Ventilatory Management of Acute Lung Injury and Acute Respiratory Distress Syndrome

CLINICAL REVIEW CLINICIAN’S CORNER

Eddy Fan, MD Context The acute lung injury and acute respiratory distress syndrome are critical Dale M. Needham, MD, PhD illnesses associated with significant morbidity and mortality. Mechanical ventilation is Thomas E. Stewart, MD the cornerstone of supportive therapy. However, despite several important advances, the optimal strategy for ventilation and adjunctive therapies for patients with acute OR NEARLY 4 DECADES SINCE THE lung injury and acute respiratory distress syndrome is still evolving. acute respiratory distress syn- Evidence Acquisition To identify reports of invasive ventilatory and adjunctive thera- drome (ARDS) was first de- pies in adult patients with acute lung injury and acute respiratory distress syndrome, scribed,1 research has been on- we performed a systematic English-language literature search of MEDLINE (1966- goingF in an effort to improve the 2005) using the Medical Subject Heading respiratory distress syndrome, adult, and outcome of this critical illness. Acute related text words, with emphasis on randomized controlled trials and meta-analyses. EMBASE and the Cochrane Central Register of Controlled Trials were similarly searched. respiratory distress syndrome is char- The search yielded 1357 potential articles of which 53 were relevant to the study ob- acterized by the acute onset of hypox- jectives and considered in this review. emia and bilateral infiltrates on chest Evidence Synthesis There is strong evidence to support the use of volume- and radiography in the absence of left atrial pressure-limited lung-protective ventilation in adult patients with acute lung injury and hypertension. Various pulmonary (eg, acute respiratory distress syndrome. The benefit of increased levels of positive end- pneumonia) and nonpulmonary (eg, expiratory pressure and recruitment maneuvers is uncertain and is being further evalu- pancreatitis) risk factors are associated in ongoing trials. Existing randomized controlled trials of alternative ventilation ated with ARDS.2 Mortality rates range modes, such as high-frequency oscillation and adjunctive therapies, including inhaled from 26% to 74%, with most deaths at- nitric oxide and prone positioning demonstrate no significant survival advantage. How- tributed to associated conditions, such ever, they may have a role as rescue therapy for patients with acute respiratory dis- as sepsis and multisystem organ fail- tress syndrome with refractory life-threatening hypoxemia. ure, rather than hypoxemia alone.2-6 Conclusions Volume- and pressure-limited ventilation strategies should be used in Some survivors of ARDS have reduced managing adult acute lung injury and acute respiratory distress syndrome patients. quality of life with physical, neurocog- Further research is needed to identify barriers to widespread adoption of this strategy, nitive, and emotional morbidity.7-10 as well as the role of alternative ventilation modes and adjunctive therapies. The original ARDS case series1 out- JAMA. 2005;294:2889-2896 www.jama.com lined a number of clinical features that Ͻ were later incorporated into more for- 300), while ARDS represents the sub- though necessary to preserve life, can po- mal definitions of this syndrome set of ALI patients with the most se- tentiate or directly injure the lungs Ͻ (TABLE 1).11-13 In 1994, the American- vere lung injury (PaO2/FIO2 ratio 200). through a variety of mechanisms col- European Consensus Conference Using these definitions, ALI and ARDS (AECC) definition was developed and are relatively common with annual in- Author Affiliations: Interdepartmental Division of Criti- is used widely by clinicians and re- cidences estimated at 20 to 50 and 15 cal Care Medicine and Department of Medicine, Uni- versity of Toronto and University Health Network and 13 searchers. Under this definition, acute to 30 cases per 100 000 persons, re- Mount Sinai Hospital, Toronto, Ontario, (Drs Fan and 2,3 lung injury (ALI) is designated for pa- spectively. Stewart) and Division of Pulmonary and Critical Care No specific pharmacologic therapy Medicine, Johns Hopkins University, Baltimore, Md (Dr tients with significant hypoxemia (par- Needham). tial pressure of arterial oxygen to frac- has proved effective for ALI or ARDS, Corresponding Author: Thomas E. Stewart, MD, Uni- and therapy is largely supportive with versity Health Network/Mount Sinai Hospital, 600 Uni- tion of inspired oxygen [PaO2/FIO2] ratio 11 versity Ave, Suite 18-206, Toronto, Ontario, Canada the use of mechanical ventilation. Per- M5G 1X5 ([email protected]). haps the most important advance in ALI Clinical Review Section Editor: Michael S. Lauer, MD. CME available online at and ARDS research has been the recog- We encourage authors to submit papers for consid- www.jama.com eration as a “Clinical Review.” Please contact Mi- nition that mechanical ventilation, al- chael S. Lauer, MD, at [email protected].

