Nosocomial Bloodstream Infections in Patients Receiving Extracorporeal Life Support: Variability in Prevention Practices: A Survey of the Extracorporeal Life Support Organization Members
Journal of Intensive Care Medicine 1-6 ª The Author(s) 2015 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0885066615571540 jic.sagepub.com
Lily B. Glater-Welt, MD1,2, James B. Schneider, MD1,2, Marcia M. Zinger, CP, NP1,2, Lisa Rosen, ScM2,3, and Todd M. Sweberg, MD1,2
Abstract Nosocomial blood stream infections (BSIs) increase both the morbidity and the mortality of patients receiving extracorporeal life support (ECLS). The aim of this study was to identify common practices for blood stream infection prevention among national Extracorporeal Membrane Oxygenation (ECMO) programs. An electronic survey that comprised of a 16-item questionnaire was sent out to all ECMO program directors and coordinators within the United States that are part of the Extracorporeal Life Support Organization (ELSO) registry. A total of 152 institutions in 40 states were surveyed, with 85 (55%) responses. Onequarter of the institutions responded that an ECMO infection-prevention bundle or checklist was used during the cannulation. Less than half responded that an ECMO infection-prevention bundle or checklist was used for cannula maintenance, although a majority (82.9%) of institutions responded that a ‘‘standard approach to cannula dressings’’ was used. Half of the respondents reported antimicrobial prophylaxis was routinely prescribed for patients on ECMO, although specific regimens varied widely. Of the institutions, 34.2% reported sending daily blood cultures as part of routine surveillance. Smaller programs were more likely to send daily surveillance blood cultures (58.8%, P < .01). We found no clear consensus on practices used to prevent BSI in patients receiving ECMO. Keywords extracorporeal membrane oxygenation (ECMO), extracorporeal life support (ECLS), blood stream infections (BSI), catheterassociated blood stream infections (CA-BSI), nosocomial infection
Introduction Extracorporeal life support (ECLS) provides life-saving therapy to critically ill patients by supporting reversible cardiac and pulmonary failure.1 The use of ECMO has been increasing annually and now is employed for over 4000 patients per year.2 Extracorporeal membrane oxygenation (ECMO) patients are exposed to the same risk factors for infection as other critically ill patients requiring invasive monitoring and interventions, such as endotracheal tubes and vascular catheters.3 In particular, the extracorporeal circuit has multiple portals of entry for microorganisms which put these patients at increased risk of significant morbidities including nosocomial blood stream infections (BSI).4 Further, severe BSIs and sepsis are often difficult to identify in patients receiving ECMO, since most of the clinical indicators clinicians rely on are altered by the circuit.5,6 To our knowledge, there are currently no data describing the variability in infection prevention practices among the Extracorporeal Life Support Organization (ELSO) community.
Utilizing a survey of national ELSO registry members, we set out to examine the common practices of ELSO centers regarding insertion and maintenance of cannulas, including antibacterial prophylaxis and infection surveillance, with their associated infection rates. We hoped to provide a basis to
1 Division of Critical Care Medicine, Steven & Alexandra Cohen Children’s Medical Center of New York, New Hyde Park, NY, USA 2 North Shore-Long Island Jewish Health System, Glen Oaks, NY, USA 3 Biostatistics Unit, Feinstein Institute for Medical Research, Manhasset, NY, USA
Received October 31, 2014, and in revised form December 22, 2014. Accepted for publication December 29, 2014. Corresponding Author: Lily B. Glater-Welt, Department of Pediatrics, Critical Care Medicine, Steven & Alexandra Cohen Children’s Medical Center of New York, NSLIJHS, 269-01 76th Ave, Rm 222, New Hyde Park, NY 10040, USA. Email: [email protected]
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
2
Journal of Intensive Care Medicine
support the development of future large-scale prospective studies on this important topic as well as devise our own institutional protocol to minimize morbidity and mortality associated with nosocomial BSIs in this particularly vulnerable patient population.
Material and Method We conducted a survey of the ECLS directors of institutions within the United States that belong to the ELSO registry using a Web-based questionnaire. The survey was developed by the investigators using a modified Delphi technique and consisted of 16 questions regarding insertion care, daily blood culture practices, antimicrobial prophylaxis practices as well as unit demographics (Appendix A). The survey was sent electronically to ECMO program medical directors and coordinators at 152 institutions. Each ECMO center received 1 initial e-mail as well as 2 follow-up e-mails prior to closure of the survey. One response per institution was recorded and completion of the survey constituted study consent. This study was approved by the institutional review board of the North Shore LIJ Health System.
Statistical Analysis The primary outcome variable was a 16-question investigatordeveloped questionnaire. Descriptive statistics (frequencies and percents) were calculated to describe the responses of each question. The chi-square test or Fisher Exact test, as appropriate, was used to compare the frequency distribution of responses between groups of interest. Spearman correlation coefficient was utilized to compare relationships between variables.
