Secondary Logo

Journal Logo

Clinical Critical Care

Use of ECMO in the Management of Severe Acute Respiratory Distress Syndrome

A Survey of Academic Medical Centers in the United States

Sharma, Nirmal S.*; Wille, Keith M.*; Zhi, Degui; Thannickal, Victor J.*; Brodie, Daniel M.; Hoopes, Charles W.§; Diaz-Guzman, Enrique*

Author Information
doi: 10.1097/MAT.0000000000000245
  • Free


Acute respiratory distress syndrome (ARDS) is a common diagnosis in the intensive care unit (ICU) and a frequent cause of morbidity and mortality among hospitalized patients.1 Cross-sectional studies have shown that patients with ARDS represent up to 5–20% of all mechanically ventilated patients, the majority of whom develop moderate-to-severe disease.2,3 Mortality of ARDS remains high, particularly among patients with severe disease (PaO2 to FiO2 ratio <100 mm Hg).4 Although changes in ventilation strategies have improved outcomes in ARDS,5 recent evidence suggests that mortality still remains very high especially in patients with severe ARDS.6

Moreover, despite wide recognition of lung injury associated with excessive tidal volume ventilation, studies have shown that variable adoption of ventilatory strategies in patients with ARDS may still exist.7,8 Furthermore, there is limited evidence to support the use of different advanced therapies frequently used in ARDS such as airway pressure release ventilation (APRV), inhaled nitric oxide (iNO), extracorporeal membrane oxygenation (ECMO), and high-frequency oscillatory ventilation (HFOV), and currently, there are no widely accepted guidelines to help standardize the approach to patients with severe ARDS.

Recent advances in ECMO technology, and improved outcomes among patients supported with ECMO,9,10 have resulted in a renewed interest in the use of this therapy for selected patients with severe ARDS.10,11 Similarly, a recent study suggesting that ventilation in the prone positioning (PP) improves mortality for patients with severe ARDS12 has prompted many to debate whether ECMO and/or PP should be the preferred therapies for patients with severe ARDS.

This study was aimed to ascertain current practices for patients with severe ARDS at academic institutions in the United States. In addition, we intended to understand the current preferences of intensivists regarding the use of different therapies for severe ARDS, including the use of ECMO.

Material and Methods

Definition of Severe ARDS

For purposes of this survey research, the following definition was used and provided to all participants of the study: 1) PaO2 to FiO2 ratio less than or equal to 100 mm Hg; 2) use of high levels of positive end-expiratory pressure more than or equal to 15 cm water; and 3) use of high FiO2 more than or equal to 0.8

Development of the Survey

After obtaining institutional review board (IRB) approval at the University of Alabama at Birmingham (IRB no. E140108008), N.S.S. developed the initial survey instrument with 17 questions using iterative process. It was then reviewed by K.W., D.B., and E.D.G. and changes were made with the addition of two more questions. Three researchers, E.D.G., K.W., and D.B. are experienced clinicians in the field of ECMO and lung transplantation. The initial questionnaire consisted of four demographics items (location of program, subspecialty, designation, and number of fellows in the program), and three other sections aimed at understanding the perceptions of adult critical care physicians in treatment of severe ARDS, use of ECMO in severe ARDS, and the role of ECMO training during critical care fellowship.

Pretesting and Pilot Survey

To validate the instrument before launching, researchers pretested the survey. This validation methodology is recommended to assess participants understanding of items and whether items appear to tap the construct of interest.13,14 The aim was for respondents to understand not only the literal meaning of survey questions but also the pragmatic meaning, to reduce response error. Pulmonary and critical care fellows and faculty at the University of Alabama at Birmingham participated in the pretest of the survey. A total of 35 participants (15 critical care fellows, 1 critical care program director [PD], and 19 critical care faculty) were sent the survey of which 23 responded. An Internet-based survey platform (SurveyMonkey, Palo Alto, CA) was used. A focus group session was conducted, and respondents were asked to comment on the appropriateness of response choices, ease of use, and understanding of each question. Response time of each member was also recorded and found to be a mean of 9 minutes.

During the structured group interview, respondents were queried about their interpretation of specific survey items and the decision processes used to answer items, to reduce response error and increase construct validity. Investigator N. S.S., who has no role in fellow grading, acted as focus group facilitator. Narrative data were noted and saved in Microsoft Word (Microsoft Corporation, Richmond, WA). Based on the responses and comments received, two questions were modified with addition of Likert scale and one question added to negate bias.