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS lectively referred to as ventilator- tomography studies (FIGURE).20 As a re- quently, mechanical ventilation can re- associated lung injury.14-16 These mecha- sult, some areas of the lung (often de- sult in barotrauma or volutrauma when nisms include exposure to high inflation pendent regions) are atelectatic, con- volumes and pressures meant for the pressures or overdistention (baro- solidated, less compliant, and thus less entire lung are forced into only a small trauma or volutrauma),17 repetitive op- available for ventilation while other portion of functional lung. In addi- ening and closing of alveoli (atelec- areas (usually nondependent regions) tion, shear forces at the interface be- trauma),18 and mechanotransduction appear and behave normally. Under- tween the open and closed lung units resulting in up-regulated cytokine re- standing this heterogeneity has led to result in atelectrauma. Both of these lease and a systemic inflammatory re- the “baby lung” concept, which sug- types of injury also can lead to release sponse (biotrauma).19 The lungs of pa- gests that, overall, a markedly re- of cytokines from the lung and have ad- tients with ALI or ARDS are particularly duced volume of lung is available for verse systemic effects, contributing to prone to ventilator-associated lung in- ventilation in ALI or ARDS, effec- the development of multisystem or- jury because they are heterogeneously tively, a functionally baby-sized lung gan failure.18,19 affected, as demonstrated in computed within an adult-sized body.21,22 Conse- This improved understanding of ALI and ARDS and ventilator-associated Table 1. Diagnostic Criteria for ARDS lung injury has been important in de- Source Oxygenation Chest Radiograph Other Criteria signing lung protective mechanical ven- Petty and Cyanosis refractory Diffuse alveolar Impaired pulmonary tilation strategies aimed at attenuat- Ashbaugh,11 to oxygen therapy infiltrates on frontal compliance ing ventilator-associated lung injury and 1971 chest radiograph Marked difference in inspired vs arterial improving outcomes. Such strategies for oxygen tensions the invasive ventilatory management of 12 Murray et al, Hypoxemia (PaO2/FIO2), No. of quadrants PEEP and respiratory adult ALI and ARDS have recently been 1988 by quintiles of alveolar system compliance consolidation (by quintiles) tested in a number of important clini- on frontal chest Preexisting direct or cal trials, which we review in this ar- radiograph indirect lung injury ticle. In addition, we discuss alterna- Nonpulmonary organ dysfunction tive invasive ventilatory modes and 13 Bernard et al, ALI, PaO2/FIO2 Յ300, Bilateral infiltrates on PCWP Յ18 mm Hg adjunctive therapies, with a focus on 1994 regardless of PEEP level frontal chest if measured or those that we believe are widely avail- ARDS, PaO2/FIO2 Յ200, radiography no clinical evidence regardless of PEEP level of left atrial able for clinical use in adults at the hypertension present time. We also highlight recent Abbreviations: ALI, acute lung injury; ARDS, acute respiratory distress syndrome; PaO2/FIO2, ratio of partial pressure of controversies and suggest areas for arterial oxygen to fraction of inspired oxygen; PCWP, pulmonary capillary wedge pressure; PEEP, positive end- expiratory pressure. future research.

Figure. Typical Chest Radiograph and Computed Tomographic Scan of a Patient With Acute Respiratory Distress Syndrome

A B

A, The chest radiograph shows bilateral pulmonary infiltrates that appear to be diffuse. B, A computed tomographic scan of the thorax from the same patient dem- onstrates that the distribution of the bilateral infiltrates is predominantly in dependent regions with more normal-appearing lung in nondependent regions.

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS

EVIDENCE ACQUISITION corporeal membrane oxygenation). A charge, and a high mortality rate (71%) To assist with this review, we system- total of 53 articles met our criteria and in the control group may have ac- atically searched MEDLINE (1966 to were considered in this review. counted for the survival difference. Nev- September 2005), using the Medical ertheless, this trial strongly suggested Subject Heading respiratory distress syn- EVIDENCE SYNTHESIS that ventilatory strategies could im- drome, adult, and related text words Conventional Lung pact mortality, and the results in the in- (acute respiratory distress syndrome, Protective-Ventilation tervention group generated interest- acute lung injury, respiratory distress syn- Conventional lung-protective ventila- ing hypotheses that have been pursued drome, adult respiratory distress syn- tion involves strategies designed to miti- in subsequent studies. The largest trial drome, ALI,orARDS). The search re- gate further lung injury in patients with of volume- and pressure-limited ven- sults were limited to English language, ALI or ARDS using a standard mechani- tilation was conducted by the ARDS human subjects, randomized con- cal ventilator. Five randomized con- Network (ARDSNet).27 This trial of 861 trolled trials, and meta-analyses. trolled trials23-27 and 3 meta-analy- patients demonstrated a 9% absolute EMBASE and the Cochrane Central ses28-30 have evaluated lung-protective decrease in mortality (31% vs 40%; Register of Controlled Trials were also ventilation compared with conven- P=.007) when patients with ALI or similarly searched. This strategy re- tional approaches, using a variety of vol- ARDS receiving mechanical ventila- sulted in 1357 potentially relevant ar- ume- and pressure-limited strategies tion have reduced tidal volumes (tar- ticles for which citations, abstracts, or (TABLE 2). Three of the randomized get of 6 mL/kg of predicted body weight both were reviewed. Bibliographies of trials, with sample sizes of 52 to 120 with a range of 4-8 mL/kg depending all selected articles and review articles patients, did not find a difference in on plateau pressure and pH) and re- were hand searched for additional rel- mortality between the treatment and duced pressures (plateau pressure, mea- evant articles. Studies evaluating non- control arms.24-26 One study by Amato sured after a 0.5-second end-inspira- 23 Յ invasive ventilation, pharmacologic et al used higher positive end- tory pause, 30 cm H2O). therapies (not part of a mechanical ven- expiratory pressure (PEEP) and recruit- Three meta-analyses28-30 of these 5 tilation strategy), and small, explor- ment maneuvers in conjunction with clinical trials have been performed.23-27 atory randomized trials (Ͻ50 pa- pressure- and volume-limited ventila- The first meta-analysis concluded that tients) were excluded. To provide a tion in the intervention group. This the control groups of the 2 trials, which focused and pragmatic review, we also study was stopped early, enrolling 53 demonstrated a survival advan- excluded studies of therapies that we of 58 patients, after demonstrating a sig- tage,23,27 did not reflect the “standard believed were not widely available for nificant reduction in 28-day mortal- of care” and were likely responsible for clinical use in adults at the present time ity. However, there was no significant the mortality difference.28 In addition, (eg, exogenous surfactant and extra- difference in mortality at hospital dis- the authors suggested that the low tidal

Table 2. Trials of Volume- and Pressure-Limited Ventilation in Acute Lung Injury and Acute Respiratory Distress Syndrome Source