Results A total of 85 (55%) institutions completed the survey. The majority of institutions surveyed reported their program to provide neonatal pulmonary (81.2%), pediatric pulmonary (75.3%), pediatric cardiac (72.9%), or neonatal cardiac support (71.8%; Table 1). Approximately one-quarter (27%) of institutions responded that an ECMO infection-prevention bundle or checklist was used during cannulation. Less than half (41.7%) of the institutions responded that an ECMO infection-prevention bundle or checklist was used for cannula maintenance, although a majority (82.9%) of institutions responded that a ‘‘standard approach to cannula dressings’’ was used. There were no significant associations between using a cannula placement bundle or checklist and program size, 1-10 patients per year: 17.7% (95% confidence interval [CI]: 3.8%, 43.4%), 11-20: 33.3% (14.6%, 57.0%), 21-30: 23.5% (6.8%, 49.9%), >30: 30.0% (14.7%, 49.4%), (P < 0.7), nor using a cannula maintenance bundle or checklist and program size, 1 to 10 patients per year: 35.3% (95% CI: 14.2%, 61.7%), 11-20: 42.9% (21.8%, 66.0%),
Table 1. Characteristics of Respondent Institutions.a Program center/type Neonatal pulmonary center Neonatal cardiac center Neonatal ECPR center Pediatric pulmonary center Pediatric cardiac center Pediatric ECPR center Adult pulmonary center Adult cardiac center Adult ECPR center Program size (cases per year) 1-10 11-20 21-30 >30 571540 a
% 81.2 71.8 41.2 75.3 72.9 43.5 37.6 42.4 21.2 20.0 24.7 20.0 35.3
n ¼ 85.
21-30: 47.1% (23.0%, 72.2%), >30: 41.4% (23.5%, 61.1%)] (p < 0.9). Respondents identified semipermeable dressings (bioclusive) as the most frequently used (57.3%) dressing to cover the cannulation sites, with most institutions (86.6%) responding that cannula site did not affect their choice of dressing. Other site dressings in use include gauze (51.2%), steri-strips (1.2%), and tape (12.2%), either alone or in combination (31.7%). When entering the ECMO circuit, the majority (48.2%) of centers use alcohol to disinfect the access ports. Other port disinfectants utilized are chlorhexadine (38.8%) and betadine (4.7%). A majority (60.5%) of centers responded that the cannulation site dressing was only changed when necessary (Table 2). Half (50%) of the respondents reported antimicrobial prophylaxis was routinely prescribed for patients on ECMO, although the specific regimens varied widely (Table 3). Most (89%) institutions reported that cannulation site did not alter their prophylactic antibiotic regimen. Of the respondents, 11% reported specific antibiotic prophylaxis for patients with transthoracic cannulation through open chests although regimens varied. The association between prophylactic antibiotic use and program size was not significant. The use of at least 1 prophylactic antibiotic was highest in the smallest programs, 1 to 10 patients per year: 70.6% (95% CI: 44.0%, 89.7%), 11 to 20: 45.0% (23.1%, 68.5%), 21 to 30: 40.0% (16.3%, 67.7%), and >30: 46.7% (28.3%, 65.7%); however, this was not statistically significant. About one-third (34.2%) of institutions reported sending daily blood cultures as a part of routine surveillance (Table 4). There was a significant association between sending daily blood cultures and program size (P < .01), with smaller programs more likely to send daily surveillance blood cultures, 1 to10 patients per year (58.8%, 95% CI: 32.9%, 81.6%), 11 to 20 patients per year (45%, 95% CI: 23.1%, 68.5%), and >30 patients per year (26.7%, 95% CI: 12.3%, 45.9%). Program sizes with 21 to 30 ECMO patients per year had the lowest
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
Glater-Welt et al
3
Table 2. Cannula Maintenance. (A) Dressing changes Frequency of Dressing Change
Frequencya
%
3
3.70
21 7 5 49
25.93 8.64 6.17 60.49
Frequency
%
13
15.29
4 33 41 0 6
4.71 38.82 48.24 0.00 7.06
Our institution does not use a dressing over the cannulation site Every 24 h Every 48 h Every 72 h Only when necessary (ie, bloody, not adhering well) (B) Surface disinfection Type Our institution does not routinely access circuit ports Betadine Chlorhexadine Alcohol Nothing A combination of the above a
n ¼ 4 missing.
Table 3. Antibacterial Prophylaxis. Type None Ampicillin þ aminoglycoside (gentamicin or amikacin) Ampicillin þ cephalosporin Vancomycin þ aminoglycoside (gentamicin or amikacin) Vancomycin þ cephalosporin Vancomycin Cephalosporin Other a
Frequencya
%
41 14
50.00 17.07
3 4
3.66 4.88
4 6 11 7
4.88 7.32 13.41 8.54
n ¼ 3 missing.
