubation. When accidental extubation occurred during a weaning trial, 30% required reintubation. When accidental extubation occurred during ventilator support, 76% required reintubation. Accidental extubation did not influence mortality, but patients requiring reintubation had longer durations of ventilator support (19 vs. 11 days), stay in the intensive care unit (21 vs. 14 days), and hospital stay (30 vs. 21 days). The authors conclude that patients requiring intubation after accidental extubation require more prolonged care.

Purro and coworkers (35) sought to determine the mechanisms of long-term ventilator dependency. Patients who were chronically dependent on mechanical ventilation had decreases in tidal volume and respiratory muscle strength, high respiratory drive, and abnormal lung mechanics. Some patients who failed to wean despite a low frequency-to-tidal volume ratio had ineffective inspiratory efforts, which artifactually lowered the ratio. The authors conclude that respiratory distress and CO2 retention resulted from an imbalance between workload and inspiratory muscle strength.

Patient Posture

The use of prone positioning can improve oxygenation in patients with ARDS, but its effect in other causes of acute respiratory failure has not been studied. Nakos and coworkers (36) found that all of eight patients with hydrostatic pulmonary edema developed a sustained improvement in oxygenation when switched from the supine to the prone position: PaO2/F i O2 ratio increased from 72 to 208 after six hours in the prone position. Of 20 patients with ARDS, 15 showed better oxygenation on being switched to the prone position: PaO2/F i O2 ratio increased from 83 to 189. The response was least favorable in patients with the later stages of ARDS, and oxygenation did not improve in any patient with pulmonary fibrosis. The authors conclude that improvement in oxygenation is greater in patients with pulmonary edema, whether cardiac or noncardiac in origin, and the response is poor or absent when fibrosis is present.

Studies have shown that turning an animal from the supine to the prone position causes better ventilation of regions of the lung adjacent to the spine without compromising ventilation in regions close to the sternum. In seven patients, Albert and Hubmayr (37) used computed tomography to measure the volume of lung found under the heart in the supine and prone position. In the region of the diaphragm, 42% of the left lung and 16% of the right lung was under the heart. Less than 4% of either lung was under the heart in the prone position. The authors conclude that better gas exchange in the prone position may be secondary to less compression of the lung by the heart.

In the supine position, a major part of the left lower lobe and a substantial part of the right lower lobe are located beneath the heart. To determine whether the heart compresses the lower lobes of the lung, Malbouisson and coworkers (38) performed computed tomography in 38 patients with ARDS and 14 healthy subjects in the supine posture. The heart of patients with ARDS was larger and heavier than in healthy subjects, especially secondary to protrusion on the left. Pressure exerted by the left heart on the lower lobe was 33% greater in patients with ARDS; consequently, only 12% of the left lower lobe consisted of gas, contrasted with 62% in healthy subjects. The authors conclude that the pressure exerted by the heart causes a dramatic loss of aeration of the lower lobes in patients with ARDS lying in the supine position.

Noninvasive Ventilation

Martin and coworkers (39) undertook a prospective, randomized comparison of noninvasive positive-pressure ventilation versus usual medical care in 61 patients with acute respiratory failure. Noninvasive ventilation decreased the rate of endotracheal intubation to less than one-third that for usual care. The decreased rate of intubation was seen in subgroups with hypoxemic respiratory failure and hypercapnic respiratory failure, but not in patients with chronic obstructive pulmonary disease (COPD). The authors conclude that noninvasive ventilation decreases the rate of intubation in patients with acute respiratory failure of various etiologies.

The decrease in the resistance to gas flow achieved by a gas of low density, such as helium, might facilitate wider use of noninvasive ventilation. Jaber and coworkers (40) studied the effect of adding a helium–oxygen mixture to noninvasive ventilation in 10 patients experiencing an acute exacerbation of COPD. The low-density gas mixture achieved about a 20% decrease in inspiratory effort, a 3-torr decrease in Pco 2, and no change in breathing pattern or oxygenation. The authors conclude that a helium–oxygen mixture enhances the ability of noninvasive ventilation to decrease respiratory effort.

In 45 patients with stable COPD (FEV1 of 0.96 liter), Garrod and coworkers (41) did a randomized controlled trial of exercise training alone versus exercise training plus noninvasive positive pressure ventilation (BiPAP). At 8 weeks, patients receiving both training and ventilation reached 72 meters more on a mean shuttle walk test than patients receiving training alone, and they had greater improvement in PO 2 (3.7 mm Hg) and the chronic respiratory disease questionnaire (12.3). The authors conclude that noninvasive ventilation can be used successfully in the home setting.

Patients with Duchenne muscular dystrophy inevitably develop hypoventilation during sleep, but it is difficult to know when to order a polysomnogram. In 19 patients, Hukins and Hillman (42) assessed the ability of daytime pulmonary function to predict nighttime desaturation. An FEV1 of less than 40% of predicted had a sensitivity of 91% and a specificity of 50%; equivalent values for Pco 2 of 45 mm Hg were 91 and 75% and, for a base excess exceeding 4 mmol per liter, 55 and 44%. Noninvasive ventilation over one year achieved a 5-torr fall in daytime Pco 2 despite a 200-ml decrease in FEV1. The authors conclude that polysomnography should be performed in patients with Duchenne muscular dystrophy when Pco 2 rises above 45 mm Hg, especially when base excess exceeds 4 mmol per liter, and the lower daytime Pco 2 with nighttime ventilation suggests that sleep hypoventilation contributes to the development of respiratory failure.

The efficacy of noninvasive ventilation in patients with stable COPD is debated by Rossi (43) and Hill (44), with rebuttals from each (45, 46).

Diagnostic Tests

Because all definitions of ARDS include a characterization of the chest X-ray, it is important that the radiographic interpretation be reliable. Meade and coworkers (47) studied the agreement among nine physicians who made a global rating of the presence of ARDS on chest X-rays. The physicians reviewed 778 chest X-rays from 99 critically ill patients. For physicians not participating in consensus training, agreement between ratings was moderate (kappa, 0.38 to 0.55). After consensus training, agreement increased markedly (kappa, 0.72 to 0.88). The authors conclude that consensus training is necessary to optimally conduct clinical trials.

In 15 patients with acute lung injury and 12 patients with cardiogenic pulmonary edema, Newman and coworkers (48) measured HTI56, a novel protein of the apical plasma membrane that is specific to the Type I cell in human alveoli. The amount of HTI56 in patients with acute lung injury was 4.3 times higher in alveolar fluid and 1.4 times higher in plasma than in patients with cardiogenic edema. The authors conclude that this specific marker, HTI56, may prove useful as a blood test for detecting alveolar epithelial damage.

Animal Models

Neutrophil elastase plays an important role in the onset of acute lung injury, but its role in the progression of the injury is not known. Kawabata and coworkers (49) used inhaled endotoxin to induce acute lung injury in hamsters, and over the subsequent 2 to 24 h gave a specific inhibitor of neutrophil elastase (sivelestat). Untreated animals developed increased inflammatory cell counts and hemorrhage, paralleled by an increase in neutrophil elastase activity in bronchoalveolar fluid. The inhibitor prevented the progression of lung inflammation and injury. Administration of dexamethasone after endotoxin inhalation was not effective, although it was effective if given before the injury. The authors conclude that neutrophil elastase plays an important role in the progression of acute lung injury.

Because hepatocyte growth factor may help with the regeneration of injured lung, Yamada and coworkers (50) determined whether lung injury induces the expression of hepatocyte growth factor. An ischemia–reperfusion injury, induced by occluding the left lung of rats, produced increases in hepatocyte growth factor in plasma, infiltrating alveolar macrophages, and lung tissue, and it also increased its mRNA expression. Treatment with an antibody directed against hepatocyte growth factor aggravated the lung injury, and reduced DNA synthesis of alveolar epithelial cells. The authors conclude that an ischemia–reperfusion injury of the lung enhances the expression of hepatocyte growth factor.

To better understand how the aspiration of gastric contents cause ARDS, Yamada and coworkers (51) investigated the influence of acid aspiration on the response to endotoxemia. Acid was instilled into the right lung of rats, and lipopolysaccharide was given 24 h later. The combination produced more severe pulmonary edema, lower Po 2, and increased tumor necrosis factor α, nitric oxide production, and accumulation of neutrophils in the alveoli. The effects were attenuated by administration of inhibitors of protein tyrosine kinase; tyrosyl phosphorylation is believed to be necessary for priming of neutrophils. The authors conclude that acid aspiration enhances the subsequent inflammatory response to endotoxin, and that the response depends in part on protein tyrosine kinase.

In an in vivo rabbit model, Nielsen and coworkers (52) investigated whether administration of a nitric oxide donor, DETANONOate, would decrease alveolar fluid clearance. DETANONOate did not change total alveolar fluid clearance. The agent, however, decreased the fraction of the clearance sensitive to amiloride and it increased the amiloride- insensitive fraction. The authors conclude that a nitric oxide donor, DETANONOate, does not decrease alveolar fluid clearance.

Cellular and Molecular Mechanisms

Although fibroproliferation is considered a late event in ARDS, more recent data challenge this view. Retrieving bronchoalveolar fluid from patients within 24 hours of a diagnosis of ARDS, Marshall and coworkers (53) found that it produced more than twice as much proliferation of human lung fibroblasts as control lavage. Levels of N-terminal peptide for type III procollagen, a marker of collagen turnover, were increased more than twofold in the patient serum and more than sixfold in lavage fluid; the values in lavage fluid were higher in nonsurvivors than in survivors. Fluid obtained by lavage at seven days continued to achieve greater stimulation of lung fibroblasts, and activity in nonsurvivors was more than twice that in survivors. The authors conclude that the early occurrence of fibroproliferation in patients with ARDS is an important therapeutic target.