ARDS Network,27 Amato et al,23 Brochard et al,24 Stewart et al,25 Brower et al,26 Study Participants 2000 1998 1998 1998 1999 No. 861 53 116 120 52 Mean age, y 52 35 57 59 49 Target intervention Tidal volume, mL/kg 6 vs 12 PBW Յ6 vs 12 ABW 6-10 vs 10-15 DBW Յ8 vs 10-15 IBW Յ8 vs 10-12 PBW Plateau pressure, Յ30 vs Յ50 Ͻ20 vs unlimited 25-30 vs Յ60 Յ30 vs Յ50* Յ30 vs Յ45-55 cm H2O Actual intervention† Tidal volume, mL/kg 6.2 vs 11.8 384 vs 768‡ 7.1 vs 10.3 7.0 vs 10.7 7.3 vs 10.2 Plateau pressure, 25 vs 33 30 vs 37 26 vs 32 22 vs 27 25 vs 31 cm H2O Outcomes mortality, % 31 vs 40§ 38 vs 71 ࿣ 47 vs 38¶ 50 vs 47# 50 vs 46# P value .007 .001 .38 .72 .61 Abbreviations: ABW, actual body weight; DBW, dry body weight; IBW, ideal body weight, calculated as 25 m2; PBW, predicted body weight, calculated as 50 plus 0.91 (height in centimeters minus 152.4) for men or 45.5 plus 0.91 (height in centimeters minus 152.4) for women. *Represents peak inspiratory pressure rather than plateau pressure. †Mean values at earliest recorded time point are reported and represent lung-protective vs control ventilation groups. ‡Tidal volume available in milliliters only. §Mortality at hospital discharge or at 180 days. ||28-Day mortality. ¶60-Day mortality. #In-hospital mortality.

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS volumes used in the intervention group ences in sedation or analgesia use be- on oxygenation. After evaluating the of the ARDSNet trial may be harm- tween lung-protective and conven- first 80 patients, recruitment maneu- ful.28 This meta-analysis has been criti- tional mechanical ventilation groups.33 vers demonstrated only a modest cized as having important methodologi- However, the effect of any potential in- effect on oxygenation with no differ- cal flaws, such as inappropriately crease in sedation or analgesia use in ence in the requirement for oxygen- grouping the individual trial re- patients receiving lung-protective ven- ation support (ie, PEEP or FIO2) and sults.29-31 In addition, its findings are tilation requires further study. were not continued for the remainder contradicted by 2 subsequent meta- In addition to pressure and volume of the study.39 The study was stopped analyses,29,30 which suggested that vol- limitation, higher PEEP and the use of early, due to futility, after enrollment ume-limited ventilation, particularly in recruitment maneuvers may be impor- of 549 of 750 patients. There was no the setting of elevated plateau pres- tant components of a lung-protective significant difference in in-hospital Ͼ sure ( 30 cm H2O), has a short-term ventilation strategy. Recruitment re- mortality, even after adjusting for survival benefit. One meta-analysis also fers to the dynamic process of reopen- important imbalances in baseline char- concluded that decreased tidal vol- ing collapsed alveoli through an inten- acteristics between the study groups umes may be advantageous below a tional increase in transpulmonary (higher, 25.1% vs lower, 27.5% PEEP; threshold level (Ͻ7.7 mL/kg pre- pressure, which may be achieved 95% confidence interval [CI], −3.6% dicted body weight).30 through a variety of mechanisms, such to 8.4%; P=.47). However, some have In many instances, lung-protective as a transient high continuous posi- suggested that a true effect may have ventilation may lead to an elevation of tive airway pressure (eg, 40 cm H2O for been missed due to stopping the study arterial carbon dioxide, referred to as 40 seconds).34 Human studies evaluat- early.45 Alternatively, a beneficial effect permissive hypercapnia. Although acute ing recruitment maneuvers have yielded of higher PEEP in some (recruitable) hypercapnic respiratory acidosis has variable results.35-39 Such factors as the patients may have been negated by a many potential adverse effects, the ex- type (eg, primary vs secondary) and detrimental effect in other (nonre- tent to which a more controlled sub- stage (eg, early vs late) of ALI or ARDS cruitable) patients. These issues acute elevation of carbon dioxide is and recruitment technique used may require further investigation. harmful remains uncertain.32 In fact, be important determinants of re- some evidence indicates that permis- sponse.40-42 The optimal pressure, du- Alternative Ventilatory Approaches sive hypercapnia is relatively be- ration, and frequency of recruitment to Lung Protection nign.32 At present, there are few data to maneuvers has not been defined and The precise role of alternative methods inform physicians on a clinically rel- tested in clinical trials. Of note, the of ventilation, such as high-frequency evant threshold of hypercapnia, acido- safety of recruitment maneuvers also re- ventilation (ie, jet, oscillation, and per- sis, or both that require specific inter- quires careful evaluation. Although cussive ventilation) and airway pres- ventions, such as increasing effective transient oxygen desaturation and hy- sure release ventilation, has not been es- ventilation, decreasing carbon diox- potension are the most common ad- tablished. High-frequency ventilation ide production, or using buffer therapy verse effects, other clinically signifi- allows for higher mean airway pres- (eg, bicarbonate). Of note, the ARDSNet cant events, such as barotrauma (eg, sures that may be advantageous for lung study27 investigators used a stepwise ap- pneumothorax), arrhythmia, and bac- recruitment. In addition, it also allows proach of increasing respiratory rates terial translocation, may occur.38,43 for markedly reduced tidal volumes (1-3 (to a maximum of 35 breaths/min), al- The isolated effect of higher PEEP mL/kg) compared with conventional lowing bicarbonate infusions (at the and recruitment maneuvers could not ventilation, which may further reduce physicians’ discretion), and increas- be determined in the small trial by ventilator-associated lung injury.46,47 ing tidal volumes for the management Amato et al.23 A much larger study, Airway pressure release ventilation not of acidosis. known as the ALVEOLI trial,44 was only provides higher mean airway pres- Other important consequences of conducted by ARDSNet to investigate sures but also allows for spontaneous lung-protective ventilation possibly in- this issue. This study provided breathing, which may be associated clude worsened oxygenation and the volume- and pressure-limited lung- with better gas exchange, hemody- need for increased sedation or analge- protective ventilation to all study namics, and reduced sedation re- sia compared with patients receiving patients and evaluated the effect on quirements.48 conventional ventilation. Of note, pa- mortality of higher PEEP (12-24 cm vs Of these alternative ventilatory tients in the intervention arm of the 5-24 cm H2O) in the treatment group. modes, only high-frequency oscilla- ARDSNet study had reduced oxygen- In addition, recruitment maneuvers tory ventilation (HFOV) has been stud- ation in the first few days but im- were conducted in patients random- ied in moderately sized randomized proved survival.27 Furthermore, a re- ized to the higher PEEP group by trials.49 In a trial of 148 patients com- cent analysis of a subset of the ARDSNet applying a pressure of 35 to 40 cm paring HFOV with conventional me- study revealed no significant differ- H2O for 30 seconds to assess the effect chanical ventilation with respect to key

