Share this article on:

Antimicrobial Prophylaxis and Infection Surveillance in Extracorporeal Membrane Oxygenation Patients: A Multi-Institutional Survey of Practice Patterns

Kao, Lillian S.*†‡; Fleming, Geoffrey M.§; Escamilla, Richard J.*; Lew, Debbie F.*; Lally, Kevin P.*†‡

doi: 10.1097/MAT.0b013e31820d19ab
Respiratory Support

The goal of this study was to characterize antimicrobial prophylaxis and infection surveillance practices at centers treating patients with extracorporeal membrane oxygenation (ECMO). A 37-question web-based survey was sent to all ECMO coordinators and directors at Extracorporeal Life Support Organization (ELSO) participating centers. Data were reported by center. The most complete response was used when multiple surveys were returned from a single center, and respondents' answers from the same center were analyzed for concordance. Responses were obtained from 76% of ELSO centers (132/173) and 41% of survey recipients (223/548). Most centers administer antibiotic prophylaxis (74%, 97/132), and almost half have a standardized protocol (49%, 64/132). Routine antibacterial but not antifungal prophylaxis is common (42%, 62/132 vs. 2/132, 2%). There is significant variation in the antibiotic choices and duration of prophylaxis, regardless of whether the center has a protocol or not. Almost half of centers (49%, 64/132) perform routine surveillance cultures but at variable intervals. There is significant heterogeneity in antibiotic prophylaxis and infection surveillance practice patterns among ELSO centers.

From the *Department of Surgery, †Department of Pediatric Surgery, ‡Center for Surgical Trials and Evidence-Based Practice (C-STEP), University of Texas Health Science Center at Houston, Houston, Texas; and §Department of Pediatrics, Vanderbilt University, Nashville, Tennessee.

Submitted for consideration September 2010; accepted for publication in revised form December 2010.

Presented at the 21st Annual ELSO Conference in St. Petersburg, FL, October 2010.

Reprint Requests: Lillian S. Kao, MD, MS, 5656 Kelley Street, Suite 30S 62008, Houston, TX 77026. Email:

Nosocomial infections are common in patients treated with extracorporeal membrane oxygenation (ECMO), with rates up to 45% in retrospective case series.1–11 They have been associated with increased length of hospital stay, length of intensive care unit stay, and increased mortality.2,4,5,8,9 A recent review of adult patients from the Extracorporeal Life Support Organization (ELSO) registry reported that despite advances in critical care, there has been no significant decrease in documented infectious complications during a 20-year time period.12 Strategies for improving outcome from nosocomial infections have been proposed but remain unproven, such as routine use of surveillance cultures for earlier diagnosis and antibiotic prophylaxis.

Patient factors such as age and autoimmune status and treatment factors such as duration of ECMO and mechanical complications are associated with nosocomial infections during ECMO treatment.13,14 Strategies used to reduce the risk for infectious complications include use of sterile techniques for cannula insertion and antibiotic prophylaxis. For other surgical procedures, evidence-based antibiotic prophylactic strategies are explicitly defined by the Surgical Care Improvement Project as the appropriate timing, spectrum, and duration of antibiotics administered perioperatively to prevent infectious complications.15,16 However, the optimal spectrum and duration of antimicrobial prophylaxis for ECMO patients are unknown. Proponents of prolonged antibiotic prophylaxis cite immune dysfunction,17 presence of invasive catheters, and multiple portals of entry into the ECMO circuit as increasing the risk for infection. Furthermore, specific high-risk subsets of patients may benefit from broadened or prolonged antimicrobial therapy such as patients with open chests.4,7 Alternatively, opponents have cited the increase in multidrug-resistant organisms over time in both ECMO patients and critically ill non-ECMO patients as a reason to avoid prolonged prophylaxis.

No randomized trials have addressed antibiotic prophylaxis in ECMO patients, and there are no data regarding the variation in infection surveillance strategies across centers. Moreover, risk for infectious complications depends on patient and ECMO characteristics, so that the optimal prophylactic regimen may differ among subgroups. The goal of this study was to survey institutions participating in the ELSO registry regarding their prophylactic antibiotic and infection surveillance practices in ECMO patients and to determine equipoise for and feasibility of a future randomized trial.