To investigate the role of apoptosis caused by CD8+ cytolytic T cells in acute lung injury, Hashimoto and coworkers (54) performed bronchoalveolar lavage in groups of critically ill patients. In the acute phase of ARDS caused by sepsis, polymerase chain reaction revealed upregulation of messenger RNAs for several apoptosis molecules, such as perforin (pore-forming protein), granzymes (serine proteases), Fas, and Fas ligand. Expression of these molecules was low or negligible in patients with sepsis who were free of ARDS and also in patients with late ARDS caused by sepsis. Expression of genes of the classic proinflammatory cytokines (interleukin 1β, tumor necrosis factor α, interleukin 6, interleukin 8, and inducible nitric oxide synthase) was most distinct in the acute phase of ARDS. In the acute phase of ARDS, two independent death pathways were operating: rapid death of target cells resulting from granzymes and perforin; and rapid apoptosis mediated by the cell surface receptor Fas in the presence of Fas ligand. The authors conclude that the dual apoptosis pathway plays an important role in the pathogenesis of acute lung injury.

Because heat shock proteins provide cells and tissues with protection against the harmful effects of cytokines and reactive oxygen species, Durand and coworkers (55) investigated the expression of cytosolic Hsp 70, a heat shock protein, in blood monocytes of 23 critically ill patients. Basal expression of the heat shock protein was similar in patients with ARDS and control subjects, but its inducibility—a measure of the ability to mount a stress response—decreased in the patients with ARDS. The only factor correlating with inducibility of the heat shock protein was the duration of ventilator support. The authors conclude that the inducibility of heat shock protein in monocytes is decreased in patients with ARDS.

To investigate the importance of alveolar hemostatic balance in severe pneumonia, Gunther and coworkers (56) performed bronchoalveolar lavage in 154 patients with pneumonia and 68 patients with ARDS. Compared with 35 healthy subjects, bronchoalveolar fluid of all patient groups contained almost twice the amount of total procoagulant activity, largely attributable to the tissue factor pathway of coagulation. This change combined with markedly decreased fibrinolytic capacity caused fibrin formation in the alveolar lining layer. Patients with pneumonia requiring ventilator support had virtually identical findings as patients with ARDS, but findings were less marked in patients with pneumonia who were ventilator free. The authors conclude that acute lung injury, whether triggered by lung infection or not, is characterized by both enhanced procoagulant and depressed fibrinolytic activity in the alveolar lining layer.

Actin is released from injured cells and, when it exceeds the protective scavenging capacity, it causes damage to the pulmonary vasculature. Erukhimov and coworkers (57) found that serum from patients with ARDS was toxic when added to cultured endothelial cells from the pulmonary arteries of sheep. Similar toxicity occurred when G-actin was added to normal serum. Toxicity was inhibited by preincubation with gelsolin, which is known to complex with actin. Staining studies also revealed filaments of F-actin in sera from patients with ARDS. The authors conclude that serum from patients with ARDS is toxic to cultured endothelial cells, and part of this toxicity is mediated by actin.

Because human serum albumin is a rich source of sulfur-containing amino acids, as used in the synthesis of glutathione, Cantin and coworkers (58) investigated whether albumin might protect cells against oxidant-mediated injury. After incubation with human serum albumin for eight hours, lung epithelial cells, fibroblasts, and lymphocytes had increased levels of glutathione. Serum depleted of albumin did not increase glutathione. The increase in glutathione was also prevented by reducing the disulfide bonds in albumin, and with administration of inhibitors of vacuolar H+-ATPase or a membrane pH-disrupting ionophore. The increase in glutathione with albumin was sufficient to protect cells against oxidant injury and to decrease activation of nuclear factor κB. The authors conclude that albumin specifically modulates cellular levels of glutathione, and this effect is sufficient to protect against oxidant injury and to downregulate activation of nuclear factor κB.

The incidence and severity of ARDS are increased in patients with a history of chronic alcohol abuse. Moss and coworkers (59) found a more than sixfold reduction in the levels of glutathione, an antioxidant, in alveolar epithelium lining fluid of patients with chronic alcohol abuse who were otherwise healthy. The patients also had a more than threefold increase in the fraction of glutathione in the oxidized form, indicating its increased utilization. The authors conclude that the decrease in glutathione may explain the increase in ARDS among chronic alcoholics.

In a rabbit model of reexpansion pulmonary edema, Nakamura and coworkers (60) found that the re-expanded lung had increased levels of interleukin 8 and its mRNA expression in the alveolar macrophages and lung epithelial cells. Pretreatment with a monoclonal antibody to interleukin 8 attenuated the leukocyte infiltration and lung injury. The authors conclude that interleukin 8 contributes to the development of reexpansion pulmonary edema, and treatment with an antibody against interleukin 8 may prove protective.

Water Transport

Chen and Yeates (61) investigated the contributions of ion transport and osmosis in transporting water into the airway lumen of mechanically ventilated dogs. A near-isosmotic (250 mOsm) aerosol of mannitol instilled into the trachea produced an increase in airway secretory output; sodium and chloride concentrations also increased. A hyperosmotic (950 Osm) aerosol of mannitol produced a greater increase in airway secretory output, without further increases in sodium and chloride contents. The authors conclude that net water transport into the airway lumen is the vectorial sum of the water fluxes associated with actively driven intracellular sodium and chloride ion transport and the flux of water due to osmosis.


Because alveolar Type II cells participate in active sodium transport, Olivera and coworkers (62) investigated the effect of aldosterone (which stimulates Na+,K+-ATPase) on the clearance of lung edema in isolated–perfused rat lungs. The alveolar Type II cells treated with aldosterone had an increase in Na+,K+-ATPase β1-subunit mRNA and protein. Aerosolized aldosterone produced a fourfold increase in Na+,K+-ATPase activity, and a 50% increase in active sodium transport and edema clearance. The authors conclude that the increase in clearance of lung edema with aldosterone is probably secondary to its action on Na+,K+-ATPase in alveolar epithelial cells, and they speculate that aldosterone may be beneficial in the management of lung edema.

Because β-agonists stimulate sodium channels and Na+, K+-ATPase in alveolar epithelial cells, Saldias and coworkers (63) investigated their ability to clear pulmonary edema caused by ventilator-induced injury. Ventilating rats with high tidal volumes caused a decrease in clearance of edema by 50%. Administration of β-agonists increased clearance of edema by as much as two- to three-fold. The beneficial effect of β-agonists was prevented by colchicine, which disrupts the microtubular transport system within cells. The authors conclude that recruitment of ion-transporting proteins from intracellular pools to the plasma membrane of the alveolar epithelial cells causes the benefit.

Endotoxemia in Animals

The accumulation of neutrophils in pulmonary capillaries in response to pathophysiologic stimuli has been attributed to mechanical retention by stiffened neutrophils. To determine whether other mechanisms might contribute to pulmonary sequestration of neutrophils, Kuebler and coworkers (64) used fluorescence microscopy to measure the kinetics of red and white cells in alveolar capillaries of rabbit lung. An infusion of lipopolysaccharide caused neutrophil sequestration and decreased the number of functioning alveolar capillaries. Pretreatment with an antibody directed against L-selectin, an adhesion molecule that mediates neutrophil rolling, completely blocked the effects. The authors conclude that leukocyte- endothelial interaction, mediated by L-selectin, is a necessary condition for the sequestration of leukocytes in this model of acute endotoxemia.

Fujii and coworkers (65) found that rats injected with lipopolysaccharide developed progressive hypotension, accompanied by increases in plasma endothelin, lung inducible nitric oxide synthase, exhaled nitric oxide, serum nitrite–nitrate, and the ratio of lung wet-to-dry weight. Pretreatment with a selective endothelin-A receptor antagonist, P1/fl, prevented the abnormal developments with the exception of the increase in plasma endothelin. The authors conclude that lipopolysaccharide produces hypotension and lung injury through increases in endothelin, which activates endothelin-A receptors and leads to increased nitric oxide synthesis.

To study the role of nitric oxide in endotoxin-induced myocardial depression, Afulukwe and coworkers (66) employed an isolated papillary preparation in a rat model. Depression of myocardial contractility was accompanied by an increase in nitric oxide in the myocardium. Nitric oxide content, however, was also elevated when contractility had recovered at 48 hours. Pretreatment with a relatively selective inhibitor of nitric oxide synthase, S-methylisothiourea, restored myocardial contractility, whereas the nonselective inhibitor, N G-nitro-l-arginine methyl ester (l-NAME), worsened contractility and increased mortality. The authors conclude that increased myocardial concentrations of nitric oxide do not necessarily correlate with depressed contractility.

Because of reports that anesthetics modulate the immune response, Giraud and coworkers (67) investigated the effect of halothane and thiopental on the inflammatory response. Intratracheal instillation of lipopolysaccharide into rat lungs caused increased cell recruitment and cytokines in bronchoalveolar lavage and lung homogenates. In rats receiving halothane, the inflammatory response to lipopolysaccharide was depressed in a dose-dependent manner: recruitment of polymorphonuclear neutrophils was decreased by 55%, and concentrations of tumor necrosis factor α, interleukin 6, and macrophage inflammatory protein 2 were decreased by more than 60%. No such change was seen with thiopental, and neither anesthetic agent modified the inflammatory response in the absence of lipopolysaccharide. The decrease in lung inflammatory response caused by halothane had disappeared by 20 hours. The authors conclude that halothane depresses the inflammatory response to endotoxin.

Little is known about changes to bronchiolar epithelium in animals challenged with lipopolysaccharide. In a rat model of acute lung injury caused by lipopolysaccharide infusion, Arsalane and coworkers (68) investigated the changes in the major secretory protein of Clara cells, CC16, which is known to have an anti-inflammatory action. Acute lung injury caused marked reductions in this secretory protein in bronchoalveolar lavage fluid and related RNA levels in the lung; serum levels of the protein were increased, consequent to increased airway permeability. Pretreatment with dexamethasone decreased lung inflammation, but did not prevent the permeability of Clara cell secretory protein. The authors conclude that acute lung inflammation caused by lipopolysaccharide results in a marked decrease in the synthesis and secretion of Clara cell secretory protein.