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS adverse outcomes (eg, new airleak, in- the supine or the prone position for at consistent with earlier studies in dem- tractable hypotension), there were no least 6 hours per day over a 10-day pe- onstrating only transient improve- significant differences between groups. riod.53 Although oxygenation signifi- ments in oxygenation without a sig- Mortality was examined as a second- cantly improved in the prone group, nificant mortality benefit (nitric oxide, ary outcome, with a nonsignificant there was no significant difference in 23% vs placebo 20%, P=.54). lower 30-day mortality in the HFOV mortality or any secondary outcome. A group (37% vs 52%, P=.10). This find- post hoc analysis of the most severely CURRENT UNRESOLVED ing must be interpreted with caution be- ill patients revealed decreased mortal- QUESTIONS AND cause the conventional ventilation pro- ity for patients randomly assigned to the PERSONAL PERSPECTIVE tocol did not use the current standard prone position (relative risk [RR, 0.54; Current ALI and ARDS Definition for volume- and pressure-limited lung 95% CI, 0.32-0.90) at day 10, but this and Clinical Trials protective ventilation and the study was benefit did not persist beyond inten- Despite a strong physiological ratio- not powered to assess mortality. A sec- sive care unit discharge. The second nale, many therapies for ALI and ARDS ond trial of 61 ARDS patients compar- study randomized 791 patients (48% have not led to a significant survival ing HFOV with conventional mechani- with ALI or ARDS) to supine or prone benefit when tested in large clinical cal ventilation also revealed no positioning for at least 8 hours per day trials. It is possible that the case defi- significant differences in survival with- and also demonstrated improved oxy- nition of ALI and ARDS may have con- out supplemental oxygen or ventila- genation without a survival benefit at 28 tributed to this. The current American- tory support (HFOV vs conventional days (supine, 31.5% vs prone 32.4%; RR, European Consensus Conference ventilation 32% vs 38%; adjusted odds 0.97; 95% CI, 0.79-1.19; P=.77).54 This definition has important limitations in ratio, 0.80; 95% CI, 0.22-2.97; P=.79), second trial revealed a significantly its reliability and validity. First, the 50 or other secondary end points. How- higher rate of adverse events in the prone PaO2/FIO2 ratio that defines hypox- ever, a post hoc analysis revealed that group, including selective (right or left emia does not consider the effect of ven- patients with the most severe hypox- mainstem bronchus) intubation, endo- tilatory settings (eg, PEEP) or adjunc- emia had a trend toward a benefit from tracheal tube obstruction, and pres- tive therapies (eg, inhaled nitric oxide), HFOV. sure sores. The third study of prone po- which can have an important acute in- sitioning in ARDS, was stopped early fluence on PaO2, nor does it consider Adjunctive Therapies to (133 of 200 patients) due to problems whether a patient meets ALI or ARDS Lung-Protective Ventilation with enrolment. It revealed a large, but criteria.64-66 Thus, variation in ventila- Adjunctive therapies are cointerven- statistically insignificant, difference in in- tion practices across institutions may tions that may result in improved out- tensive care unit mortality between the lead to systematic differences in defin- comes when combined with lung- 2 groups (supine, 58.6% vs prone, ing this disease entity for clinical stud- protective ventilatory strategies. Prone 44.4%, P=.43).55 ies and patient management. Second, positioning and inhaled nitric oxide are Inhaled nitric oxide provides selec- there is only moderate interobserver 2 adjunctive therapies that are cur- tive vasodilation in ventilated lung units agreement in interpreting the chest ra- rently widely available and have been thus improving ventilation-perfusion diograph for ARDS, and ventilator set- studied in large randomized trials of mismatch, hypoxemia, and pulmo- tings also can influence the degree of adult patients with ALI or ARDS. nary hypertension.56 This therapy has infiltrates appearing on the radio- The use of prone positioning with been studied in 6 randomized, placebo- graph.13,67 Finally, in comparison to au- mechanical ventilation was first de- controlled trials of adults with ALI or topsy findings of nonsurvivors, a single- scribed in 1974.51 Placing the patient ARDS.57-62 None of these studies dem- center study demonstrated only in a prone position has potential physi- onstrated a sustained benefit. A Coch- moderate accuracy of the ARDS defi- ological benefits, including the recruit- rane meta-analysis of more than 500 pa- nition.68 Refinement of the current defi- ment of dorsal (nondependent) atelec- tients (80% adults) concluded that nition of ALI and ARDS to achieve more tatic lung units, improved respiratory inhaled nitric oxide led to a transient homogeneous patient populations may mechanics, decreased ventilation- improvement in oxygenation for up to be beneficial in order to detect a sig- perfusion mismatch, increased secre- 72 hours but no survival benefit (RR, nificant treatment effect in clinical trials tion drainage, reduced and improved 0.98; 95% CI, 0.66-1.44).63 After this of ALI and ARDS therapies. However, distribution of injurious mechanical meta-analysis was completed, the larg- a more restrictive definition also may forces, and improved fit of the lungs est trial of inhaled nitric oxide was pub- exclude patients who may potentially within the thorax.52 lished.62 This study randomly as- benefit from a particular intervention. There are 3 randomized trials of prone signed 385 nonseptic patients with ALI Further consideration of the current positioning in adults with ALI or or ARDS to receive either continuous ALI and ARDS definitions are neces- ARDS.53-55 In the first study, 304 pa- low-dose inhaled nitric oxide at 5 ppm sary before potentially important thera- tients were randomly assigned either to or placebo. The results of this trial were pies are rejected as nonbeneficial.