Table 4. Infection Surveillance. Daily blood cultures Yes No
% 34.1 65.9
percentage of daily blood cultures (6.7%, 95% CI: 0.2%, 31.9%). Of the 85 institutions who participated in our survey, 57 responded to the optional question requesting their program’s rate of BSI (% per year ¼ # culture proven infections/# cases per year). The rates given ranged between 0% and 18%, with a median rate of 0.05%. There were a total of 28 (32.9%) participants who response was ‘‘unknown’’ or the question was not answered. Among the 57 responses, there was a significant correlation between the size of the program and BSI rates; as the size of the program increased, the rate of
BSI increased (Spearman correlation coefficient r ¼ .52973; P < .01). When limiting the analysis to those programs reporting BSI rates, there was no significant association between BSI rates and whether the institution utilized an ECMO infection prevention bundle/checklist for cannula placement, or whether the institution used a cannula maintenance bundle (Fisher Exact, P < .3997). Further, no significant association was found between reported BSI rates and whether protective gear was worn during port access, type of disinfectant used when cleaning ports, dressing type, cannula site, or use of prophylactic antibiotics (regardless of the reported antibiotics used).
Discussion The results of our survey highlight the variability in practice regarding prevention of bloodstream infections for patients requiring ECLS. Although there are definite trends in antimicrobial prophylaxis and infection surveillance, consensus on infection control practices is still lacking. In 2003, Odetola et al reported the startling frequency of catheter-related BSIs (CR-BSIs) in this patient population.7 The ECLS population had a 10-fold increased risk of developing CR-BSI, higher than that observed in patients with central venous catheters inserted for other indications. Culture proven infection rates reported by the ELSO registry in January 2014 range from 6% to 18%, depending upon age-group and diagnosis. Nosocomial BSI leads to prolonged duration of ECLS and increased risk of mortality.1,6,8,9 In 2011, Kao et al published the results of a large, 132 center survey describing antimicrobial prophylaxis and infection surveillance in ECMO patients. The results of this comprehensive survey demonstrated significant variability in practice patterns among ELSO centers with specific focus on antibiotic use and frequency of surveillance cultures. However, this report did not discuss measures for infection prevention on ECMO, such as cannula maintenance and insertion practices. In the last decade, focus has been placed upon the prevention of hospital-acquired conditions (HACs), including catheter-associated BSIs (CA-BSIs). Importantly, rates of these HACs have been improving due to the implementation of particular ‘‘bundled’’ care interventions. Bundles have been defined as specific practices essential for safe patient care and have been utilized for insertion and maintenance of central venous catheters as well as the prevention of many other HACs. In 2008, the ‘‘ELSO Infectious Disease (ID) Task Force’’ produced recommendations based upon extensive literature review and expert consensus, namely, that the ECMO circuit should be treated as a ‘‘protected central line.’’ Their recommendations included cleaning the access ports with chlorhexadine rather than alcohol or betadine, frequent hand washing, and removal of unnecessary central access.3 Despite the demonstrated success of bundled care interventions in reducing CA-BSI, and the recommendations from the ELSO ID
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
4
Journal of Intensive Care Medicine
Task Force, it is unclear whether a standardized approach to prevention of BSI in patients receiving ECLS is being employed. In addition, the efficacy of these recommendations has not been prospectively studied to our knowledge. A majority of ELSO centers who participated in our survey are using alcohol as a disinfectant rather than the recommended chlorhexadine. In addition, the ELSO guidelines discourage the routine use of prophylactic antibiotics (excluding open chest patients) and surveillance cultures.3 Despite these suggestions, half of the responding programs use antimicrobial prophylaxis and 34% send surveillance blood cultures. Previous surveys have documented 74% of ELSO centers use antimicrobial prophylaxis, higher than the 50% found in this cohort.10 Of the centers, 34% send daily blood cultures, despite evidence questioning their clinical utility.6,10 In 2010, the National Association of Children’s Hospitals and Related Institutions (NACHRI) published the results of their pediatric intensive care unit (PICU) CA-BSI collaborative which demonstrated significant, sustained decreases in pediatric CA-BSI in the PICU setting when implementing central line insertion and maintenance bundles.11 Although there have been a number of other studies in pediatric and adult patients documenting sustained reductions in central line associated BSIs with the use of a care bundle, less than half of centers utilize a similar bundle for patients receiving ECMO.9,12-15 Only a quarter of respondents to our survey reported use of a bundle or checklist during ECMO cannula insertion and less than half utilize a bundle or checklist for cannula maintenance. Brown et al were able to reduce ECMO infections from 29.3 to 20.1 episodes per 1000 ECMO days in their study population using a multidisciplinary surveillance program.9 Although techniques used to minimize hospitalacquired infections may benefit ECMO patients, it is clear that despite prior efforts, a consensus approach to BSI control on ECMO has not been established.16
Conclusion The variability in practice this survey reveals presents an opportunity for practice improvement across the ECMO community. We hope that the common clinical practices found among ELSO centers will enable us to propose an infection prevention ‘‘bundle’’ for this patient population in our own institution to lay the groundwork for development of future prospective randomized controlled clinical studies to evaluate use of standardized bundles. In conclusion, our results demonstrate significant variability and a lack of consensus on BSI prevention in patients receiving ECMO. Of particular note is the inconsistency in practice among the ECLS community with regard to antibiotic prophylaxis and surveillance blood cultures which is consistent with previous studies. In an era of significant effort dedicated to the prevention of HACs, a prospective study is warranted to evaluate the application of proven infection control methods to the ECMO population.