Because increased generation of nitric oxide by inducible nitric oxide synthase may contribute to acute lung injury, Evgenov and coworkers (69) investigated the effect of aminoguanidine, an inhibitor of inducible nitric oxide synthase, on lung injury resulting from endotoxemia. In chronically instrumented sheep, endotoxin caused increases in pulmonary artery pressure, pulmonary capillary pressure, extravascular lung water, protein clearance, and lung lymph flow. Administration of the inhibitor after two hours produced further increases in lymph flow and protein clearance, but no change in lung water despite an increase in capillary pressure; the inhibitor also decreased venous admixture and improved oxygenation. The authors conclude that the failure to increase extravascular lung water, despite increased capillary pressure, was caused by increased lymphatic drainage resulting from the consequences of blocking inducible nitric oxide synthase.

Endotoxin is thought to contribute to pulmonary hyperresponsiveness in asthma and ARDS. In an isolated mouse lung, Held and Uhlig (70) studied the effect of endotoxin on airway and vascular hyperreactivity. Infusion of lipopolysaccharide induced hyperreactivity of the airways to methacholine and of the pulmonary vasculature to platelet-activating factor. Both forms of hyperreactivity were completely prevented by blocking the thromboxane/endoperoxide receptor. Blocking cyclooxygenase 2 abolished vascular hyperreactivity, but it had only a marginal effect on airway hyperreactivity. Pretreatment with an oxygen radical scavenger, N-acetylcysteine, partly protected against the increase in airway hyperreactivity. The authors conclude that endotoxin induces airway and vascular hyperreactivity by activating the thromboxane/endoperoxide receptor, that vascular hyperactivity depends on cyclooxygenase 2 activity, and that airway hyperactivity largely depends on other mechanisms.

Tissue factor pathway inhibitor blocks Factor Xa, Factor VIIa, and the synthesis of interleukin 8. In rats given lipopolysaccharide, Enkhbaatar and coworkers (71) determined whether a recombinant form of this protease inhibitor could decrease pulmonary vascular injury. Pre- or posttreatment with the inhibitor decreased the pulmonary vascular injury induced by lipopolysaccharide, as well as the coagulation abnormalities. It prevented the increases in tissue levels of tumor necrosis factor α, cytokine-induced neutrophil chemoattractant, and myeloperoxidase. The inhibitor blocked the increased expression of CD11b and CD18 on the surface of neutrophils in vitro. It inhibited increases in levels of intracellular calcium, a second messenger of neutrophil activation. Two selective inhibitors of Factor Xa and Factor VIIa had no effect on the pulmonary vascular injury despite having potent anticoagulant effects. The authors conclude that a recombinant form of tissue factor pathway inhibitor reduces the pulmonary vascular injury resulting from lipopolysaccharide by inhibiting leukocyte activation and coagulation abnormalities.

Sepsis in Animals

Inadequate splanchnic blood flow is thought to play a pivotal role in the development of multiple organ failure associated with septic shock. In 12 patients with septic shock, Kiefer and coworkers (72) used dopexamine, a combined dopamine and β-adrenergic agent, to study the effect of a 25% increase in cardiac output on splanchnic hemodynamics. Splanchnic blood flow increased by 12%, but the contribution of splanchnic blood flow to total blood flow decreased by 10%. Both global and regional oxygen delivery increased, but regional energy balance, as reflected by regional oxygen consumption and Pco 2 gradients across the gastric mucosa or liver, were not affected. The authors conclude that dopexamine does not have a preferential effect on intestinal and hepatic oxygenation in patients with septic shock.

Recruitment of neutrophils to sites of inflammation depends on adhesion of neutrophils to the endothelium, achieved by upregulation of adhesion molecules on the neutrophils and the endothelium. In a rabbit model of ARDS induced by acute peritonitis, Gardinali and coworkers (73) found that pretreatment with monoclonal antibody against the adhesion molecule on neutrophils, CD11–CD18, decreased animal mortality by 40%. The lungs of rabbits pretreated with the antibody had fewer granulocytes, less alveolocapillary membrane damage, less malonyldialdehyde, a by-product of membrane lipid peroxidation by neutrophils, and more α-tocopherol, an antioxidant. The authors conclude that an antibody against the CD11–CD18 complex, which blocks neutrophil adherence, decreased lung damage and increased survival in rabbits with peritonitis complicated by acute respiratory failure.

Cardiovascular dysfunction and shock develops in 40% of patients presenting with sepsis. Pinsky and Rico (74) investigated changes in left ventricular function in the first four hours after induction of endotoxic shock in dogs. Infusion of endotoxin produced immediate cardiovascular collapse, but no change in several indices of left ventricular contractility. The authors conclude that the cardiovascular collapse in septic dogs is not even partly due to altered left ventricular function.

Hemofiltration may remove some of the soluble inflammatory mediators of sepsis and help as an adjuvant treatment in sepsis. Bellomo and coworkers (75) studied the effect of continuous venovenous hemofiltration (80 ml per kg per hour) on hemodynamics in a canine model of septic shock. Compared with a sham circuit, hemofiltration attenuated the hypotension that followed an infusion of endotoxin. Hemofiltration also reduced the level of endothelin 1, but it had no effect on other soluble mediators (endotoxin, tumor necrosis factor α, and 6-keto prostaglandin F). The authors conclude that continuous venovenous hemofiltration attenuates the early component of endotoxin-induced hypotension, but this effect is not explained by convective clearance of inflammatory mediators.

To investigate the role of inducible nitric oxide synthase in septic lung injury, Okamoto and coworkers (76) administered a selective inhibitor in rats undergoing cecal ligation and puncture. All rats died within 48 hours of the injury. Survival was increased only when the inhibitor was given 12 hours after the injury; giving it immediately or 6 hours after the procedure did not increase survival. The cecal injury increased the activity of nitric oxide synthase in the lung for up to 24 hours, but activity decreased thereafter; because mRNA expression was still elevated at 42 hours, the earlier decrease in activity of nitric oxide synthase was attributed to a posttranscriptional mechanism. The authors conclude that selective inhibition of inducible nitric oxide synthase improves survival after cecal injury, but the self-limited expression of the synthase in the lung suggests that it plays only a limited role in sepsis.

The nuclear factor κB complex directs high-level transcription of many cytokines and proinflammatory genes. Because of limitations with available methods for assessing the in vivo activation and function of the factor complex, Blackwell and coworkers (77) used transgenic mice that expressed Photius luciferase complementary DNA, which is under the control of nuclear factor κB. (A transgenic animal bears a foreign gene [termed a transgene] that is usually spliced to a tissue-specific or cell-specific promotor. The transgene is inserted into a fertilized egg in vitro, and thus becomes integrated into the animal's germ line.) Injection of lipopolysaccharide into the peritoneal cavity of mice produced luciferase gene expression in many organs; expression in the liver was blocked by an inhibitor of nuclear factor κB, confirming that luciferase gene expression is a surrogate marker for activation of the transcription factor. Lipopolysaccharide also increased lung and serum levels of messenger mRNA for cytokines thought to be dependent on nuclear κB. The authors conclude that endotoxin activates gene transcription that is dependent on nuclear factor κB in multiple organs, thus contributing to systemic inflammation, and that transgenic mice can be used as an in vivo assay of this activity.

To investigate the effect of oxygen-derived free radicals on diaphragmatic contractility, Fujimura and coworkers (78) used a rat model of septic peritonitis achieved by cecal ligation and perforation. Force–frequency curves performed on a removed hemidiaphragm revealed decreased contractility. Levels of malondialdehyde, an index of lipid peroxidation mediated by oxygen-derived free radicals, were increased in the diaphragm, as were two of the main antioxidant enzymes, superoxide dismutase and glutathione peroxidase. Diaphragmatic contractility was improved by the administration of polyethylene glycol-absorbed superoxide dismutase (a scavenger of superoxide ions), polyethylene glycol-absorbed catalase (a scavenger of hydrogen peroxide), and dimethyl sulfoxide (a scavenger of hydroxyl radicals) before the cecal injury; these agents also prevented the increase in malondialdehyde in the diaphragm. The authors conclude that several oxygen-derived free radicals contribute to the decreased contractility of the diaphragm in septic peritonitis.

In a rat model of septic peritonitis resulting from cecal ligation and perforation, Fujimura and coworkers (79) investigated the effect of a β-agonist, isoproterenol, on diaphragmatic contractility. Sixteen hours after injury, a hemidiaphragm was removed and placed in an organ bath. Infusion of isoproterenol improved diaphragmatic contractility and it also accelerated recovery from fatigue (induced by repetitive stimulation of the muscle strip). Isoproterenol produced increased levels of cyclic AMP in the muscle; a derivative of cyclic AMP mimicked the inotropic action of isoproterenol; and a β-blocker, propranolol, abolished the effect of isoproterenol. The authors conclude that isoproterenol improves diaphragmatic contractility and accelerates recovery from fatigue in septic peritonitis by activating the adenyl cyclase system.

To determine whether the inducible nitric oxide synthase pathway contributes to impaired skeletal muscle contractility in sepsis, Lanone and coworkers (80) studied samples from the rectus abdominis in 16 septic patients and 21 control subjects. The muscles of the patients had lower contractile force, and increases in inducible nitric oxide synthase expression (mRNA and protein) and activity. An inhibitor of nitric oxide synthase, N-monomethyl-l-arginine (l-NMMA), reversed the impaired contractility. Immunohistochemical studies revealed the generation of peroxynitrite (a highly reactive oxidant formed by the reaction of nitric oxide with superoxide anion); exposure of control muscles to the amount of peroxynitrite found in patients produced a reduction in force generation that was nonreversible. The authors conclude that the inducible isoform of nitric oxide synthase may contribute to the impaired muscle contractility in septic patients via the local generation of peroxynitrite.