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS

Widespread Adoption of the Role for Rescue Therapy has been the most important advance ARDSNet Ventilation Protocol There are many alternative ventila- in the management of patients with ALI Given that the ARDSNet study27 pro- tory approaches and adjunctive thera- and ARDS. Volume- and pressure- vided the only ventilation protocol that pies that have some evidence of short- limited ventilation clearly leads to im- had a significant and sustained short- term physiological benefit, but no proved patient survival. The role of re- term mortality benefit, we believe consistent evidence for a survival ad- cruitment maneuvers, higher levels of patients with ALI or ARDS would ben- vantage when studied with large ran- PEEP, or both remain controversial and efit if all institutions used a lung- domized trials. Consequently, decid- are the subject of 2 ongoing multi- protective ventilation protocol based on ing the exact role of such therapies in center clinical trials (ExPress trial this volume- and pressure-limited strat- managing individual patients is diffi- [France], LOVS trial [Canada, Austra- egy. Although the ARDSnet protocol can cult. One approach is to completely lia, and Saudi Arabia]). At this time, use be successfully implemented in clini- avoid all of these therapies outside of of alternative modes of ventilation (eg, cal practice,69 widespread adoption has clinical trials until their efficacy has HFOV) and adjunctive therapies (eg, in- not rapidly occurred.70-73 One possible been adequately demonstrated. Assum- haled nitric oxide and prone position- explanation for this finding is a per- ing their physiological benefit and safety ing) should be limited to future clini- ception by caregivers that implemen- have been appropriately evaluated in cal trials and rescue therapy for patients tation could be harmful to patients,73 prior human studies, another ap- with ALI or ARDS with life-threaten- despite the lack of evidence to support proach is to limit use of these thera- ing hypoxemia failing maximal con- this notion. pies to well-defined rescue situations ventional lung-protective ventilation. Some argue that adoption of the ex- in which a patient is deemed to be fail- Although agreement on the current act ARDSNet protocol (available at ing, or at risk of harm, from conven- definition of ALI and ARDS has been a www.ardsnet.org) may be unnecessary tional ventilation. In these instances, fundamental step forward in research and use of other modes of ventilation, rescue therapy may be considered to po- and clinical practice, further refine- which achieve similar volume and pres- tentially avoid significant morbidity or ment is required to ensure that the ef- sure limitations (ie, tidal volume 4-8 mortality. However, the controversy lies ficacy of new and existing therapies are mL/kg predicted body weight and pres- in determining which rescue thera- evaluated in the most appropriate pa- Ͻ sure 30 cm H2O), may be equally ben- pies should be used (either alone or in tient population. eficial.74 Such alternate volume- and combination) and at what point they Finally, after decades of ALI and pressure-limited protocols may be easier should be initiated and terminated. For ARDS research, it is important to en- to implement and follow, especially if example, preliminary data suggest that sure widespread uptake of efficacious they are already consistent with local earlier institution of HFOV may por- volume- and pressure-limited mechani- practice patterns. However, the use of tend a better prognosis76 and that there cal ventilation. It is critical that clini- custom-made protocols, which may may be synergistic benefits of combin- cians who treat patients with ALI or seem logical, should be approached with ing different rescue modalities for ARDS ARDS reassess their current ventila- caution, for it remains unclear what spe- patients with severe, life-threatening hy- tion strategies, assimilate the available cific aspect of the ARDSNet protocol poxemia.77,78 Although some proto- evidence, and modify their practices to confers the survival advantage. For in- cols have been reported,79 there is in- ensure the highest quality of care and stance, a secondary analysis of the sufficient evidence to support the best outcomes for their patients. Un- ARDSNet data75 suggested that tidal vol- superiority of any particular approach derstanding the existing barriers to use ume reduction even benefited patients to rescue therapy. Whenever possible, of lung-protective ventilation and meth- with safe plateau pressures of 31 cm H2O such patients should be considered for ods for increasing implementation of or less of water, meaning that pressure- transfer to institutions with signifi- current research findings are impor- limited ventilation, without concomi- cant experience in ALI and ARDS man- tant opportunities for future study. tant volume-limitation, may be detri- agement to allow further expert evalu- Financial Disclosures: None reported. mental. Irrespective of this controversy ation and treatment. Additional Funding/Support: Dr Needham is supported by the Ca- as to whether the exact ARDSNet pro- research is needed to build on the an- nadian Institutes of Health Research (Clinician- Scientist Award), Royal College of Physicians and Sur- tocol should be adopted, the existing evi- ecdotal evidence supporting the ben- geons of Canada (Detweiler Fellowship), and grant P050 dence (including 2 meta-analyses of all efit of rescue therapies for the most se- HL 73994-01 from the National Institutes of Health. trials) supports that clinicians should verely ill patients with ALI or ARDS. Role of the Sponsor: None of the funding organiza- tions or sponsors had any role in the design and con- change their practice and adopt vol- duct of the study; the collection, management, analy- ume- and pressure-limited ventilation for Conclusions and sis, or interpretation of the data; or preparation, review, Future Considerations or approval of the manuscript. patients with ALI or ARDS. As addi- Acknowledgment: We thank A. Slutsky, MD, Inter- tional evidence emerges, ongoing reas- Recognition that mechanical ventila- departmental Division of Critical Care Medicine and Department of Medicine, and A. Detsky, MD, PhD, sessment and evolution of these proto- tion, although life-saving, can contrib- Department of Medicine, both at the University of cols will be necessary. ute to patient morbidity and mortality Toronto, Toronto, Ontario, and C. Soong, MD, De-