National Launch of Survey

The refined survey instrument was launched nationally via e-mail. The link to the survey was sent to participating PDs or program coordinators of 302 adult critical care fellowship programs in the United States. This included the following subspecialties: pulmonary and critical care (PCC), anesthesia critical care (ACC), surgical critical care (SCC), internal medicine critical care (IMCC), and emergency medicine critical care (EMCC). Fellowship programs were identified from the Fellowship and Residency Electronic Interactive Database on the American Medical Association website. Program directors were asked to forward the survey link to their own faculty and trainees. After an initial e-mail inviting PDs to participate, two subsequent reminders were sent. The survey link was kept active between February 16 and April 16, 2014. Responses were compiled by the Internet-based survey program (SurveyMonkey). Participant identity was kept confidential during the data collection process, and no personally identifiable material was collected. To protect respondent’s identity, we requested the survey link to be forwarded by the PD or the program coordinator to all critical trainees and faculty.

Statistical Analysis

We performed descriptive analyses of measured variables using means and standard deviations or medians with interquartile ranges as appropriate. Descriptive statistics were used to report categorical data, and associations between categorical variables were tested with a 2 × 2 or Fisher’s exact test. Variables that were statistically significant by univariate analysis were tested for significance by multivariate analysis. Associations between program demographic characteristics and survey responses were explored via 2 × 2 analysis. Responses of “agree” and “strongly agree,” as well as “disagree” and “strongly disagree,” were grouped together for categorical analysis. Because of the nonnormality of the data, the Wilcoxon rank sum test was used to examine and compare the survey responses for continuous data. Statistical analysis was performed with JMP, version 10.0 software (JMP, Cary, NC).


The survey was distributed among 302 critical care training programs in the United States: PCC (n = 130, 43%), ACC (n = 45, 15%), SCC (n = 94, 31%), IMCC (n = 30, 10%), and EMCC (n = 3, 1%). A total of 381 responses were received: 83 PD, 120 critical care faculty, 174 critical care trainees, and “four other” (Table 1). The PD response rate was 27.4%. Survey respondents represented programs located in 35 states (Figure 1).

Table 1
Table 1:
Baseline Respondent Characteristics
Figure 1
Figure 1:
Geographical distribution of the respondents.

Treatment of Severe ARDS

A majority of respondents (72%) described caring for patients with severe ARDS very often (weekly) or often (>3–4 times/month). These responses did not vary by designation of the respondent or by subspecialty (p = 0.13). In addition to mechanical ventilation, the following modalities were available and used by the respondents at their institutions (% of the respondents): ECMO (68%), iNO (65%), APRV (64%), PP (45%), and HFOV (29%).

When asked to rank (1–5), their preferred modality of treatment in patients with severe ARDS who deteriorate despite optimal mechanical ventilation strategies, the following was reported: APRV was selected by most respondents as the first choice (41%), followed by iNO (29%), PP (20%), HFOV (5%), and ECMO (4%) (Figure 2). Responses were not statistically different between faculty and trainees when ranking APRV, iNO, HFOV, and PP. However, fellows were less averse to using ECMO for severe ARDS compared with faculty (p < 0.04; Figure 3). Among subspecialties, APRV was preferred least by PCC (p = 0.001), PP was favored least by ACC, and SCC (p = 0.02) and iNO use as the initial choice was reported more by PCC and ACC (p = 0.01).

Figure 2
Figure 2:
Preferred modality of treatment for severe acute respiratory distress syndrome once conventional mechanical ventilation fails (ranking in descending order of preference). APRV, airway pressure release ventilation; HFOV, high-frequency oscillator ventilation; NO, nitric oxide; PP, prone positioning.
Figure 3
Figure 3:
Preferred modality of treatment for severe acute respiratory distress syndrome once conventional mechanical ventilation fails (ranking in descending order of preference). A comparison between facultyversus fellow responses. APRV, airway pressure release ventilation; HFOV, high-frequency oscillator ventilation; INO, inhaled nitric oxide; PP, prone positioning.