Appendix A ELSO Survey: Preventing Blood Stream Infections in Patients Receiving ECLS 1)
2)
Limitations As a survey, our results are limited and subjective. Our response rate of 55% may have limited the ability to identify all practice differences that exist in the ELSO community. Further, due to limitations in obtaining program data for all recipients of the survey, we were unable to describe and compare programs that responded versus those who did not respond. Some institutions chose not to answer all questions which made some of our results incomplete. In addition, the questions in the survey did not delve into other important infection prevention practices such as central venous catheter management and use of biomarkers for diagnosis of infection.17 Finally, the survey was comprised of questions based on ECMO experience at our institution, which differs from some of the respondent institutions. This includes differences in terminology, such as ‘‘bundle,’’ which could be interpreted differently based upon local practice.
3)
4)
5)
How would you categorize your program? (check all that apply) i. Neonatal pulmonary center ii. Neonatal cardiac center iii. Neonatal ECPR iv. Pediatric pulmonary center v. Pediatric cardiac center vi. Pediatric ECPR center vii. Adult pulmonary center viii. Adult cardiac center ix. Adult ECPR center How many patients are placed on ECMO (per year) at your institution? i. 1-10 ii. 11-20 iii. 21-30 iv. >30 Does your institution utilize an ECMO infection prevention ‘‘bundle’’/checklist for cannula placement? i. Yes ii. No Does your institution utilize an ECMO infection prevention ‘‘bundle’’/checklist for cannula maintenance? i. Yes ii. No When accessing ports on the ECMO circuit (i.e., drawing blood or administering medications) do you wear: (check all that apply) i. Our institution does not routinely access circuit ports ii. Hat iii. Mask iv. Gloves
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
Glater-Welt et al
6)
7)
8)
9)
5
v. Gown vi. All of the above vii. None of the above What do you clean the access ports with prior to accessing them? i. Our institution does not routinely access circuit ports ii. Betadine iii. Chlorhexadine iv. Alcohol v. Nothing vi. A combination of the above What type of dressing (if any) do you use to cover/secure the cannulation site? (femoral & neck sites) i. Gauze ii. Bioclusive dressing (i.e., Tegaderm) iii. Steri-strips iv. Tape v. A combination of the above vi. Other vii. None Does cannula site affect your choice of dressing(s)? i. Yes ii. No If you answered ‘‘yes’’ to question #8, please describe how.
13)
Does cannulation site alter your prophylactic antibiotic regimen? i. Yes ii. No 14) If you answered ‘‘yes’’ to question #13, please describe how cannula site affects your prophylaxis choice.
15)
Do you send daily blood cultures as part of your routine surveillance? i. Yes ii. No 16) What is your BSI rate for ECMO patients at your institution? (% per year ¼ # culture proven infection(s) divided by # of cases per year)
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.
References 10)
Do you utilize a standard approach to cannula dressing(s) (i.e., the same for all patients regardless of practitioner)? i. Yes ii. No 11) How often is the dressing over the cannulation site changed? (femoral and neck sites) i. Our institution does not use a dressing over the cannulation site ii. Every 24 h iii. Every 48 h iv. Every 72 h v. Only when necessary (i.e., bloody, not adhering well) 12) What antibacterial prophylaxis do you place your ECMO patients on? i. None ii. Ampicillin þ Aminoglycoside (Gentamicin or Amikacin) iii. Ampicillin þ Cephalosporin iv. Vancomycin þ Aminoglycoside (Gentamicin or Amikacin) v. Vancomycin þ Cephalosporin vi. Vancomycin vii. Cephalosporin viii. Other
1. Bizzarro MJ, Conrad SA, Kaufman DA, Rycus P. Infections acquired during extracorporeal membrane oxygenation in neonates, children, and adults. Pediatr Crit Care Med. 2011;12(3): 277. 2. Extracorporeal life support organization: ECLS registry report international summary. Updated January 2014. 3. Annich G, Lynch W, MacLaren G, Wilson J, Bartlett R. ECMO: Extracorporeal cardiopulmonary support in critical care. 4th ed. Ann Arbor, Michigan: Extracorporeal Life Support Organization; 2012. 4. Tse-Chang A, Midodzi W, Joffe AR, Robinson JL. Infections in children receiving extracorporeal life support. Infect Control Hosp Epidemiol. 2011;32(2):115-120. 5. 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. 2009;25(2):169-173. 6. Elerian LF, Sparks JW, Meyer TA, et al. Usefulness of surveillance cultures in neonatal extracorporeal membrane oxygenation. ASAIO J. 2001;47(3):220-223. 7. Odetola FO, Moler FW, Dechert RE, VanDerElzen K, Chenoweth C. Nosocomial catheter-related bloodstream infections in a pediatric intensive care unit: Risk and rates associated with various intravascular technologies. Pediatr Crit Care Med. 2003;4(4): 432-436.