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS partment of Medicine, William Osler Health Center, has computed tomography taught us about the acute 39. Meade MO, Guyatt GH, Cook DJ, et al. Physi- Etobicoke, Ontario, for their review of early drafts of respiratory distress syndrome? Am J Respir Crit Care ologic randomized pilot study of a lung recruitment this article. Med. 2001;164:1701-1711. maneuver in acute lung injury. Am J Respir Crit Care 21. Gattinoni L, Pesenti A. ARDS: the non- Med. 2001;165:A683. homogeneous lung; facts and hypothesis. Intensive 40. Lim SC, Adams AB, Simonson DA, et al. Inter- REFERENCES Crit Care Dig. 1987;6:1-4. comparison of recruitment maneuver efficacy in three 22. Gattinoni L, Pesenti A. The concept of “baby lung.” models of acute lung injury. Crit Care Med. 2004;32: 1. Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Intensive Care Med. 2005;31:776-784. 2371-2377. Acute respiratory distress in adults. Lancet. 1967;2: 23. Amato MB, Barbas CS, Medeiros DM, et al. Effect 41. Gattinoni L, Pelosi P, Suter PM, Pedoto A, Vercesi 319-323. of a protective-ventilation strategy on mortality in the P, Lissoni A. Acute respiratory distress syndrome caused 2. Frutos-Vivar F, Nin N, Esteban A. Epidemiology of acute respiratory distress syndrome. N Engl J Med. by pulmonary and extrapulmonary disease: different acute lung injury and acute respiratory distress 1998;338:347-354. syndromes? Am J Respir Crit Care Med. 1998;158:3-11. syndrome. Curr Opin Crit Care. 2004;10:1-6. 24. Brochard L, Roudot-Thoraval F, Roupie E, et al. 42. Pelosi P, D’Onofrio D, Chiumello D, et al. Pul- 3. Rubenfeld GD. Epidemiology of acute lung injury. Tidal volume reduction for prevention of ventilator- monary and extrapulmonary acute respiratory dis- Crit Care Med. 2003;31:S276-S284. induced lung injury in acute respiratory distress tress syndrome are different. Eur Respir J Suppl. 2003; 4. Esteban A, Anzueto A, Frutos F, et al. Character- syndrome. Am J Respir Crit Care Med. 1998;158: 42:48S-56S. istics and outcomes in adult patients receiving me- 1831-1838. 43. Cakar N, Akinci O, Tugrul S, et al. Recruitment chanical ventilation: a 28-day international study. 25. Stewart TE, Meade MO, Cook DJ, et al. Evalua- maneuver: does it promote bacterial translocation? Crit JAMA. 2002;287:345-355. tion of a ventilation strategy to prevent barotrauma Care Med. 2002;30:2103-2106. 5. Montgomery AB, Stager MA, Carrico CJ, Hudson in patients at high risk for acute respiratory distress 44. The National Heart, Lung, and Blood Institute LD. Causes of mortality in patients with the adult res- syndrome: pressure and volume limited ventilation ARDS Clinical Trials Network. Higher verus lower posi- piratory distress syndrome. Am Rev Respir Dis. 1985; strategy group. N Engl J Med. 1998;338:355-361. tive end-expiratory pressures in patients with the acute 132:485-491. 26. Brower RG, Shanholtz CB, Fessler HE, et al. Pro- respiratory distress syndrome. N Engl J Med. 2004;351: 6. Milberg JA, Davis DR, Steinberg KP, Hudson LD. spective, randomized controlled clinical trial compar- 327-336. Improved survival of patients with acute respiratory ing traditional versus reduced tidal volume ventila- 45. Levy MM. PEEP in ARDS—how much is enough? distress syndrome (ARDS): 1983-1993. JAMA. 1995; tion in acute respiratory distress syndrome patients. N Engl J Med. 2004;351:389-391. 273:306-309. Crit Care Med. 1999;27:1492-1498. 46. Ritacca FV, Stewart TE. Clinical review: high fre- 7. Davidson TA, Caldwell ES, Curtis JR. Reduced qual- 27. The Acute Respiratory Distress Syndrome Network. quency-oscillatory ventilation in adults—a review of ity of life in survivors of acute respiratory distress syn- Ventilation with lower tidal volumes as compared with the literature and practical applications. Crit Care. 2003; drome compared with critically ill control patients. traditional tidal volumes for acute lung injury and the 7:385-390. JAMA. 1999;281:354-360. acute respiratory distress syndrome. N Engl J Med. 47. Imai Y, Slutsky AS. High-frequency oscillatory ven- 8. Herridge MS, Cheung AM, Tansey CM, et al. One- 2000;342:1301-1308. tilation and ventilator-induced lung injury. Crit Care year outcomes in survivors of the acute respiratory 28. Eichacker PQ, Gerstenberger EP, Banks SM, Cui Med. 2005;33:S129-S134. distress syndrome. N Engl J Med. 2003;348:683- X, Natanson C. Meta-analysis of acute lung injury and 48. Habashi NM. Other approaches to open-lung ven- 693. acute respiratory syndrome trials testing low tidal tilation: airway pressure release ventilation. Crit Care 9. Hopkins RO, Weaver LK, Pope D, Orme JF, Bigler Am J Respir Crit Care Med Med ED, Larsen-Lohr V. Neuropsychologiocal sequelae and volumes. . 2002;166: . 2005;33:S228-S240. impaired health status in survivors of severe acute res- 1510-1514. 49. Derdak S, Mehta S, Stewart TE, et al. High- piratory distress syndrome. Am J Respir Crit Care Med. 29. Petrucci N, Iacovelli W. Ventilation with smaller frequency oscillatory ventilation for acute respiratory 1999;160:50-56. tidal volumes: a quantitative systematic review of ran- distress syndrome in adults: a randomized, con- 10. Hopkins RO, Weaver LK, Collingridge D, Parkin- domized controlled trials. Anesth Analg. 