To define the pathophysiology of lactic acidosis in sepsis, Chrusch and coworkers (81) measured lactate metabolism and splanchnic and hepatic hemodynamics during Escherichia coli sepsis in dogs. Splanchnic production of lactate increased with sepsis, although splanchnic oxygen delivery did not fall. Extraction of lactate by the liver increased during sepsis, but the increase was less than that seen during lactic acid loading in nonseptic dogs. Vasopressors had no effect on splanchnic production or hepatic extraction of lactate. The authors conclude that lactic acidosis in sepsis results from an increase in splanchnic production of lactate and impaired extraction of lactate by the liver.

Sepsis in Vitro

Because patients with sepsis have impaired microvascular blood flow, Kirschenbaum and coworkers (82) compared cell rheology in 16 critically ill septic patients, 13 critically ill noninfected patients, and 13 healthy subjects. The septic patients had decreased filtration of whole blood secondary to decreases in erythrocyte and neutrophil deformability, increased erythrocyte aggregation, and increased platelet–neutrophil interaction; the latter action was the greatest. The authors conclude that rheologic abnormalities may contribute to the impaired microvascular blood flow in patients with sepsis.

To understand the intracellular mechanisms involved in immunodepression of leukocytes in sepsis, Adib-Conquy and coworkers (83) studied expression of nuclear factor κB in peripheral blood monocytes from patients with sepsis and trauma. Expression of the active form of nuclear factor κB, the p65p50 heterodimer, was decreased in all patients, and survivors had a decrease in an inhibitory form of the transcription factor (the p50p50 homodimer). Stimulation of patient monocytes with lipopolysaccharide did not induce further translocation of nuclear factor κB. Nonsurvivors showed a predominance of the inactive homodimer, and the ratio of the active to inactive forms was inversely correlated with interleukin 10 levels (r = −0.8). The decreased expression of nuclear factor κB was not due to its cytoplasmic inhibitor, I-κBα, because its expression was low. The authors conclude that large amounts of the inactive form of nuclear factor κB are found in nonsurvivors of sepsis, and that the hyporeactivity in patients with severe sepsis and trauma resembles the pattern reported for endotoxin tolerance.

Because tachykinins, especially substance P acting via the neurokinin 1 receptor, mediates the adhesion of leukocytes to endothelial and epithelial cells, Kuo and coworkers (84) assessed the effect of endotoxin on this action. Neither lipopolysaccharide nor substance P alone increased the adherence of neutrophils to epithelial cells, but their combination caused significant enhancement. Adherence of neutrophils caused epithelial cells to release interleukin 1β and tumor necrosis factor α, either spontaneously or with stimulation by substance P or lipopolysaccharide. Enhanced release of interleukin 1β and tumor necrosis factor α was diminished by an antagonist of the neurokinin 1 receptor, and by prevention of neutrophil adherence with a CD11b/CD18 antibody or a serine protease inhibitor cocktail. The authors conclude that lipopolysaccharide enhances the substance P-mediated adherence of leukocytes to epithelial cells and cytokine release partly through the neurokinin 1 receptor.

Vincent and coworkers discuss the effects of nitric oxide in septic shock in a clinical commentary (85).


The incidence of ventilated-associated pneumonia in patients with ARDS varies among reports, perhaps because of the use of inaccurate diagnostic and microbial techniques. In 134 patients with ARDS and 744 patients without ARDS, Markowicz and coworkers (86) diagnosed ventilator-associated pneumonia by bronchoscopy and quantitative bacteriology. Pneumonia occurred in 37% of 134 patients with ARDS and in 23% of 744 patients without ARDS. In patients developing pneumonia, mortality was identical in those with (57%) and without ARDS (59%). Patients with pneumonia required a longer duration of mechanical ventilation. Contrary to previous studies, use of sucralfate was associated with a 4.4-fold increased risk of pneumonia. The authors conclude that development of pneumonia prolongs the duration of ventilator support without affecting mortality.


Blot and coworkers (87) compared two approaches to making an early diagnosis of ventilator-associated pneumonia: nonquantitative endotracheal aspirates and bronchoscopic catheter brushes (a Gram stain was included with both). Of 27 episodes of pneumonia diagnosed by clinicoradiologic evolution and bronchoscopic culture results, Gram stain of endotracheal aspirates had a sensitivity of 89% and specificity of 62%. Performance was considered similar for Gram stain of bronchoscopic specimens: sensitivity of 67% and specificity of 95%. Of microbiologically proven episodes of pneumonia, Gram stain of endotracheal aspirates had a sensitivity of 91% and specificity of 64%; bronchoscopic specimens had a sensitivity of 70% and specificity of 96%. The authors conclude that Gram staining of the combination of endotracheal aspirates and bronchoscopic specimens allows early diagnosis in about two-thirds of patients with ventilator-associated pneumonia.

Ruiz and coworkers (88) assessed whether the approach used in the microbial diagnosis of ventilator-associated pneumonia influenced outcome. Patients with suspected pneumonia were randomized to a bronchoscopic diagnostic strategy (37 patients) or to tracheobronchial aspiration (39 patients). Initial empiric antibiotic therapy was administered according to American Thoracic Society guidelines. Length of stay in the intensive care unit, duration of mechanical ventilation, and 30-day mortality did not differ. The authors conclude that the outcome of ventilator-associated pneumonia was not influenced by the approach to microbial investigation.


The emergence of antibiotic-resistant microorganisms has led to a need for new strategies in the use of antibiotics. To determine whether rotation and restricted use of antibiotics would decrease the incidence of ventilator-associated pneumonia, Gruson and coworkers (89) undertook a four-year prospective study of 3,455 patients admitted to a single ICU. Supervised rotation of antibiotic and restricting the use of ceftazidime and ciprofloxacin decreased episodes of pneumonia from 22 to 16%. The number of gram-negative bacilli that were potentially resistant to antibiotics also decreased from 140 to 79; of pneumonias caused by Staphylococcus aureus, however, those resistant to methicillin increased from 40 to 60%. The authors conclude that rotation and restriction of antibiotics decrease the incidence of ventilator-associated pneumonia caused by resistant bacteria.

In 171 patients with nosocomial pneumonia caused by gram-positive pathogens, Fagon and coworkers (90) did a prospective randomized comparison of quinupristin/dalfopristin (7.5 mg per kg every 8 hours) versus vancomycin (1 g every 12 hours). Between the seventh and thirteenth day, 56% of the patients receiving quinupristin/dalfopristin and 58% of those receiving vancomycin had improved clinically. Success rates were equivalent for Streptococcus pneumoniae, Staphylococcus aureus, and methicillin-resistant Staphylococcus aureus. Therapy was discontinued because of adverse effects in 15% of the quinupristin/dalfopristin group and in 10% of the vancomycin group. The authors conclude that quinupristin/dalfopristin was equivalent to vancomycin in the treatment of nosocomial pneumonia caused by gram-positive organisms.

To decrease overuse of antibiotics for treating pulmonary infiltrates in critically ill patients, Singh and coworkers (91) compared two management approaches. Patients with a low likelihood of pneumonia were randomized to standard management (choice and duration of antibiotics at physician discretion), or to ciprofloxacin alone for only three days unless patients scored a higher likelihood of pneumonia on that day. Antibiotics were continued for more than three days in 28% of 39 patients receiving protocolized care, and in 90% of 42 patients receiving usual care. In patients scoring a low likelihood of pneumonia on the third day, antibiotics were still continued in 90% of patients receiving usual care but none of those receiving protocolized care. Mortality or length of stay did not differ between the groups, but antimicrobial resistance or superinfection occurred in 35% of patients receiving usual care and 15% of patients receiving protocol care. The authors conclude that their operational criteria identify patients with pulmonary infiltrates for whom short-course monotherapy is appropriate.

To determine the effect of catheter-related infection on clinical outcome, Rello and coworkers (92) did a prospective cohort study. Of 49 cases, 31 were caused by coagulase-negative staphylococci. Hospital mortality was the same for infected patients and control subjects, but hospital stay was increased by 20 days in infected patients. Each episode of catheter- related infection represented additional costs of more than 3,000 euros. The authors conclude that catheter-related infection did not increase attributable mortality, but it did increase hospital stay and costs.

Maury and coworkers (93) investigated whether providing ICU staff with the option of rubbing their hands with alcohol solution would improve compliance of hand disinfection. When staff was offered only conventional hand washing, compliance was 42%. When the alternative of rubbing the hands with alcohol was offered, compliance rose to 61%. Nurses, senior physicians, and residents all showed improved compliance. The authors conclude that offering an alcohol solution for hand disinfection achieved better compliance than hand washing.

Karam and Heffner discuss emerging issues in antibiotic resistance in blood-borne infections in an update on nonpulmonary critical care (94).

Pressure–Volume Curves

To assess interobserver and intraobserver variation in interpreting pressure–volume curves, Harris and coworkers (95) analyzed 24 sets of curves obtained in patients with ARDS. Seven clinicians varied considerably in their calculation of the lower inflection point, differing by as much as 11 cm H2O for a given patient; the difference increased to 17 cm H2O when curves were generated with fewer data points. Use of a sigmoidal equation to fit the data markedly decreased the variability of calculations to within 2 cm H2O. The authors conclude that the usual graphical determination of the lower inflection point rarely coincides with more precise detection of the point of a rapid increase in compliance on the inflation limb of a pressure–volume curve.