Use of ECMO in Severe ARDS

Although 80% of the respondents described ECMO availability at their institutions, 13% of them reported that ECMO was not used for severe ARDS. When respondents were asked whether they would consider using ECMO if it was available at their institutions, a majority (83%) agreed. These responses did not vary by designation of the respondent or by subspecialty. When asked whether ECMO should be favored over other modalities, a majority (70%) disagreed, and there were no differences in opinion between fellows and faculty or among respondents from various subspecialties (Figures 4 and 5). However, 57% of respondents believed that ECMO provides better patient outcomes when compared with other treatment modalities for severe ARDS, whereas 60% of respondents also believed that ECMO would provide a better chance for the patient to receive lung protective ventilation (tidal volume <6 ml/kg body weight; Figures 4 and 5). Although these responses were not statistically different between fellows and faculty, differences existed among respondents from various subspecialties. Respondents from SCC, IMCC, and EMCC believed more strongly that there is enough scientific evidence to support ECMO use compared with respondents from PCC and ACC (p = 0.02). Eighty-seven percent of respondents believed that ECMO was unavailable in most hospitals in the country, whereas half of the respondents believed that ECMO was associated with high costs and poor patient outcomes with no differences in opinion between fellows and faculty (Figures 4 and 5). These responses differed between the various critical care subspecialties, with those from SCC, IMCC, and EMCC more in disagreement with the statement that ECMO was associated with high costs and poor outcomes compared with PCC and ACC (p = 0.01). Among those respondents who would consider using ECMO for patients with severe ARDS, 88% believed that it was beneficial for patients with influenza pneumonia, whereas 57% believed that it is useful for all patients regardless of the etiology of their ARDS. Early referral for ECMO was preferred by most (93%) across all groups with 58% answering that they would use it within 72 hours and 35% within a week of severe ARDS diagnosis. Among respondents who described that ECMO was not available at their institutions, a majority reported never (43%) or rarely (42%) transferring patients to the nearest ECMO facility for support.

Figure 4
Figure 4:
Use of ECMO in management of severe ARDS: Perceptions of the respondents regarding the use of ECMO in severe ARDS. ARDS, acute respiratory distress syndrome.
Figure 5
Figure 5:
Use of ECMO in management of severe ARDS: Perceptions of the respondents regarding the use of ECMO in severe ARDS. A comparison between facultyversus fellow responses. ARDS, acute respiratory distress syndrome.

ECMO and Fellowship Training

Of respondents, 72% described having some exposure to ECMO in their training program for various indications including support for severe ARDS, cardiogenic shock, as a bridge to lung transplant, or use during the perioperative period. However, a majority (83%) of the faculty and fellows across all surveyed critical care subspecialties believed that there was not enough ECMO experience and training among critical care physicians, and 62% of all respondents reported having limited knowledge about ECMO. Faculty responses indicated more self-described knowledge about ECMO compared with the fellows (p = 0.01), and among subspecialties, respondents from ACC felt most knowledgeable about different ECMO technologies (p = 0.001). Ninety percent of respondents agreed with having specific educational objectives about ECMO as part of a critical care training curriculum, although more fellows agreed with this statement than faculty (95 vs. 85%, p < 0.001). A majority of respondents (86%) agreed that management of patients with severe ARDS using ECMO was within the purview of critical care training.


The treatment of patients with severe ARDS who deteriorate despite the use of conventional mechanical ventilation has not been well studied, and definitive guidelines to direct the use of other supportive therapies are lacking. In this study, we confirmed that, among our respondents, severe ARDS is a common ICU diagnosis and found that there are significant variations in therapeutic choices among critical care physicians. Airway pressure release ventilation was considered by most as the initial therapy, whereas PP seems to be the most common second choice. These results contrast with the available data to support the use of APRV in patients with severe ARDS and suggest that PP remains underused despite new evidence indicating that it may be associated with improved mortality.12,15

Recent experience with ECMO in patients with ARDS during the 2009 influenza A (H1N1) pandemic has rejuvenated the interest in this technology as an alternate therapy for patients with severe ARDS.11,16 However, evidence remains limited, and a recent literature review concluded that the benefit of ECMO on hospital mortality is unclear.17 Despite these limitations, there has been a significant increase in the number of patients supported by ECMO for acute respiratory failure and the number of centers offering ECMO in the United States.18 Interestingly, a large proportion of our survey respondents believed that ECMO was not widely available, which stands in contrast to the fact that it was reported as being available in more than 80% of the academic programs surveyed. This illustrates the misperception among physicians practicing in tertiary care centers, of ECMO being an uncommon technology. Importantly, although a majority of the respondents agreed that ECMO could facilitate low tidal volume ventilation in patients with high ventilatory requirements and that 76% believed that with recent technology advances, ECMO use was safer, only 4% of the respondents chose ECMO as the next step in treating patients with severe ARDS. This suggests that ECMO is perceived as salvage therapy among physicians practicing in academic medical centers where this technology seems to be widely available.