Back to Top | Article Outline


A 37-question web-based survey was devised addressing antimicrobial prophylaxis and surveillance practice patterns. All ECMO coordinators and directors were identified and email addresses obtained from the ELSO website. Institutional Review Board approval was obtained from the University of Texas Health Science Center at Houston (HSC-MS-10-0215). The survey was conducted using SurveyMonkey. The original email was sent on July 2, 2010, with reminders sent on July 19 and August 17, 2010. The survey (see Appendix) included questions on a) standardized protocols for antimicrobial prophylaxis, b) indications for and choice and duration of antibacterial and/or antifungal prophylaxis, and c) nosocomial infection surveillance practices.

The data were anonymously compiled and analyzed by institution. When more than one survey was received per institution, the survey that was most complete was used for analysis. If more than one complete survey was received, the director's responses or one set of responses was arbitrarily chosen (if more than one director or no director was identified). Although the survey was designed, such that follow-up questions were skipped based on previous answers, some respondents answered questions that contradicted their prior answers, so not all the total numbers of respondents for each question are equivalent. Answers are reported as percentages for categorical data and ranges for continuous data.

Back to Top | Article Outline


A total of 556 surveys were sent out to 172 ELSO participating centers world wide; eight surveys had the wrong email address. A total of 223 (41%) recipients responded, of whom 198 (89%) completed the survey in its entirety. The majority of respondents identified themselves as ECMO directors (66%) or coordinators (30%). The rest of the respondents were subspecialty or codirectors or team members (i.e., team leader, nurse, surgeon, pharmacist, or quality improvement personnel). Eighty-two percent of centers were from the United States.

Back to Top | Article Outline

Antibiotic Protocols

The majority of respondents agreed that antibiotic prophylaxis is administered to ECMO patients at their centers as part of a protocol or physician preference (68%, 151/223). Among those who administered antibiotic prophylaxis, 65% responded that there was a standardized protocol. Among centers with multiple respondents, 32% (20/62) disagreed on the answer to the first question and 34% (11/32) disagreed on the answer to the second question. Out of 59 centers with a standardized antibiotic protocol, 42 (71%) reported monitoring compliance. The estimated compliance rates ranged from 70% to 100% with 22 centers reporting 100% compliance. Among 34 centers stating that physicians prescribe antibacterial prophylaxis to ECMO patients outside of a protocol, 12 estimated that 100% of physicians prescribe antibacterial prophylaxis. Nine centers reported antifungal prophylaxis outside of a protocol, with ranges of physician use from 2% to 100%. Fifteen respondents answered that their center should adopt a standardized protocol.

Back to Top | Article Outline

Antibacterial Regimens.

Forty-two percent of centers routinely administer antibacterial prophylaxis to all ECMO patients, although there was disagreement among the respondents from two of the institutions. Among the 3% of centers that only administer antibacterial prophylaxis to a subset of patients, the main types were cardiac patients (i.e., postcardiotomy, open sternum, and congenital heart disease). Several of the listed indications were more for empiric coverage than prophylaxis (i.e., suspected sepsis).

Among centers citing a standardized protocol or antibacterial prophylaxis based on physician preference, less than half use a single agent. The classes and distribution of centers using those antibiotics as single agents or in combination are listed in Table 1. No centers with a standardized protocol reported using carbapenems, tigecycline, or linezolid as prophylaxis, whereas 14% without a protocol reported using carbapenems.

The majority of centers reporting routine antibacterial prophylaxis prescribe antibiotics for the duration of ECMO (21% with a protocol). The duration for the other centers with a protocol included precannulation and for 24 hours (8%), a limited number of days (6%), and precannulation only (4%). Six percent listed other conditions that determine duration of therapy, including open chests, presence of drains or chest tubes, underlying diagnosis, and time until negative “rule-out” cultures.

Back to Top | Article Outline

Antifungal Regimens.