The pressure–volume curve has a steep linear segment between the lower and upper inflections, and compliance is assumed to remain constant when tidal volume and PEEP are titrated to cause end-inspiratory volume to reside in this segment. Lichtwarck-Aschoff and coworkers (96) tested the validity of this assumption in 14 surfactant-depleted piglets. PEEP was increased from 3 to 24 cm H2O, and volume was increased in 10 increments up to full tidal volume. With PEEP set at the lower inflection point, compliance was minimal and it decreased over tidal volume, indicating overdistension. At intermediate levels of PEEP, compliance both increased and decreased within the same breath. The authors conclude that it is difficult to find a level of PEEP that produces a constant compliance over the full range of tidal volume; moreover, that level of PEEP does not coincide with maximal gas exchange. An editorial commentary by Dreyfuss and Saumon (97) accompanies this purpose.

Intrinsic PEEP and Respiratory Muscles

Intrinsic PEEP in patients with ARDS has been attributed to airway resistance because expiratory flow limitation has not been envisaged in this condition. Koutsoukou and coworkers (98) used the negative expiratory pressure technique to determine whether patients with ARDS exhibit airflow limitation. When semirecumbent, 8 of the 10 patients displayed airflow limitation and all had intrinsic PEEP (range, 0.4 to 7.7 cm H2O). In the supine position, intrinsic PEEP and the extent of flow limitation were further increased, although other respiratory variables did not change. Administration of albuterol had no effect. The authors conclude that most patients with ARDS exhibit airflow limitation with concomitant intrinsic PEEP.

In patients breathing spontaneously, expiratory muscle contraction causes intrinsic PEEP to be higher than the value recorded when the expiratory muscles are silent. Zakynthinos and coworkers (99) used a Campbell diagram to obtain a reference value of intrinsic PEEP and then compared three methods of correcting for expiratory muscle activity against this reference. Measuring either the increase in gastric pressure over the course of expiration or the entire decay of gastric pressure during inspiration, and subtracting either value from intrinsic PEEP, gave satisfactory results: corrected intrinsic PEEP was within 0.7 cm H2O of the reference value. In contrast, measuring the decay in gastric pressure between the onset of inspiratory effort and the point of zero flow was not satisfactory: the corrected value differed from the reference measurement by 3.1 cm H2O. The authors conclude that it is possible to correct for the contribution of expiratory muscles to intrinsic PEEP by subtracting either the increase in gastric pressure over the course of expiration or the entire decay of gastric pressure during inspiration from recorded intrinsic PEEP. An editorial commentary by Younes (100) accompanies this article.

The relative contribution of ribcage and abdominal compartments to ventilation may differ among critically ill patients. Aliverti and coworkers (101) tested the accuracy of optoelectric plethysmography for measuring lung volume and ribcage– abdominal contribution. In 11 healthy subjects and 13 ventilator-supported patients with acute lung injury, volumes were within 5% of those measured with a spirometer or pneumotachograph. The abdominal contribution to inspired volume was less in patients ventilated with pressure support than in normal subjects (43 and 63%); in patients receiving continuous positive pressure ventilation, it increased with tidal volume (61%). The authors conclude that use of optoelectric plethysmography is feasible in critically ill patients.

Hemodynamic Monitoring

Studies suggesting an increased mortality with use of a pulmonary artery catheter are confounded by the failure to take into account the influence of treatment when calculating disease severity scores. To assess the effect of pulmonary artery catheterization on outcome, Vieillard-Baron and coworkers (102) compared patients managed in two ICUs: one where catheters were used on demand and another where their use was prohibited. Mortality was higher in 35 patients monitored with a catheter (83%) than in 63 patients managed without a catheter (37%). Logistic regression analysis revealed that use of vasopressors and a nonpulmonary cause of ARDS were independently associated with death. Only when use of vasopressors was omitted from the model did pulmonary artery catheterization, septic shock, and disease severity become independent predictors of mortality. The authors conclude that it is the use of vasopressors, not the pulmonary artery catheter per se, that predicts mortality.

To clarify the rarity of right ventricular failure in patients with ARDS despite the frequent occurrence of pulmonary hypertension, Pagnamenta and coworkers (103) studied pulmonary hemodynamics in dogs with lung injury secondary to oleic acid. The animals developed an increase in pulmonary vascular resistance, but the narrowing occurred in the smallest pulmonary arteries; the latter finding combined with a decrease in vascular impedance minimize the afterload on the right ventricle. The authors conclude that right ventricular failure is uncommon with acute lung injury because the site of the increase in vascular resistance is in the smallest pulmonary arteries.

In 40 mechanically ventilated patients with acute circulatory failure related to sepsis, Michard and coworkers (104) investigated the usefulness of respiration-associated changes in arterial pressure to predict response to fluid therapy. Volume expansion produced an increase in cardiac index of at least 15% in 16 patients. Before fluid resuscitation, patients responding to volume expansion had greater respiration-associated variation in pulse pressure than nonresponders (24 vs. 7%). A respiration-associated variation in pulse pressure of 13% discriminated between responders and nonresponders (sensitivity of 94%, specificity of 96%). The changes in cardiac index and respiration-associated variation in pulse pressure with fluid therapy were correlated (r2 = 0.85). The authors conclude that measuring changes in pulse pressure associated with respiration provides a reliable means of predicting the response to volume expansion.

Gas Exchange

When healthy subjects breathe 100% oxygen during general anesthesia, they develop reabsorption atelectasis because of denitrogenation of alveolar units with low ventilation–perfusion ratios. To determine the effect of breathing 100% oxygen in patients with acute respiratory failure requiring mechanical ventilation, Santos and coworkers (105) used the multiple inert gas elimination technique in eight patients with acute lung injury and four patients with COPD. In the patients with acute lung injury, breathing 100% oxygen produced an increase in intrapulmonary shunt, from 16 to 23%, due to collapse of unstable alveolar units with low ventilation–perfusion ratios. In the patients with COPD, also receiving mechanical ventilation, oxygen caused an increase in abnormal dispersion of the blood flow distribution (from 1.3 to 1.8), suggesting release of hypoxic pulmonary vasoconstriction. The authors conclude that the change in pulmonary gas exchange as a result of breathing 100% oxygen depends on the underlying disease.

Acid–Base Disorders

In arterial blood samples from 152 patients and 9 healthy subjects, Fencl and coworkers (106) compared two widely used diagnostic approaches—plasma bicarbonate concentration with anion gap, and base excess—against a diagnostic approach based on physiochemical principles. Base excess and bicarbonate were normal in about one-sixth of the patients, but the third method detected simultaneous acidifying and alkalinizing disturbances in the vast majority of these patients. Customary measures become unreliable when concentrations of nonbicarbonate buffers (mainly albumin) are abnormal; serum albumin was more than 3 standard deviations below the mean in 96% of the samples. The authors conclude that a third method of diagnosing acid–base disturbances, based on physicochemical principles and adjustment for hypoalbuminemia, identifies components of complex acid–base disturbances missed by usual approaches.

When inhaled nitric oxide was abruptly discontinued in 31 critically ill patients, Christenson and coworkers (107) noted three response patterns: almost half showed worsening oxygenation; one-quarter developed cardiovascular collapse (a greater than 20% decrease in cardiac output); and one-quarter showed no change. The cardiovascular collapse was secondary to an acute increase in right ventricular afterload, and was more common in sicker and older patients. In all cases, reinstitution of inhaled nitric oxide promptly reversed worsened hemodynamics and gas exchange. The authors conclude that careful and monitored discontinuation of inhaled nitric oxide will identify patients at risk of complications during abrupt discontinuation.

In rabbits receiving oleic acid intravenously, Zhou and coworkers (108) studied strategies to prevent acute lung injury. Compared with pressure support alone, both inhaled nitric oxide and exogenous surfactant achieved better oxygenation and longer survival. The lung injury score was lowest in the rabbits receiving inhaled nitric oxide. The authors conclude that inhaled nitric oxide and exogenous surfactant alleviate acute lung injury.

The standard prehospital treatment of carbon monoxide poisoning is the administration of 100% oxygen. Takeuchi and coworkers (109) studied seven healthy subjects who developed carboxyhemoglobin levels of 10–12% on exposure to carbon monoxide. While breathing 100% oxygen, the half-time of decrease in carboxyhemoglobin was 78 minutes. When ventilation was increased twofold to sixfold by the addition of 4–5% carbon dioxide, the half-time fell to 31 minutes. The authors conclude that hyperpnea induced by an oxygen–carbon dioxide mixture may improve the management of carbon monoxide poisoning.

Luce and Alpers (110) discuss the legal aspects of withholding and withdrawing life support and providing palliative care in a critical care perspective.

Renal Disorders

Schortgen and coworkers (111) compared problems during intermittent hemodialysis in 45 patients managed before the implementation of practice guidelines and 76 patients managed after their implementation. The two groups were largely similar, although the latter group had a higher disease severity score. The guidelines resulted in better hemodynamic tolerance of hemodialysis, with fewer drops in arterial pressure (21 vs. 33%), decreased ICU stay, and a decrease in predicted mortality (47 vs. 66%). The authors conclude that implementing practice guidelines for intermittent hemodialysis lessens hemodynamic instability.

To investigate the effects of initial versus delayed occurrence of acute renal failure, Guerin and coworkers (112) undertook a prospective epidemiological study in 28 ICUs. Although patients with delayed renal failure had a higher mortality, it was not an independent predictor of mortality on logistic regression analysis. Instead, survival was related to disease severity on admission, number of episodes of renal failure, and to the absence of oliguria, hemodialysis, and ischemic acute tubular necrosis.

To determine the epidemiology and outcome of acute renal failure within “closed” ICUs (where input from outside consultants is minimal), Cole and coworkers (113) undertook a prospective multicenter study of ICUs in Australia. Among 116 patients with severe acute renal failure, 111 received continuous renal replacement therapy, 91 required mechanical ventilation, and 95 needed vasoactive drugs. Mortality was 49%, which compared favorably with predicted mortality of 60%. The authors conclude that the findings support the safety and efficacy of closed ICUs.

Renal replacement therapy for acute renal failure is discussed in an update in nonpulmonary critical care by Murray and Hall (114).