2004;99: trolled trial. Am J Respir Crit Care Med. 2002;166: son RB, Chan KJ, Orme JF Jr. Two-year cognitive, emo- 193-200. 801-808. tional, and quality-of-life outcomes in acute respira- 30. Moran JL, Bersten AD, Solomon PJ. Meta- 50. Bollen CW, van Well GT, Sherry T, et al. High fre- tory distress syndrome. Am J Respir Crit Care Med. analysis of controlled trials of ventilator therapy in acute quency oscillatory ventilation compared with conven- 2005;171:340-347. lung injury and acute respiratory distress syndrome: tional mechanical ventilation in adult respiratory dis- 11. Petty TL, Ashbaugh DG. The adult respiratory dis- an alternative perspective. Intensive Care Med. 2005; tress syndrome: a randomized controlled trial. Crit Care. tress syndrome: clinical features, factors influencing 31:227-235. 2005;9:R430-R439. prognosis and principles of management. Chest. 1971; 31. Brower RG, Matthay M, Schoenfeld D. Meta- 51. Bryan AC. Comments of a devil’s advocate. Am 60:233-239. analysis of acute lung injury and acute respiratory dis- Rev Respir Dis. 1974;110:143-144. 12. Murray JF, Matthay MA, Luce JM, Flick MR. An tress syndrome trials. Am J Respir Crit Care Med. 2002; 52. Pelosi P, Brazzi L, Gattinoni L. Prone position in expanded definition of the adult respiratory distress 166:1515-1517. acute respiratory distress syndrome. Eur Respir J. 2002; syndrome. Am Rev Respir Dis. 1988;138:720-723. 32. Laffey JG, O’Croinin D, McLoughlin P, Ka- 20:1017-1028. 13. Bernard GR, Artigas A, Brigham KL, et al. The vanagh BP. Permissive hypercapnia—role in lung pro- 53. Gattinoni L, Tognoni G, Pesenti A, et al. Effect of American-European Consensus Conference on ARDS: tective ventilatory strategies. Intensive Care Med. 2004; prone positioning on the survival of patients with definitions, mechanisms, relevant outcomes, and clini- 30:347-356. acute respiratory failure. N Engl J Med. 2001;345: cal trial coordination. Am J Respir Crit Care Med. 1994; 33. Kahn JM, Andersson L, Karir V, Polissar NL, Neff 568-573. 149:818-824. MJ, Rubenfeld GD. Low tidal volume ventilation does 54. Guerin C, Gaillard S, Lemasson S, et al. Effects of 14. American Thoracic Society. International consen- not increase sedation use in patients with acute lung systematic prone positioning in hypoxemic acute res- sus conferences in intensive care medicine: ventilator- injury. Crit Care Med. 2005;33:766-771. piratory failure: a randomized controlled trial. JAMA. associated lung injury in ARDS. Am J Respir Crit Care 34. Lapinsky SE, Mehta S. Bench-to-bedside review: 2004;292:2379-2387. Med. 1999;160:2118-2124. recruitment and recruiting maneuvers. Crit Care. 2005; 55. Mancebo J, Rialp G, Fernandez R, et al; Random- 15. Crimi E, Slutsky AS. Inflammation and the acute 9:60-65. ized multicenter trial in ARDS. Supine vs prone position. respiratory distress syndrome. Best Pract Res Clin 35. Lapinsky SE, Aubin M, Mehta S, Boiteau P, Slut- Intensive Care Med. 2003;29:S64. Anaesthesiol. 2004;18:477-492. sky A. Safety and efficacy of a sustained inflation for 56. Rossaint R, Falke KJ, Lopez F, Slama K, Pison U, 16. Plotz FB, Slutsky AS, van Vught AJ, Heijnen CJ. alveolar recruitment in adults with respiratory failure. Zapol WM. Inhaled nitric oxide for the adult respira- Ventilator-induced lung injury and multiple system or- Intensive Care Med. 1999;25:1297-1301. tory distress syndrome. N Engl J Med. 1993;328: gan failure: a critical review of facts and hypotheses. 36. Pelosi P, Cadringher P, Bottino N, et al. Sigh in 399-405. Intensive Care Med. 2004;30:1865-1872. acute respiratory distress syndrome. Am J Respir Crit 57. Dellinger RP, Zimmerman JL, Taylor RW, et al. Ef- 17. Dreyfuss D, Soler P, Basset G, Saumon G. High Care Med. 1999;159:872-880. fects of inhaled nitric oxide in patients with acute res- inflation pressure pulmonary edema: respective ef- 37. Foti G, Cereda M, Sparacino M, DeMarchi ME, piratory distress syndrome: results of a randomized fects of high airway pressure, high tidal volume, and Villa F, Pesenti A. Effects of periodic lung recruitment phase II trial. Crit Care Med. 1998;26:15-23. positive end-expiratory pressure. Am Rev Respir Dis. maneuvers on gas exchange and respiratory mechan- 58. Michael JR, Barton RG, Saffle JR, et al. Inhaled 1988;137:1159-1164. ics in mechanically ventilated ARDS patients. Inten- nitric oxide versus conventional therapy: effect on oxy- 18. Slutsky AS. Lung injury caused by mechanical sive Care Med. 2000;26:501-507. genation in ARDS. Am J Respir Crit Care Med. 1998; ventilation. Chest. 1999;116:9S-15S. 38. The ARDS Clinical Trials Network. Effects of re- 157:1372-1380. 19. Tremblay LN, Slutsky AS. Ventilator-induced lung cruitment maneuvers in patients with acute lung in- 59. Troncy E, Collet JP, Shapiro S, et al. Inhaled nitric injury: from barotrauma to biotrauma. Proc Assoc Am jury and acute respiratory distress syndrome venti- oxide in acute respiratory distress syndrome: a pilot Physicians. 1998;110:482-488. lated with high positive end-expiratory pressure. Crit randomized controlled study. Am J Respir Crit Care 20. Gattinoni L, Caironi P, Pelosi P, Goodman LR. What Care Med. 2003;31:2592-2597. Med. 1998;157:1483-1488.