A minority of centers reported administering routine (2%) or selective antifungal agents (13%) to ECMO patients. Only one center reported using a triazole (i.e., fluconazole or voriconazole) for routine prophylaxis and one center reported using nystatin. Fifteen centers reported selectively administering antifungal prophylaxis for indications such as prolonged or broad-spectrum antibacterial coverage, posttransplantation, immunocompromise, and open chests. The majority of centers prescribing antifungal prophylaxis do so dependent on the duration of ECMO or the duration of antibacterial prophylaxis, whereas a few centers prescribe antifungals for a limited number of days (5–10 days).

Back to Top | Article Outline

Infection Surveillance

A total of 64 centers reported performing surveillance cultures. Sites that are cultured include blood (100%), sputum (48%), urine (44%), and other (9%) (i.e., wound, cannulation site, circuit cultures, throat, and rectal swabs). The periodicity of surveillance cultures varied every 24 hours (42%), every 48 hours (16%), every 72 hours (25%), once a week (2%), on specific days of the week or during the ECMO run (9%), only on admission (2%), and per physician preference (3%).

Back to Top | Article Outline

Willingness to Participate in a Prospective Study

Seventy-four centers (56%) were willing to participate in a prospective study on antibiotic prophylaxis in ECMO patients.

Back to Top | Article Outline


The results of our survey demonstrated that there is marked variability between ELSO centers in antimicrobial prophylaxis and infection surveillance practices. Although the majority of respondents cited that antibiotic prophylaxis is administered to ECMO patients at their centers, <50% of all respondents had or knew of a standardized protocol at their institution. Even within single centers, there was wide variability in responses, with respondents either describing physician-dependent practices or having discordant answers regarding reported standardized practices.

Multiple practice surveys have demonstrated wide variability in physician practices in critically ill patients, particularly with respect to antimicrobial prophylaxis and therapy,18–20 even when evidence-based guidelines exist. The effect of protocols implementing evidence-based guidelines in critically ill patients to reduce variability and to improve outcome is controversial. Although before and after studies have reported improved mortality as a result of education about and implementation of evidence-based protocols (i.e., for diagnosis and management of sepsis), these studies cannot determine whether a cause-and-effect relationship exists.21 When tested in multicenter prospective trials, multifaceted change strategies (i.e., to improve nutritional adequacy or to decrease time to extubation) can improve adherence to evidence-based guidelines but may not improve outcome (i.e., mortality).22,23

Although routine antibiotic prophylaxis is recommended for ECMO patients,24 there are no specific guidelines, and there is a lack of high-quality evidence on which to devise such guidelines. There are no randomized trials evaluating infection prevention strategies in ECMO patients. Furthermore, there are no observational cohort studies comparing different antibiotic regimens on the incidence of infectious complications or on the microbiology of nosocomial pathogens. Comparison of center case series whereby standardized antimicrobial regimens are used is difficult due to heterogeneity in patient populations and cointerventions. Multicenter registries contain insufficient data about individual center infection prevention practices, limiting their use in guiding practice recommendations.

From an epidemiology standpoint, appropriate antibiotic prophylaxis should target known nosocomial pathogens in ECMO patients. Based on observational studies, the microbiology of nosocomial infections in both ECMO patients and critically ill non-ECMO patients has changed over time. Studies from the late 1990s and early 2000s reported bacterial nosocomial infections in ECMO patients to be predominantly caused by gram-positive cocci such as coagulase-negative Staphylococci.3,8 Although coagulase-negative Staphylococci are still the most predominantly cultured organism in ECMO patients,13 more recent studies have reported an increased percentage of infections due to gram-negative bacilli.9,14 An increased use of prophylactic glycopeptides such as vancomycin may account in part for this trend.9 In a series of more than 500 adult ECMO patients at an institution using routine vancomycin prophylaxis, the predominant organisms accounting for bloodstream infections were gram-negative bacilli and Candida.14 Moreover, a recent review of the ELSO registry reported that species of Candida were the most common causal organism for nosocomial infections in pediatric and adult ECMO patients.13 Among the respondents to our survey, routine antibacterial but not antifungal prophylaxis was common. There was no predominant antibacterial regimen—a combination regimen was used more frequently than a single antibiotic for prophylaxis. Centers without protocols were more likely to use combination regimens and broader spectrum antibiotics such as carbapenems than those with protocols. One goal of an antimicrobial prophylaxis protocol for ECMO patients would be to advocate for antibiotic stewardship and to minimize infections from multidrug-resistant organisms.