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
6
Journal of Intensive Care Medicine
8. Steiner CK, Stewart DL, Bond SJ, Hornung CA, McKay VJ. Predictors of acquiring a nosocomial bloodstream infection on extracorporeal membrane oxygenation. J Pediatr Surg. 2001;36(3):487-492. 9. 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. 2006;7(6):546-550. 10. Kao LS, Fleming GM, Escamilla RJ, Lew DF, Lally KP. Antimicrobial prophylaxis and infection surveillance in extracorporeal membrane oxygenation patients: a multi-institutional survey of practice patterns. ASAIO J. 2011;57(3):231. 11. Miller MR, Griswold M, Harris JM, et al. Decreasing PICU catheter-associated bloodstream infections: NACHRI’s quality transformation efforts. Pediatrics. 2010;125(2):206-213. 12. Miller MR, Niedner MF, Huskins WC, et al. Reducing PICU central line–associated bloodstream infections: 3-year results. Pediatrics. 2011;128(5):e1077-e1083.
13. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. 14. Rinke ML, Chen AR, Bundy DG, et al. Implementation of a central line maintenance care bundle in hospitalized pediatric oncology patients. Pediatrics. 2012;130(4):e996-e1004. 15. Schulman J, Stricof R, Stevens TP, et al. Statewide NICU centralline-associated bloodstream infection rates decline after bundles and checklists. Pediatrics. 2011;127(3):436-444. 16. Rehder KJ, Turner DA, Cheifetz IM. Extracorporeal membrane oxygenation for neonatal and pediatric respiratory failure: an evidence-based review of the past decade (2002-2012). Pediatr Crit Care Med. 2013;14(9):851-861. 17. Pieri M, Greco T, De Bonis M, et al. Diagnosis of infection in patients undergoing extracorporeal membrane oxygenation: a case-control study. J Thorac Cardiovasc Surg. 2012;143(6): 1411-1416. e1.
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
Journal of Intensive Care Medicine 1-6 ª The Author(s) 2015 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0885066615571540 jic.sagepub.com
Lily B. Glater-Welt, MD1,2, James B. Schneider, MD1,2, Marcia M. Zinger, CP, NP1,2, Lisa Rosen, ScM2,3, and Todd M. Sweberg, MD1,2
Abstract Nosocomial blood stream infections (BSIs) increase both the morbidity and the mortality of patients receiving extracorporeal life support (ECLS). The aim of this study was to identify common practices for blood stream infection prevention among national Extracorporeal Membrane Oxygenation (ECMO) programs. An electronic survey that comprised of a 16-item questionnaire was sent out to all ECMO program directors and coordinators within the United States that are part of the Extracorporeal Life Support Organization (ELSO) registry. A total of 152 institutions in 40 states were surveyed, with 85 (55%) responses. Onequarter of the institutions responded that an ECMO infection-prevention bundle or checklist was used during the cannulation. Less than half responded that an ECMO infection-prevention bundle or checklist was used for cannula maintenance, although a majority (82.9%) of institutions responded that a ‘‘standard approach to cannula dressings’’ was used. Half of the respondents reported antimicrobial prophylaxis was routinely prescribed for patients on ECMO, although specific regimens varied widely. Of the institutions, 34.2% reported sending daily blood cultures as part of routine surveillance. Smaller programs were more likely to send daily surveillance blood cultures (58.8%, P < .01). We found no clear consensus on practices used to prevent BSI in patients receiving ECMO. Keywords extracorporeal membrane oxygenation (ECMO), extracorporeal life support (ECLS), blood stream infections (BSI), catheterassociated blood stream infections (CA-BSI), nosocomial infection
Introduction Extracorporeal life support (ECLS) provides life-saving therapy to critically ill patients by supporting reversible cardiac and pulmonary failure.1 The use of ECMO has been increasing annually and now is employed for over 4000 patients per year.2 Extracorporeal membrane oxygenation (ECMO) patients are exposed to the same risk factors for infection as other critically ill patients requiring invasive monitoring and interventions, such as endotracheal tubes and vascular catheters.3 In particular, the extracorporeal circuit has multiple portals of entry for microorganisms which put these patients at increased risk of significant morbidities including nosocomial blood stream infections (BSI).4 Further, severe BSIs and sepsis are often difficult to identify in patients receiving ECMO, since most of the clinical indicators clinicians rely on are altered by the circuit.5,6 To our knowledge, there are currently no data describing the variability in infection prevention practices among the Extracorporeal Life Support Organization (ELSO) community.