Gastroenterological Disorders

To determine factors associated with poor outcome in advanced liver disease, Kress and coworkers (115) performed a multivariate analysis of clinical variables in 183 patients denied a liver transplant. Mortality was 56% as opposed to 12% in patients listed for a transplant. Increased mortality was predicted by disease severity score, sepsis, and need for mechanical ventilation. Gastrointestinal bleeding was associated with a decreased mortality (OR 0.44). The authors conclude that mortality is higher in critically ill patients with advanced liver disease who are denied transplantation than those listed.

Advances in critical care gastroenterology are discussed in an update in nonpulmonary critical care by Pratt and Epstein (116).

Cardiac Disorders

Cardiac arrhythmias are discussed in an update in nonpulmonary critical care by Hammill and Hubmayr (117).

Hematological Disorders

Thrombocytopenic disorders in critically ill patients are discussed in an update on nonpulmonary care by Drews and Weinberger (118).

Supported by a Merit Review grant from the Veterans Affairs Research Service.

1. Esteban A, Anzueto A, Alia I, Gordo F, Apezteguia C, Palizas F, Cide D, Goldwaser R, Soto L, Bugedo G, Rodrigo C, Pimentel J, Raimondi G, Tobin MJHow is mechanical ventilation employed in the intensive care unit? An international utilization review. Am J Respir Crit Care Med161200014501458
2. Ricard JD, Le Miere E, Markowicz P, Lasry S, Saumon G, Djedaini K, Coste F, Dreyfuss DEfficiency and safety of mechanical ventilation with a heat and moisture exchanger changed only once a week. Am J Respir Crit Care Med1612000104109
3. Parthasarathy S, Jubran A, Tobin MJAssessment of neural inspiratory time in ventilator-supported patients. Am J Respir Crit Care Med1622000546552
4. Mehta S, Liu PP, Fitzgerald FS, Allidina YK, Douglas BTEffects of continuous positive airway pressure on cardiac volumes in patients with ischemic and dilated cardiomyopathy. Am J Respir Crit Care Med1612000128134
5. Lu Q, Capderou A, Cluzel P, Mourgeon E, Abdennour L, Law-Koune JD, Straus C, Grenier P, Zelter M, Rouby JJA computed tomographic scan assessment of endotracheal suctioning-induced bronchoconstriction in ventilated sheep. Am J Respir Crit Care Med162200018981904
6. Otis ABMy initiation into respiratory physiology. Am J Respir Crit Care Med1612000345346
7. Neumann P, Berglund JE, Andersson LG, Maripu E, Magnusson A, Hedenstierna GEffects of inverse ratio ventilation and positive end-expiratory pressure in oleic acid-induced lung injury. Am J Respir Crit Care Med161200015371545
8. Grasso S, Puntillo F, Mascia L, Ancona G, Fiore T, Bruno F, Slutsky AS, Ranieri VMCompensation for increase in respiratory workload during mechanical ventilation. Pressure-support versus proportional-assist ventilation. Am J Respir Crit Care Med1612000819826
9. Dojat M, Harf A, Touchard D, Lemaire F, Brochard LClinical evaluation of a computer-controlled pressure support mode. Am J Respir Crit Care Med161200011611166
10. Hoffman LA, Miro AM, Tasota FJ, Delgado E, Zullo TG, Lutz J, Pinsky MRTracheal gas insufflation. Limits of efficacy in adults with acute respiratory distress syndrome. Am J Respir Crit Care Med1622000387392
11. Rossi N, Musch G, Sangalli F, Verweij M, Patroniti N, Fumagalli R, Pesenti AReverse-thrust ventilation in hypercapnic patients with acute respiratory distress syndrome. Acute physiological effects. Am J Respir Crit Care Med1622000363368
12. Nam AJ, Brower RG, Fessler HE, Simon BABiologic variability in mechanical ventilation rate and tidal volume does not improve oxygenation or lung mechanics in canine oleic acid lung injury. Am J Respir Crit Care Med161200017971804
13. Mutch WA, Harms S, Ruth GM, Kowalski SE, Girling LG, Lefevre GRBiologically variable or naturally noisy mechanical ventilation recruits atelectatic lung. Am J Respir Crit Care Med1622000319323
14. Van der Kloot TE, Blanch L, Melynne YA, Weinert C, Adams AB, Marini JJ, Shapiro RS, Nahum ARecruitment maneuvers in three experimental models of acute lung injury. Effect on lung volume and gas exchange. Am J Respir Crit Care Med161200014851494
15. Feihl F, Eckert P, Brimioulle S, Jacobs O, Schaller MD, Melot C, Naeije RPermissive hypercapnia impairs pulmonary gas exchange in the acute respiratory distress syndrome. Am J Respir Crit Care Med1622000209215
16. Weber T, Tschernich H, Sitzwohl C, Ullrich R, Germann P, Zimpfer M, Sladen RN, Huemer GTromethamine buffer modifies the depressant effect of permissive hypercapnia on myocardial contractility in patients with acute respiratory distress syndrome. Am J Respir Crit Care Med162200013611365
17. Kallet RH, Jasmer RM, Luce JM, Lin LH, Marks JDThe treatment of acidosis in acute lung injury with tris-hydroxymethyl aminomethane (THAM). Am J Respir Crit Care Med161200011491153
18. Lange NR, Kozlowski JK, Gust R, Shapiro SD, Schuster DPEffect of partial liquid ventilation on pulmonary vascular permeability and edema after experimental acute lung injury. Am J Respir Crit Care Med1622000271277
19. Ricard JD, Martin-Lefevre L, Dreyfuss D, Saumon GAlveolar permeability and liquid absorption during partial liquid ventilation of rats with perflubron. Am J Respir Crit Care Med16120004449
20. Fessler HE, Pearse DAccuracy of hemodynamic measurements during partial liquid ventilation with perflubron. Am J Respir Crit Care Med162200013721376
21. Fujino Y, Goddon S, Chiche JD, Hromi J, Kacmarek RMPartial liquid ventilation ventilates better than gas ventilation. Am J Respir Crit Care Med1622000650657
22. Lopes Cardozo RH, Steendijk P, Baan J, Brouwers HA, De Vroomen M, Van Bel F. Right ventricular function in respiratory distress syndrome and subsequent partial liquid ventilation. Homeometric autoregulation in the right ventricle of the newborn animal. Am J Respir Crit Care Med 2000;162:374–379.
23. Tschumperlin DJ, Oswari J, Margulies ASDeformation-induced injury of alveolar epithelial cells. Effect of frequency, duration, and amplitude. Am J Respir Crit Care Med1622000357362
24. Laffey JG, Engelberts D, Kavanagh BPBuffering hypercapnic acidosis worsens acute lung injury. Am J Respir Crit Care Med1612000141146
25. Murphy DB, Cregg N, Tremblay L, Engelberts D, Laffey JG, Slutsky AS, Romaschin A, Kavanagh BPAdverse ventilatory strategy causes pulmonary-to-systemic translocation of endotoxin. Am J Respir Crit Care Med16220002733
26. Laffey JG, Engelberts D, Kavanagh BPInjurious effects of hypocapnic alkalosis in the isolated lung. Am J Respir Crit Care Med1622000399405
27. Laffey JG, Tanaka M, Engelberts D, Luo X, Yuan S, Tanswell AK, Post M, Lindsay T, Kavanagh BPTherapeutic hypercapnia reduces pulmonary and systemic injury following in vivo lung reperfusion. Am J Respir Crit Care Med162200022872294
28. Hickling KGLung-protective ventilation in acute respiratory distress syndrome: protection by reduced lung stress or by therapeutic hypercapnia? Am J Respir Crit Care Med162200020212022
29. Hotchkiss JR, Blanch L, Murias G, Adams AB, Olson DA, Wangensteen OD, Leo PH, Marini JJEffects of decreased respiratory frequency on ventilator-induced lung injury. Am J Respir Crit Care Med1612000463468
30. Welsh DA, Summer WR, Dobard EP, Nelson S, Mason CMKeratinocyte growth factor prevents ventilator-induced lung injury in an ex vivo rat model. Am J Respir Crit Care Med162200010811086
31. Deslee G, Brichet A, Lebuffe G, Copin MC, Ramon P, Marquette CHObstructive fibrinous tracheal pseudomembrane. A potentially fatal complication of tracheal intubation. Am J Respir Crit Care Med162200011691171
32. Maldonado A, Bauer TT, Ferrer M, Hernandez C, Arancibia F, Rodriguez-Roisin R, Torres ACapnometric recirculation gas tonometry and weaning from mechanical ventilation. Am J Respir Crit Care Med1612000171176
33. Coplin WM, Pierson DJ, Cooley KD, Newell DW, Rubenfeld GDImplications of extubation delay in brain-injured patients meeting standard weaning criteria. Am J Respir Crit Care Med161200015301536
34. Epstein SK, Nevins ML, Chung JEffect of unplanned extubation on outcome of mechanical ventilation. Am J Respir Crit Care Med161200019121916
35. Purro A, Appendini L, De Gaetano A, Gudjonsdottir M, Donner CF, Rossi APhysiologic determinants of ventilator dependence in long-term mechanically ventilated patients. Am J Respir Crit Care Med161200011151123
36. Nakos G, Tsangaris I, Kostanti E, Nathanail C, Lachana A, Koulouras V, Kastani DEffect of the prone position on patients with hydrostatic pulmonary edema compared with patients with acute respiratory distress syndrome and pulmonary fibrosis. Am J Respir Crit Care Med1612000360368
37. Albert RK, Hubmayr RDThe prone position eliminates compression of the lungs by the heart. Am J Respir Crit Care Med161200016601665