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006 MANAGEMENT OF ACUTE LUNG INJURY AND ARDS

60. Lundin S, Mang H, Smithies M, Stenqvist O, Frost- 67. Rubenfeld GD, Caldwell E, Granton J, Hudson LD, BT, Hudson LD. Barriers to providing lung-protective ell C. Inhalation of nitric oxide in acute lung injury: re- Matthay MA. Interobserver variability in applying ventilation to patients with acute lung injury. Crit Care sults of a European multicentre study. Intensive Care a radiographic definition for ARDS. Chest. 1999;116: Med. 2004;32:1289-1293. Med. 1999;25:911-919. 1347-1353. 74. Marini JJ, Gattinoni L. Ventilatory management 61. Mehta S, Simms HH, Levy MM, et al. Extended 68. Esteban A, Fernandez-Segoviano P, Frutos- of the acute respiratory distress syndrome: a therapy with inhaled nitric oxide fails to improve oxy- Vivar F, et al. Comparison of clinical criteria for the consensus of two. Crit Care Med. 2004;32:250-255. genation in patients with acute respiratory distress syn- acute respiratory distress syndrome with autopsy 75. Hager DN, Krishnan JA, Hayden DL, Brower RG. drome: a randomized controlled trial. J Appl Res. 2001; findings. Ann Intern Med. 2004;141:440-445. Tidal volume reduction in patients with acute lung in- 1:73-84. 69. Kallet RH, Jasmer RM, Pittet JF, et al. Clinical imple- jury when plateau pressures are not high. Am J Respir 62. Taylor RW, Zimmerman JL, Dellinger RP, et al. mentation of the ARDS network protocol is associ- Crit Care Med. 2005;172:1241-1245. Low-dose inhaled nitric oxide in patients with acute ated with reduced hospital mortality compared 76. Mehta S, Granton J, MacDonald RJ, et al. High- lung injury: a randomized controlled trial. JAMA. 2004; with historical controls. Crit Care Med. 2005;33:925- frequency oscillatory ventilation in adults: the Toronto 291:1603-1609. 929. experience. Chest. 2004;126:518-527. 63. Sokol J, Jacobs SE, Bohn D. Inhaled nitric oxide 70. Brower RG, Thompson BT, Ancukiewicz M. Clini- 77. Fan E, Mehta S. High-frequency oscillatory ven- for acute respiratory failure in children and adults: a cal trial of mechanical ventilation with traditional ver- tilation and adjunctive therapies: inhaled nitric oxide meta-analysis. Anesth Analg. 2003;97:989-998. sus lower tidal volumes in acute lung injury and acute and prone positioning. Crit Care Med. 2005;33: 64. Shapiro BA, Crane RD, Harrison RA, Steiner MC. respiratory distress syndrome: effects on physician’s S182-S187. Changes in intrapulmonary shunting with adminis- practices. Am J Respir Crit Care Med. 2004;169:A17. 78. Ferguson ND, Chiche JD, Kacmarek RM, et al. tration of 100 percent oxygen. Chest. 1980;77: 71. Weinert CR, Gross CR, Marinelli WA. Impact of Combining high-frequency oscillatory ventilation 138-141. randomized trial results on acute lung injury ventila- and recruitment maneuvers in adults with early 65. Santos C, Ferrer M, Roca J, Torres A, Hernandez tor therapy in teaching hospitals. Am J Respir Crit Care acute respiratory distress syndrome: the Treatment C, Rodriguez-Roisin R. Pulmonary gas exchange re- Med. 2003;167:1304-1309. with Oscillation and an Open Lung Strategy (TOOLS) sponse to oxygen breathing in acute lung injury. Am 72. Young MP, Manning HL, Wilson DL, et al. Ven- Trial Pilot study. Crit Care Med. 2005;33:479- J Respir Crit Care Med. 2000;161:26-31. tilation of patients with acute lung injury and acute 486. 66. Ferguson ND, Kacmarek RM, Chiche JD, et al. respiratory distress syndrome: has new evidence 79. Medoff BD, Shepard JO, Smith RN, Kratz A. Case Screening of ARDS patients using standardized ven- changed clinical practice? Crit Care Med. 2004;32: 17-2005—a 22-year old woman with back and leg tilator settings: influence on enrollment in a clinical trial. 1260-1265. pain and respiratory failure. N Engl J Med. 2005;352: Intensive Care Med. 2004;30:1111-1116. 73. Rubenfeld GD, Cooper C, Carter G, Thompson 2425-2434.

Art may be served by morality: it can never be its ser- vant. For the principles of art are eternal, while the principles of morality fluctuate with the spiritual ebb and flow of the ages. —Arthur Symons (1865-1945)

Downloaded from www.jama.com at K C Library Svs 142d, on January 28, 2006