Utilizing a survey of national ELSO registry members, we set out to examine the common practices of ELSO centers regarding insertion and maintenance of cannulas, including antibacterial prophylaxis and infection surveillance, with their associated infection rates. We hoped to provide a basis to
1 Division of Critical Care Medicine, Steven & Alexandra Cohen Children’s Medical Center of New York, New Hyde Park, NY, USA 2 North Shore-Long Island Jewish Health System, Glen Oaks, NY, USA 3 Biostatistics Unit, Feinstein Institute for Medical Research, Manhasset, NY, USA
Received October 31, 2014, and in revised form December 22, 2014. Accepted for publication December 29, 2014. Corresponding Author: Lily B. Glater-Welt, Department of Pediatrics, Critical Care Medicine, Steven & Alexandra Cohen Children’s Medical Center of New York, NSLIJHS, 269-01 76th Ave, Rm 222, New Hyde Park, NY 10040, USA. Email: [email protected]
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
2
Journal of Intensive Care Medicine
support the development of future large-scale prospective studies on this important topic as well as devise our own institutional protocol to minimize morbidity and mortality associated with nosocomial BSIs in this particularly vulnerable patient population.
Material and Method We conducted a survey of the ECLS directors of institutions within the United States that belong to the ELSO registry using a Web-based questionnaire. The survey was developed by the investigators using a modified Delphi technique and consisted of 16 questions regarding insertion care, daily blood culture practices, antimicrobial prophylaxis practices as well as unit demographics (Appendix A). The survey was sent electronically to ECMO program medical directors and coordinators at 152 institutions. Each ECMO center received 1 initial e-mail as well as 2 follow-up e-mails prior to closure of the survey. One response per institution was recorded and completion of the survey constituted study consent. This study was approved by the institutional review board of the North Shore LIJ Health System.
Statistical Analysis The primary outcome variable was a 16-question investigatordeveloped questionnaire. Descriptive statistics (frequencies and percents) were calculated to describe the responses of each question. The chi-square test or Fisher Exact test, as appropriate, was used to compare the frequency distribution of responses between groups of interest. Spearman correlation coefficient was utilized to compare relationships between variables.
Results A total of 85 (55%) institutions completed the survey. The majority of institutions surveyed reported their program to provide neonatal pulmonary (81.2%), pediatric pulmonary (75.3%), pediatric cardiac (72.9%), or neonatal cardiac support (71.8%; Table 1). Approximately one-quarter (27%) of institutions responded that an ECMO infection-prevention bundle or checklist was used during cannulation. Less than half (41.7%) of the institutions responded that an ECMO infection-prevention bundle or checklist was used for cannula maintenance, although a majority (82.9%) of institutions responded that a ‘‘standard approach to cannula dressings’’ was used. There were no significant associations between using a cannula placement bundle or checklist and program size, 1-10 patients per year: 17.7% (95% confidence interval [CI]: 3.8%, 43.4%), 11-20: 33.3% (14.6%, 57.0%), 21-30: 23.5% (6.8%, 49.9%), >30: 30.0% (14.7%, 49.4%), (P < 0.7), nor using a cannula maintenance bundle or checklist and program size, 1 to 10 patients per year: 35.3% (95% CI: 14.2%, 61.7%), 11-20: 42.9% (21.8%, 66.0%),
Table 1. Characteristics of Respondent Institutions.a Program center/type Neonatal pulmonary center Neonatal cardiac center Neonatal ECPR center Pediatric pulmonary center Pediatric cardiac center Pediatric ECPR center Adult pulmonary center Adult cardiac center Adult ECPR center Program size (cases per year) 1-10 11-20 21-30 >30 571540 a
% 81.2 71.8 41.2 75.3 72.9 43.5 37.6 42.4 21.2 20.0 24.7 20.0 35.3
n ¼ 85.
21-30: 47.1% (23.0%, 72.2%), >30: 41.4% (23.5%, 61.1%)] (p < 0.9). Respondents identified semipermeable dressings (bioclusive) as the most frequently used (57.3%) dressing to cover the cannulation sites, with most institutions (86.6%) responding that cannula site did not affect their choice of dressing. Other site dressings in use include gauze (51.2%), steri-strips (1.2%), and tape (12.2%), either alone or in combination (31.7%). When entering the ECMO circuit, the majority (48.2%) of centers use alcohol to disinfect the access ports. Other port disinfectants utilized are chlorhexadine (38.8%) and betadine (4.7%). A majority (60.5%) of centers responded that the cannulation site dressing was only changed when necessary (Table 2). Half (50%) of the respondents reported antimicrobial prophylaxis was routinely prescribed for patients on ECMO, although the specific regimens varied widely (Table 3). Most (89%) institutions reported that cannulation site did not alter their prophylactic antibiotic regimen. Of the respondents, 11% reported specific antibiotic prophylaxis for patients with transthoracic cannulation through open chests although regimens varied. The association between prophylactic antibiotic use and program size was not significant. The use of at least 1 prophylactic antibiotic was highest in the smallest programs, 1 to 10 patients per year: 70.6% (95% CI: 44.0%, 89.7%), 11 to 20: 45.0% (23.1%, 68.5%), 21 to 30: 40.0% (16.3%, 67.7%), and >30: 46.7% (28.3%, 65.7%); however, this was not statistically significant. About one-third (34.2%) of institutions reported sending daily blood cultures as a part of routine surveillance (Table 4). There was a significant association between sending daily blood cultures and program size (P < .01), with smaller programs more likely to send daily surveillance blood cultures, 1 to10 patients per year (58.8%, 95% CI: 32.9%, 81.6%), 11 to 20 patients per year (45%, 95% CI: 23.1%, 68.5%), and >30 patients per year (26.7%, 95% CI: 12.3%, 45.9%). Program sizes with 21 to 30 ECMO patients per year had the lowest
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
Glater-Welt et al
3
Table 2. Cannula Maintenance. (A) Dressing changes Frequency of Dressing Change
Frequencya
%
3
3.70
21 7 5 49
25.93 8.64 6.17 60.49
Frequency
%
13
15.29
4 33 41 0 6
4.71 38.82 48.24 0.00 7.06
Our institution does not use a dressing over the cannulation site Every 24 h Every 48 h Every 72 h Only when necessary (ie, bloody, not adhering well) (B) Surface disinfection Type Our institution does not routinely access circuit ports Betadine Chlorhexadine Alcohol Nothing A combination of the above a
n ¼ 4 missing.