38. Malbouisson LM, Busch CJ, Puybasset L, Lu Q, Cluzel P, Rouby JJRole of the heart in the loss of aeration characterizing lower lobes in acute respiratory distress syndrome. CT Scan ARDS Study Group. Am J Respir Crit Care Med161200020052012
39. Martin TJ, Hovis JD, Costantino JP, Bierman MI, Donahoe MP, Rogers RM, Kreit JW, Sciurba FC, Stiller RA, Sanders MHA randomized, prospective evaluation of noninvasive ventilation for acute respiratory failure. Am J Respir Crit Care Med1612000807813
40. Jaber S, Fodil R, Carlucci A, Boussarsar M, Pigeot J, Lemaire F, Harf A, Lofaso F, Isabey D, Brochard LNoninvasive ventilation with helium–oxygen in acute exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med161200011911200
41. Garrod R, Mikelsons C, Paul EA, Wedzicha JARandomized controlled trial of domiciliary noninvasive positive pressure ventilation and physical training in severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med162200013351341
42. Hukins CA, Hillman DRDaytime predictors of sleep hypoventilation in Duchenne muscular dystrophy. Am J Respir Crit Care Med1612000166170
43. Rossi ANoninvasive ventilation has not been shown to be ineffective in stable COPD. Am J Respir Crit Care Med1612000688689
44. Hill NSNoninvasive ventilation has been shown to be ineffective in stable COPD. Am J Respir Crit Care Med1612000689690
45. Rossi ARebuttal. Am J Respir Crit Care Med1612000691
46. Hill NSRebuttal. Am J Respir Crit Care Med1612000691
47. Meade MO, Cook RJ, Guyatt GH, Groll R, Kachura JR, Bedard M, Cook DJ, Slutsky AS, Stewart TEInterobserver variation in interpreting chest radiographs for the diagnosis of acute respiratory distress syndrome. Am J Respir Crit Care Med16120008590
48. Newman V, Gonzalez RF, Matthay MA, Dobbs LGA novel alveolar Type I cell-specific biochemical marker of human acute lung injury. Am J Respir Crit Care Med1612000990995
49. Kawabata K, Hagio T, Matsumoto S, Nakao S, Orita S, Aze Y, Ohno HDelayed neutrophil elastase inhibition prevents subsequent progression of acute lung injury induced by endotoxin inhalation in hamsters. Am J Respir Crit Care Med161200020132018
50. Yamada T, Hisanaga M, Nakajima Y, Mizuno S, Matsumoto K, Nakamura T, Nakano HEnhanced expression of hepatocyte growth factor by pulmonary ischemia–reperfusion injury in the rat. Am J Respir Crit Care Med1622000707715
51. Yamada H, Miyazaki H, Kikuchi T, Fujimoto J, Kudoh IAcid instillation enhances the inflammatory response to subsequent lipopolysaccharide challenge in rats. Am J Respir Crit Care Med162200013661371
52. Nielsen VG, Baird MS, Chen L, Matalon SDETANONOate, a nitric oxide donor, decreases amiloride-sensitive alveolar fluid clearance in rabbits. Am J Respir Crit Care Med161200011541160
53. Marshall RP, Bellingan G, Webb S, Puddicombe A, Goldsack N, McAnulty RJ, Laurent GJFibroproliferation occurs early in the acute respiratory distress syndrome and impacts on outcome. Am J Respir Crit Care Med162200017831788
54. Hashimoto S, Kobayashi A, Kooguchi K, Kitamura Y, Onodera H, Nakajima HUpregulation of two death pathways of perforin/granzyme and FasL/Fas in septic acute respiratory distress syndrome. Am J Respir Crit Care Med1612000237243
55. Durand P, Bachelet M, Brunet F, Richard MJ, Dhainaut JF, Dall'Ava J, Polla BSInducibility of the 70 kD heat shock protein in peripheral blood monocytes is decreased in human acute respiratory distress syndrome and recovers over time. Am J Respir Crit Care Med1612000286292
56. Gunther A, Mosavi P, Heinemann S, Ruppert C, Muth H, Markart P, Grimminger F, Walmrath D, Temmesfeld-Wollbruck B, Seeger WAlveolar fibrin formation caused by enhanced procoagulant and depressed fibrinolytic capacities in severe pneumonia. Comparison with the acute respiratory distress syndrome. Am J Respir Crit Care Med1612000454462
57. Erukhimov JA, Tang ZL, Johnson BA, Donahoe MP, Razzack JA, Gibson KF, Lee WM, Wasserloos KJ, Watkins SA, Pitt BRActin-containing sera from patients with adult respiratory distress syndrome are toxic to sheep pulmonary endothelial cells. Am J Respir Crit Care Med1622000288294
58. Cantin AM, Paquette B, Richter M, Larivee PAlbumin-mediated regulation of cellular glutathione and nuclear factor kappa B activation. Am J Respir Crit Care Med162200015391546
59. Moss M, Guidot DM, Wong-Lambertina M, Ten Hoor T, Perez RL, Brown LA. The effects of chronic alcohol abuse on pulmonary glutathione homeostasis. Am J Respir Crit Care Med 2000;161:414–419.
60. Nakamura M, Fujishima S, Sawafuji M, Ishizaka A, Oguma T, Soejima K, Matsubara H, Tasaka S, Kikuchi K, Kobayashi K, Ikeda E, Sadick M, Hebert CA, Aikawa N, Kanazawa M, Yamaguchi KImportance of interleukin-8 in the development of reexpansion lung injury in rabbits. Am J Respir Crit Care Med161200010301036
61. Chen BT, Yeates DBDifferentiation of ion-associated and osmotically driven water transport in canine airways. Am J Respir Crit Care Med162200017151722
62. Olivera WG, Ciccolella DE, Barquin N, Ridge KM, Rutschman DH, Yeates DB, Sznajder JIAldosterone regulates Na,K-ATPase and increases lung edema clearance in rats. Am J Respir Crit Care Med1612000567573
63. Saldias FJ, Lecuona E, Comellas AP, Ridge KM, Rutschman DH, Sznajder JIBeta-adrenergic stimulation restores rat lung ability to clear edema in ventilator-associated lung injury. Am J Respir Crit Care Med1622000282287
64. Kuebler WM, Borges J, Sckell A, Kuhnle GE, Bergh K, Messmer K, Goetz AERole of L-selectin in leukocyte sequestration in lung capillaries in a rabbit model of endotoxemia. Am J Respir Crit Care Med16120003643
65. Fujii Y, Magder S, Cernacek P, Goldberg P, Guo Y, Hussain SNEndothelin receptor blockade attenuates lipopolysaccharide-induced pulmonary nitric oxide production. Am J Respir Crit Care Med1612000982989
66. Afulukwe IF, Cohen RI, Zeballos GA, Iqbal M, Scharf SMSelective NOS inhibition restores myocardial contractility in endotoxemic rats; however, myocardial NO content does not correlate with myocardial dysfunction. Am J Respir Crit Care Med16220002126
67. Giraud O, Seince PF, Rolland C, Lecon-Malas V, Desmonts JM, Aubier M, Dehoux MHalothane reduces the early lipopolysaccharide-induced lung inflammation in mechanically ventilated rats. Am J Respir Crit Care Med162200022782286
68. Arsalane K, Broeckaert F, Knoops B, Wiedig M, Toubeau G, Bernard AClara cell specific protein (CC16) expression after acute lung inflammation induced by intratracheal lipopolysaccharide administration. Am J Respir Crit Care Med161200016241630
69. Evgenov OV, Hevroy O, Bremnes KE, Bjertnaes LJEffect of aminoguanidine on lung fluid filtration after endotoxin in awake sheep. Am J Respir Crit Care Med1622000465470
70. Held HD, Uhlig SMechanisms of endotoxin-induced airway and pulmonary vascular hyperreactivity in mice. Am J Respir Crit Care Med162200015471552
71. Enkhbaatar P, Okajima K, Murakami K, Uchiba M, Okabe H, Okabe K, Yamaguchi YRecombinant tissue factor pathway inhibitor reduces lipopolysaccharide-induced pulmonary vascular injury by inhibiting leukocyte activation. Am J Respir Crit Care Med162200017521759
72. Kiefer P, Tugtekin I, Wiedeck H, Bracht H, Geldner G, Georgieff M, Radermacher PEffect of a dopexamine-induced increase in cardiac index on splanchnic hemodynamics in septic shock. Am J Respir Crit Care Med1612000775779
73. Gardinali M, Borrelli E, Chiara O, Lundberg C, Padalino P, Conciato L, Cafaro C, Lazzi S, Luzi P, Giomarelli PP, Agostoni AInhibition of CD11–CD18 complex prevents acute lung injury and reduces mortality after peritonitis in rabbits. Am J Respir Crit Care Med161200010221029
74. Pinsky MR, Rico PCardiac contractility is not depressed in early canine endotoxic shock. Am J Respir Crit Care Med161200010871093
75. Bellomo R, Kellum JA, Gandhi CR, Pinsky MR, Ondulik BThe effect of intensive plasma water exchange by hemofiltration on hemodynamics and soluble mediators in canine endotoxemia. Am J Respir Crit Care Med161200014291436
76. Okamoto I, Abe M, Shibata K, Shimizu N, Sakata N, Katsuragi T, Tanaka KEvaluating the role of inducible nitric oxide synthase using a novel and selective inducible nitric oxide synthase inhibitor in septic lung injury produced by cecal ligation and puncture. Am J Respir Crit Care Med1622000716722
77. Blackwell TS, Yull FE, Chen CL, Venkatakrishnan A, Blackwell TR, Hicks DJ, Lancaster LH, Christman JW, Kerr LDMultiorgan nuclear factor kappa B activation in a transgenic mouse model of systemic inflammation. Am J Respir Crit Care Med162200010951101
78. Fujimura N, Sumita S, Aimono M, Masuda Y, Shichinohe Y, Narimatsu E, Namiki AEffect of free radical scavengers on diaphragmatic contractility in septic peritonitis. Am J Respir Crit Care Med162200021592165