Our survey found that a majority of U.S. critical care programs provide limited ECMO-related education and exposure to trainees, and a majority of respondents favored inclusion of ECMO training as part of the curriculum for critical care physicians. These findings are particularly relevant as there has been a 433% increase in the number of adult patients supported with ECMO in the United States since 2006.19 Interestingly, respondents from ACC reported to be more knowledgeable regarding ECMO compared with other specialties, suggesting that there is a gap in critical care training regarding ECMO exposure. A previous survey conducted among critical care physicians (in Australia and New Zealand) demonstrated lack of training and availability as a factor in limiting the ECMO use in ARDS.20 We speculate that increased training of critical care physicians may improve optimal utilization of this resource and provide increased access to patients who may benefit from this technology.

The mortality associated with severe ARDS remains high despite significant advances in critical care, including widespread acceptance of a low-volume, low-pressure ventilation strategy.4 Our survey findings emphasize the need for evidence comparing alternative therapies for patients with severe ARDS. Currently, there is no consensus about which of these therapies should be used, and the use of PP seems to be limited at this time despite good evidence for its use in experienced centers. Some approaches such as HFOV continue to be applied by some physicians despite recent studies suggesting no benefit or increased mortality.21,22

The survey results include responses from both faculty and fellows. However, faculty responses may provide a more realistic understanding of the current status of ECMO use in the management of ARDS, as many fellows may not have the requisite training or experience in the management of patients on ECMO. A comparison of faculty and fellow responses has been provided in detail in Figures 3 and 5.

Our study is not without limitations. Survey results reflect critical care faculty and fellow’s perceptions of use of various treatment modalities for severe ARDS including ECMO, which are thus subject to errors of interpretation and misperception.23 The PD response rate was low. However, this response rate is in line with means reported for e-mail and Web-based surveys both in health care research and in other fields.24–27 As surveys were distributed anonymously by PDs, we were unable to determine the exact number of trainees who received surveys, and therefore, we were unable to calculate a maximum response rate. In addition, as responses from trainees were anonymous, sampling bias is possible as small fractions of trainee participants from some programs may be skewing perceptions. However, strength of this survey include that responses were received from participants in 36 different states. Cook et al.28 have noted that representativeness of responses is considered to be more critical than response rate alone.


Our study shows that there is a significant variation among clinicians at U.S. academic medical centers in the selection of treatment modalities for patients with severe ARDS. Airway pressure release ventilation appears to be the most common first-line approach to severe ARDS. Most critical care physicians would consider ECMO in patients who fail optimal mechanical ventilation strategies and believe that ECMO use can facilitate lung protective ventilation; however, few favor ECMO as a first-line modality. Critical care physicians and trainees report limited knowledge of ECMO and desire to have specific ECMO educational objectives as part of the critical care curriculum. Future efforts are needed to produce guidelines to direct different therapies for patients with severe ARDS.