Routine surveillance for nosocomial infections among ECMO patients is advocated by some clinicians due to the unreliability of usual signs and symptoms and the subsequent difficulties in diagnosing sepsis. No randomized trials of routinely sending surveillance cultures have been reported. Observational studies using routine surveillance cultures may be subject to detection bias, and none of these studies compare the effect of different infection diagnosis strategies on mortality. Elerian et al. performed a retrospective four-center review of routine daily blood cultures surveillance and every other day tracheal aspirates and urine cultures in neonatal ECMO patients. Neither routine blood or urine cultures were useful in managing ECMO patients, but tracheal aspirates were helpful in guiding antibiotic therapy in neonates on ECMO more than 5 days, particularly when a chest radiograph could not exclude pneumonia.6 Steiner et al.8 suggested surveillance blood cultures beginning on the 10th day of ECMO due to the associated increase in bacteremia. Two single-center studies reported results with daily surveillance blood cultures in the absence of routine continuous antibiotic prophylaxis.4,10 Kaczala et al. reported that biological markers were not predictive of a nosocomial infection and that the incidence of bloodstream infection in the absence of antibiotic prophylaxis was higher than ELSO reports; thus, the authors supported the use of routine daily blood cultures when antibiotic prophylaxis is not used.10 Once again, wide variability existed among survey respondents, consistent with the literature. Despite the lack of evidence, survey respondents were most likely to perform routine blood cultures.

The survey had limitations in that no outcome data were obtained. Detailed patient-level data would have been required to link risk-adjusted infectious complications to antimicrobial prophylaxis and infection surveillance practices. Moreover, the survey was not comprehensive with respect to all possible infection prevention practices—i.e., targeted antimicrobial prophylaxis based on admission nasal swabs, central venous catheter management, use of sterile techniques during insertion of cannulas, management of mechanical complications, etc. Finally, the survey was based on directors' and coordinators' knowledge of ECMO practices, which may not be accurate as suggested by the high degree of discordance among respondents from the same centers. However, the survey is the first to demonstrate and to attempt to quantify the range of practice patterns among ELSO centers, and the response rate by institution was high.

With the recent H1N1 outbreak, there have been expanding indications and use of ECMO with minimal high-quality evidence to guide management, particularly with respect to infection prevention and diagnosis. Given the morbidity and mortality of nosocomial infections in these patients, there is a need to perform rigorous trials to determine the most effective and cost-effective strategies for preventing and diagnosing nosocomial infections, and to determine their impact on the emergence of multidrug-resistant pathogens. Data on antimicrobial prophylaxis regimens using standardized definitions should be prospectively collected to evaluate their impact on the incidence, timing, and microbiology of infectious complications. Furthermore, randomized trials should be performed to evaluate the effectiveness of different durations and spectrums of antibiotics on complications and mortality. Prospective trials among ELSO centers should be considered given the high proportion willing to participate.

Back to Top | Article Outline


Supported by a grant from the Extracorporeal Life Support Organization (ELSO).