79. Fujimura N, Sumita S, Narimatsu E, Nakayama Y, Shitinohe Y, Namiki AEffects of isoproterenol on diaphragmatic contractility in septic peritonitis. Am J Respir Crit Care Med1612000440446
80. Lanone S, Mebazaa A, Heymes C, Henin D, Poderoso JJ, Panis Y, Zedda C, Billiar T, Payen D, Aubier M, Boczkowski JMuscular contractile failure in septic patients:role of the inducible nitric oxide synthase pathway. Am J Respir Crit Care Med162200023082315
81. Chrusch C, Bands C, Bose D, Li X, Jacobs H, Duke K, Bautista E, Eschun G, Light RB, Mink SNImpaired hepatic extraction and increased splanchnic production contribute to lactic acidosis in canine sepsis. Am J Respir Crit Care Med1612000517526
82. Kirschenbaum LA, Aziz M, Astiz ME, Saha DC, Rackow ECInfluence of rheologic changes and platelet-neutrophil interactions on cell filtration in sepsis. Am J Respir Crit Care Med161200016021607
83. Adib-Conquy M, Adrie C, Moine P, Asehnoune K, Fitting C, Pinsky MR, Dhainaut JF, Cavaillon JMNF-kappaB expression in mononuclear cells of patients with sepsis resembles that observed in lipopolysaccharide tolerance. Am J Respir Crit Care Med162200018771883
84. Kuo HP, Lin HC, Hwang KH, Wang CH, Lu LCLipopolysaccharide enhances substance P-mediated neutrophil adherence to epithelial cells and cytokine release. Am J Respir Crit Care Med162200018911897
85. Vincent JL, Zhang H, Szabo C, Preiser JCEffects of nitric oxide in septic shock. Am J Respir Crit Care Med161200017811785
86. Markowicz P, Wolff M, Djedaini K, Cohen Y, Chastre J, Delclaux C, Merrer J, Herman B, Veber B, Fontaine A, Dreyfuss DMulticenter prospective study of ventilator-associated pneumonia during acute respiratory distress syndrome. Incidence, prognosis, and risk factors. ARDS Study Group. Am J Respir Crit Care Med161200019421948
87. Blot F, Raynard B, Chachaty E, Tancrede C, Antoun S, Nitenberg GValue of Gram stain examination of lower respiratory tract secretions for early diagnosis of nosocomial pneumonia. Am J Respir Crit Care Med162200017311737
88. Ruiz M, Torres A, Ewig S, Marcos MA, Alcon A, Lledo R, Asenjo MA, Maldonaldo ANoninvasive versus invasive microbial investigation in ventilator-associated pneumonia: evaluation of outcome. Am J Respir Crit Care Med1622000119125
89. Gruson D, Hilbert G, Vargas F, Valentino R, Bebear C, Allery A, Bebear C, Gbikpi-Benissan G, Cardinaud JPRotation and restricted use of antibiotics in a medical intensive care unit. Impact on the incidence of ventilator-associated pneumonia caused by antibiotic-resistant gram-negative bacteria. Am J Respir Crit Care Med1622000837843
90. Fagon J, Patrick H, Haas DW, Torres A, Gibert C, Cheadle WG, Falcone RE, Anholm JD, Paganin F, Fabian TC, Lilienthal FTreatment of gram-positive nosocomial pneumonia. Prospective randomized comparison of quinupristin/dalfopristin versus vancomycin. Nosocomial Pneumonia Group. Am J Respir Crit Care Med1612000753762
91. Singh N, Rogers P, Atwood CW, Wagener MM, Yu VLShort-course empiric antibiotic therapy for patients with pulmonary infiltrates in the intensive care unit. A proposed solution for indiscriminate antibiotic prescription. Am J Respir Crit Care Med1622000505511
92. Rello J, Ochagavia A, Sabanes E, Roque M, Mariscal D, Reynaga E, Valles JEvaluation of outcome of intravenous catheter-related infections in critically ill patients. Am J Respir Crit Care Med162200010271030
93. Maury E, Alzieu M, Baudel JL, Haram N, Barbut F, Guidet B, Offenstadt GAvailability of an alcohol solution can improve hand disinfection compliance in an intensive care unit. Am J Respir Crit Care Med1622000324327
94. Karam GH, Heffner JEEmerging issues in antibiotic resistance in blood-borne infections. Am J Respir Crit Care Med162200016101616
95. Harris RS, Hess DR, Venegas JGAn objective analysis of the pressure–volume curve in the acute respiratory distress syndrome. Am J Respir Crit Care Med1612000432439
96. Lichtwarck-Aschoff M, Mols G, Hedlund AJ, Kessler V, Markstrom AM, Guttmann J, Hedenstierna G, Sjostrand UHCompliance is nonlinear over tidal volume irrespective of positive end-expiratory pressure level in surfactant-depleted piglets. Am J Respir Crit Care Med162200021252133
97. Dreyfuss D, Saumon GPressure–volume curves: searching for the Grail or laying patients with adult respiratory distress syndrome on Procrustes' bed? Am J Respir Crit Care Med163200123
98. Koutsoukou A, Armaganidis A, Stavrakaki-Kallergi C, Vassilakopoulos T, Lymberis A, Roussos C, Milic-Emili JExpiratory flow limitation and intrinsic positive end-expiratory pressure at zero positive end-expiratory pressure in patients with adult respiratory distress syndrome. Am J Respir Crit Care Med161200015901596
99. Zakynthinos SG, Vassilakopoulos T, Zakynthinos E, Mavrommatis A, Roussos CContribution of expiratory muscle pressure to dynamic intrinsic positive end-expiratory pressure: validation using the Campbell diagram. Am J Respir Crit Care Med162200016331640
100. Younes MDynamic intrinsic PEEP (PEEPi,dyn): is it worth saving? Am J Respir Crit Care Med162200016081609
101. Aliverti A, Dellaca R, Pelosi P, Chiumello D, Pedotti A, Gattinoni LOptoelectronic plethysmography in intensive care patients. Am J Respir Crit Care Med161200015461552
102. Vieillard-Baron A, Girou E, Valente E, Brun-Buisson C, Jardin F, Lemaire F, Brochard LPredictors of mortality in acute respiratory distress syndrome. Focus on the role of right heart catheterization. Am J Respir Crit Care Med161200015971601
103. Pagnamenta A, Bouckaert Y, Wauthy P, Brimioulle S, Naeije RContinuous versus pulsatile pulmonary hemodynamics in canine oleic acid lung injury. Am J Respir Crit Care Med1622000936940
104. Michard F, Boussat S, Chemla D, Anguel N, Mercat A, Lecarpentier Y, Richard C, Pinsky MR, Teboul JLRelation between respiratory changes in arterial pulse pressure and fluid responsiveness in septic patients with acute circulatory failure. Am J Respir Crit Care Med1622000134138
105. Santos C, Ferrer M, Roca J, Torres A, Hernandez C, Rodriguez-Roisin RPulmonary gas exchange response to oxygen breathing in acute lung injury. Am J Respir Crit Care Med16120002631
106. Fencl V, Jabor A, Kazda A, Figge JDiagnosis of metabolic acid–base disturbances in critically ill patients. Am J Respir Crit Care Med162200022462251
107. Christenson J, Lavoie A, O'Connor M, Bhorade S, Pohlman A, Hall JBThe incidence and pathogenesis of cardiopulmonary deterioration after abrupt withdrawal of inhaled nitric oxide. Am J Respir Crit Care Med161200014431449
108. Zhou ZH, Sun B, Lin K, Zhu LWPrevention of rabbit acute lung injury by surfactant, inhaled nitric oxide, and pressure support ventilation. Am J Respir Crit Care Med1612000581588
109. Takeuchi A, Vesely A, Rucker J, Sommer LZ, Tesler J, Lavine E, Slutsky AS, Maleck WH, Volgyesi G, Fedorko L, Iscoe S, Fisher JAA simple “new” method to accelerate clearance of carbon monoxide. Am J Respir Crit Care Med161200018161819
110. Luce JM, Alpers ALegal aspects of withholding and withdrawing life support from critically ill patients in the United States and providing palliative care to them. Am J Respir Crit Care Med162200020292032
111. Schortgen F, Soubrier N, Delclaux C, Thuong M, Girou E, Brun-Buisson C, Lemaire F, Brochard LHemodynamic tolerance of intermittent hemodialysis in critically ill patients: usefulness of practice guidelines. Am J Respir Crit Care Med1622000197202
112. Guerin C, Girard R, Selli JM, Perdrix JP, Ayzac LInitial versus delayed acute renal failure in the intensive care unit. A multicenter prospective epidemiological study. Rhone-Alpes Area Study Group on Acute Renal Failure. Am J Respir Crit Care Med1612000872879
113. Cole L, Bellomo R, Silvester W, Reeves JHA prospective, multicenter study of the epidemiology, management, and outcome of severe acute renal failure in a “closed” ICU system. Am J Respir Crit Care Med1622000191196
114. Murray P, Hall JRenal replacement therapy for acute renal failure. Am J Respir Crit Care Med1622000777781
115. Kress JP, Rubin A, Pohlman AS, Hall JBOutcomes of critically ill patients denied consideration for liver transplantation. Am J Respir Crit Care Med1622000418423
116. Pratt DS, Epstein SKRecent advances in critical care gastroenterology. Am J Respir Crit Care Med161200014171420
117. Hammill SC, Hubmayr RDThe rapidly changing management of cardiac arrhythmias. Am J Respir Crit Care Med161200010701073
118. Drews RE, Weinberger SEThrombocytopenic disorders in critically ill patients. Am J Respir Crit Care Med1622000347351
Correspondence and requests for reprints should be addressed to Martin J. Tobin, M.D., Division of Pulmonary and Critical Care Medicine, Room 438E, Building 1, Hines Veterans Affairs Hospital, 5th Avenue & Roosevelt Road, Hines, IL 60141. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record