Table 3. Antibacterial Prophylaxis. Type None Ampicillin þ aminoglycoside (gentamicin or amikacin) Ampicillin þ cephalosporin Vancomycin þ aminoglycoside (gentamicin or amikacin) Vancomycin þ cephalosporin Vancomycin Cephalosporin Other a
Frequencya
%
41 14
50.00 17.07
3 4
3.66 4.88
4 6 11 7
4.88 7.32 13.41 8.54
n ¼ 3 missing.
Table 4. Infection Surveillance. Daily blood cultures Yes No
% 34.1 65.9
percentage of daily blood cultures (6.7%, 95% CI: 0.2%, 31.9%). Of the 85 institutions who participated in our survey, 57 responded to the optional question requesting their program’s rate of BSI (% per year ¼ # culture proven infections/# cases per year). The rates given ranged between 0% and 18%, with a median rate of 0.05%. There were a total of 28 (32.9%) participants who response was ‘‘unknown’’ or the question was not answered. Among the 57 responses, there was a significant correlation between the size of the program and BSI rates; as the size of the program increased, the rate of
BSI increased (Spearman correlation coefficient r ¼ .52973; P < .01). When limiting the analysis to those programs reporting BSI rates, there was no significant association between BSI rates and whether the institution utilized an ECMO infection prevention bundle/checklist for cannula placement, or whether the institution used a cannula maintenance bundle (Fisher Exact, P < .3997). Further, no significant association was found between reported BSI rates and whether protective gear was worn during port access, type of disinfectant used when cleaning ports, dressing type, cannula site, or use of prophylactic antibiotics (regardless of the reported antibiotics used).
Discussion The results of our survey highlight the variability in practice regarding prevention of bloodstream infections for patients requiring ECLS. Although there are definite trends in antimicrobial prophylaxis and infection surveillance, consensus on infection control practices is still lacking. In 2003, Odetola et al reported the startling frequency of catheter-related BSIs (CR-BSIs) in this patient population.7 The ECLS population had a 10-fold increased risk of developing CR-BSI, higher than that observed in patients with central venous catheters inserted for other indications. Culture proven infection rates reported by the ELSO registry in January 2014 range from 6% to 18%, depending upon age-group and diagnosis. Nosocomial BSI leads to prolonged duration of ECLS and increased risk of mortality.1,6,8,9 In 2011, Kao et al published the results of a large, 132 center survey describing antimicrobial prophylaxis and infection surveillance in ECMO patients. The results of this comprehensive survey demonstrated significant variability in practice patterns among ELSO centers with specific focus on antibiotic use and frequency of surveillance cultures. However, this report did not discuss measures for infection prevention on ECMO, such as cannula maintenance and insertion practices. In the last decade, focus has been placed upon the prevention of hospital-acquired conditions (HACs), including catheter-associated BSIs (CA-BSIs). Importantly, rates of these HACs have been improving due to the implementation of particular ‘‘bundled’’ care interventions. Bundles have been defined as specific practices essential for safe patient care and have been utilized for insertion and maintenance of central venous catheters as well as the prevention of many other HACs. In 2008, the ‘‘ELSO Infectious Disease (ID) Task Force’’ produced recommendations based upon extensive literature review and expert consensus, namely, that the ECMO circuit should be treated as a ‘‘protected central line.’’ Their recommendations included cleaning the access ports with chlorhexadine rather than alcohol or betadine, frequent hand washing, and removal of unnecessary central access.3 Despite the demonstrated success of bundled care interventions in reducing CA-BSI, and the recommendations from the ELSO ID
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
4
Journal of Intensive Care Medicine
Task Force, it is unclear whether a standardized approach to prevention of BSI in patients receiving ECLS is being employed. In addition, the efficacy of these recommendations has not been prospectively studied to our knowledge. A majority of ELSO centers who participated in our survey are using alcohol as a disinfectant rather than the recommended chlorhexadine. In addition, the ELSO guidelines discourage the routine use of prophylactic antibiotics (excluding open chest patients) and surveillance cultures.3 Despite these suggestions, half of the responding programs use antimicrobial prophylaxis and 34% send surveillance blood cultures. Previous surveys have documented 74% of ELSO centers use antimicrobial prophylaxis, higher than the 50% found in this cohort.10 Of the centers, 34% send daily blood cultures, despite evidence questioning their clinical utility.6,10 In 2010, the National Association of Children’s Hospitals and Related Institutions (NACHRI) published the results of their pediatric intensive care unit (PICU) CA-BSI collaborative which demonstrated