Back to Top | Article Outline


1. Schutze GE, Heulitt MJ: Infections during extracorporeal life support. J Pediatr Surg 30: 809–812, 1995.
2. Coffin SE, Bell LM, Manning M, Polin R: Nosocomial infections in neonates receiving extracorporeal membrane oxygenation. Infect Control Hosp Epidemiol 18: 93–96, 1997.
3. Burket JS, Bartlett RH, Vander Hyde K, Chenoweth CE: Nosocomial infections in adult patients undergoing extracorporeal membrane oxygenation. Clin Infect Dis 28: 828–833, 1999.
4. Brown KL, Ridout DA, Shaw M, et al: Healthcare-associated infection in pediatric patients on extracorporeal life support: The role of multidisciplinary surveillance. Pediatr Crit Care Med 7: 546–550, 2006.
5. Montgomery VL, Strotman JM, Ross MP: Impact of multiple organ system dysfunction and nosocomial infections on survival of children treated with extracorporeal membrane oxygenation after heart surgery. Crit Care Med 28: 526–531, 2000.
6. Elerian LF, Sparks JW, Meyer TA, et al: Usefulness of surveillance cultures in neonatal extracorporeal membrane oxygenation. ASAIO J 47: 220–223, 2001.
7. O'Neill JM, Schutze GE, Heulitt MJ, et al: Nosocomial infections during extracorporeal membrane oxygenation. Intensive Care Med 27: 1247–1253, 2001.
8. Steiner CK, Stewart DL, Bond SJ, et al: Predictors of acquiring a nosocomial bloodstream infection on extracorporeal membrane oxygenation. J Pediatr Surg 36: 487–492, 2001.
9. Hsu MS, Chiu KM, Huang YT, et al: Risk factors for nosocomial infection during extracorporeal membrane oxygenation. J Hosp Infect 73: 210–216, 2009.
10. Kaczala GW, Paulus SC, Al-Dajani N, et al: Bloodstream infections in pediatric ECLS: Usefulness of daily blood culture monitoring and predictive value of biological markers. The British Columbia experience. Pediatr Surg Int 25: 169–173, 2009.
11. Douglass BH, Keenan AL, Purohit DM: Bacterial and fungal infection in neonates undergoing venoarterial extracorporeal membrane oxygenation: An analysis of the registry data of the Extracorporeal Life Support Organization. Artif Organs 20: 202–208, 1996.
12. Brogan TV, Thiagarajan RR, Rycus PT, et al: Extracorporeal membrane oxygenation in adults with severe respiratory failure: A multi-center database. Intensive Care Med 35: 2105–2114, 2009.
13. Bizzarro MJ, Conrad SA, Kaufman DA, Rycus P: Infections acquired during extracorporeal membrane oxygenation in neonates, children, and adults. Pediatr Crit Care Med 2010 (in press).
14. Sun HY, Ko WJ, Tsai PR, et al: Infections occurring during extracorporeal membrane oxygenation use in adult patients. J Thorac Cardiovasc Surg 140: 1125.e2–1132.e2, 2010.
15. Bratzler DW, Houck PM: Antimicrobial prophylaxis for surgery: An advisory statement from the National Surgical Infection Prevention Project. Clin Infect Dis 38: 1706–1715, 2004.
16. Bratzler DW, Hunt DR: The surgical infection prevention and surgical care improvement projects: National initiatives to improve outcomes for patients having surgery. Clin Infect Dis 43: 322–330, 2006.
17. Hocker JR, Wellhausen SR, Ward RA, et al: Effect of extracorporeal membrane oxygenation on leukocyte function in neonates. Artif Organs 15: 23–28, 1991.
18. Aarts MA, Granton J, Cook DJ, et al: Empiric antimicrobial therapy in critical illness: Results of a surgical infection society survey. Surg Infect (Larchmt) 8: 329–336, 2007.
19. Corona A, Bertolini G, Ricotta AM, et al: Variability of treatment duration for bacteraemia in the critically ill: A multinational survey. J Antimicrob Chemother 52: 849–852, 2003.
20. Djurkovic S, Baracaldo JC, Guerra JA, et al: A survey of clinicians addressing the approach to the management of severe sepsis and septic shock in the United States. J Crit Care 25: 658.e1–6, 2010.
21. Levy MM, Dellinger RP, Townsend SR, et al: The Surviving Sepsis Campaign: Results of an international guideline-based performance improvement program targeting severe sepsis. Crit Care Med 38: 367–374, 2010.
22. Doig GS, Simpson F, Finfer S, et al: Effect of evidence-based feeding guidelines on mortality of critically ill adults: A cluster randomized controlled trial. JAMA 300: 2731–2741, 2008.
23. Robertson TE, Mann HJ, Hyzy R, et al: Multicenter implementation of a consensus-developed, evidence-based, spontaneous breathing trial protocol. Crit Care Med 36: 2753–2762, 2008.
24. Van Meurs K, Lally KP, Peek G, Zwischenberger JB (eds): ECMO Extracorporeal Cardiopulmonary Support in Critical Care, 3rd ed. Ann Arbor, MI, Extracorporeal Life Support Organization, 2005.
Back to Top | Article Outline

Cited Here...

Copyright © 2011 by the American Society for Artificial Internal Organs