1. Walkey AJ, Summer R, Ho V, Alkana P. Acute respiratory distress syndrome: Epidemiology and management approaches. Clin Epidemiol. 2012;4:159–169
2. Esteban A, Anzueto A, Frutos F, et al.Mechanical Ventilation International Study Group. Characteristics and outcomes in adult patients receiving mechanical ventilation: A 28-day international study. JAMA. 2002;287:345–355
3. Estenssoro E, Dubin A, Laffaire E, et al. Incidence, clinical course, and outcome in 217 patients with acute respiratory distress syndrome. Crit Care Med. 2002;30:2450–2456
4. Ranieri VM, Rubenfeld GD, Thompson BT, et al. Acute respiratory distress syndrome: The Berlin Definition. JAMA. 2012;307:2526–2533
5. Eisner MD, Thompson T, Hudson LD, et al.Acute Respiratory Distress Syndrome Network. Efficacy of low tidal volume ventilation in patients with different clinical risk factors for acute lung injury and the acute respiratory distress syndrome. Am J Respir Crit Care Med. 2001;164:231–236
6. Villar J, Blanco J, Añón JM, et al.ALIEN Network. The ALIEN study: Incidence and outcome of acute respiratory distress syndrome in the era of lung protective ventilation. Intensive Care Med. 2011;37:1932–1941
7. Umoh NJ, Fan E, Mendez-Tellez PA, et al. Patient and intensive care unit organizational factors associated with low tidal volume ventilation in acute lung injury. Crit Care Med. 2008;36:1463–1468
8. Walkey AJ, Wiener RS. Risk factors for underuse of lung-protective ventilation in acute lung injury. J Crit Care. 2012;27:323.e1–323.e9
9. Paden ML, Rycus PT, Thiagarajan RRELSO Registry. . Update and outcomes in extracorporeal life support. Semin Perinatol. 2014;38:65–70
10. Peek GJ, Mugford M, Tiruvoipati R, et al.CESAR Trial Collaboration. Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): A multicentre randomised controlled trial. Lancet. 2009;374:1351–1363
11. Roch A, Lepaul-Ercole R, Grisoli D, et al. Extracorporeal membrane oxygenation for severe influenza A (H1N1) acute respiratory distress syndrome: A prospective observational comparative study. Intensive Care Med. 2010;36:1899–1905
12. Guérin C, Reignier J, Richard JC, et al.PROSEVA Study Group. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. 2013;368:2159–2168
13. Desimone LM, Le Floch KC. Are we asking the right questions? Using cognitive interviews to improve surveys in education research. Educ Eval Policy Anal. 2004;26:1–22
14. Willis GB, Royston P, Bercini D. The use of verbal report methods in the development and testing of survey questionnaires. Appl Cogn Psychol. 1991;5:251–267
15. Henderson WR, Griesdale DE, Dominelli P, Ronco JJ. Does prone positioning improve oxygenation and reduce mortality in patients with acute respiratory distress syndrome? Can Respir J. 2014;21:213–215
16. Mitchell MD, Mikkelsen ME, Umscheid CA, Lee I, Fuchs BD, Halpern SD. A systematic review to inform institutional decisions about the use of extracorporeal membrane oxygenation during the H1N1 influenza pandemic. Crit Care Med. 2010;38:1398–1404
17. Zampieri FG, Mendes PV, Ranzani OT, et al. Extracorporeal membrane oxygenation for severe respiratory failure in adult patients: A systematic review and meta-analysis of current evidence. J Crit Care. 2013;28:998–1005
18. Paden ML, Conrad SA, Rycus PT, Thiagarajan RRELSO Registry. . Extracorporeal Life Support Organization Registry Report 2012. ASAIO J. 2013;59:202–210
19. Sauer CM, Yuh DD, Bonde P. Extracorporeal membrane oxygenation (ECMO) use has increased by 433% in adults in the United States from 2006 to 2011. ASAIO J. 2014;61:31–36
20. Hastings SL, Pellegrino VA, Preovolos A, Salamonsen RF. Survey of adult extracorporeal membrane oxygenation (ECMO) practice and attitudes among Australian and New Zealand intensivists. Crit Care Resusc. 2008;10:46
21. Ferguson ND, Cook DJ, Guyatt GH, et al.OSCILLATE Trial Investigators; Canadian Critical Care Trials Group. High-frequency oscillation in early acute respiratory distress syndrome. N Engl J Med. 2013;368:795–805
22. Young D, Lamb SE, Shah S, et al.OSCAR Study Group. High-frequency oscillation for acute respiratory distress syndrome. N Engl J Med. 2013;368:806–813
23. Brunkhorst FM, Engel C, Ragaller M, et al.German Sepsis Competence Network (SepNet). Practice and perception—A nationwide survey of therapy habits in sepsis. Crit Care Med. 2008;36:2719–2725
24. Cohn SM, Price MA, Stewart RM, et al. Surgical critical care and private practice surgeons: A different world out there! J Am Coll Surg. 2008;206:419–425
25. Kahn JM, Asch RJ, Iwashyna TJ, et al. Physician attitudes toward regionalization of adult critical care: A national survey. Crit Care Med. 2009;37:2149–2154
26. Scott A, Jeon SH, Joyce CM, et al. A randomised trial and economic evaluation of the effect of response mode on response rate, response bias, and item non-response in a survey of doctors. BMC Med Res Methodol. 2011;11:126
27. Vaporciyan AA, Reed CE, Erikson C, et al. Factors affecting interest in cardiothoracic surgery: Survey of North American general surgery residents. Ann Thorac Surg. 2009;87:1351–1359
28. Cook C, Heath F, Thompson RL. A meta-analysis of response rates in web- or internet-based surveys. Educ Psychol Meas. 2000;606:821–836

ECMO; severe ARDS; APRV; prone positioning; inhaled nitric oxide; ECMO survey

Copyright © 2015 by the American Society for Artificial Internal Organs