significant, sustained decreases in pediatric CA-BSI in the PICU setting when implementing central line insertion and maintenance bundles.11 Although there have been a number of other studies in pediatric and adult patients documenting sustained reductions in central line associated BSIs with the use of a care bundle, less than half of centers utilize a similar bundle for patients receiving ECMO.9,12-15 Only a quarter of respondents to our survey reported use of a bundle or checklist during ECMO cannula insertion and less than half utilize a bundle or checklist for cannula maintenance. Brown et al were able to reduce ECMO infections from 29.3 to 20.1 episodes per 1000 ECMO days in their study population using a multidisciplinary surveillance program.9 Although techniques used to minimize hospitalacquired infections may benefit ECMO patients, it is clear that despite prior efforts, a consensus approach to BSI control on ECMO has not been established.16
Conclusion The variability in practice this survey reveals presents an opportunity for practice improvement across the ECMO community. We hope that the common clinical practices found among ELSO centers will enable us to propose an infection prevention ‘‘bundle’’ for this patient population in our own institution to lay the groundwork for development of future prospective randomized controlled clinical studies to evaluate use of standardized bundles. In conclusion, our results demonstrate significant variability and a lack of consensus on BSI prevention in patients receiving ECMO. Of particular note is the inconsistency in practice among the ECLS community with regard to antibiotic prophylaxis and surveillance blood cultures which is consistent with previous studies. In an era of significant effort dedicated to the prevention of HACs, a prospective study is warranted to evaluate the application of proven infection control methods to the ECMO population.
Appendix A ELSO Survey: Preventing Blood Stream Infections in Patients Receiving ECLS 1)
2)
Limitations As a survey, our results are limited and subjective. Our response rate of 55% may have limited the ability to identify all practice differences that exist in the ELSO community. Further, due to limitations in obtaining program data for all recipients of the survey, we were unable to describe and compare programs that responded versus those who did not respond. Some institutions chose not to answer all questions which made some of our results incomplete. In addition, the questions in the survey did not delve into other important infection prevention practices such as central venous catheter management and use of biomarkers for diagnosis of infection.17 Finally, the survey was comprised of questions based on ECMO experience at our institution, which differs from some of the respondent institutions. This includes differences in terminology, such as ‘‘bundle,’’ which could be interpreted differently based upon local practice.
3)
4)
5)
How would you categorize your program? (check all that apply) i. Neonatal pulmonary center ii. Neonatal cardiac center iii. Neonatal ECPR iv. Pediatric pulmonary center v. Pediatric cardiac center vi. Pediatric ECPR center vii. Adult pulmonary center viii. Adult cardiac center ix. Adult ECPR center How many patients are placed on ECMO (per year) at your institution? i. 1-10 ii. 11-20 iii. 21-30 iv. >30 Does your institution utilize an ECMO infection prevention ‘‘bundle’’/checklist for cannula placement? i. Yes ii. No Does your institution utilize an ECMO infection prevention ‘‘bundle’’/checklist for cannula maintenance? i. Yes ii. No When accessing ports on the ECMO circuit (i.e., drawing blood or administering medications) do you wear: (check all that apply) i. Our institution does not routinely access circuit ports ii. Hat iii. Mask iv. Gloves
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
Glater-Welt et al
6)
7)
8)
9)
5
v. Gown vi. All of the above vii. None of the above What do you clean the access ports with prior to accessing them? i. Our institution does not routinely access circuit ports ii. Betadine iii. Chlorhexadine iv. Alcohol v. Nothing vi. A combination of the above What type of dressing (if any) do you use to cover/secure the cannulation site? (femoral & neck sites) i. Gauze ii. Bioclusive dressing (i.e., Tegaderm) iii. Steri-strips iv. Tape v. A combination of the above vi. Other vii. None Does cannula site affect your choice of dressing(s)? i. Yes ii. No If you answered ‘‘yes’’ to question #8, please describe how.
13)
Does cannulation site alter your prophylactic antibiotic regimen? i. Yes ii. No 14) If you answered ‘‘yes’’ to question #13, please describe how cannula site affects your prophylaxis choice.
15)
Do you send daily blood cultures as part of your routine surveillance? i. Yes ii. No 16) What is your BSI rate for ECMO patients at your institution? (% per year ¼ # culture proven infection(s) divided by # of cases per year)
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.
References 10)
Do you utilize a standard approach to cannula dressing(s) (i.e., the same for all patients regardless of practitioner)? i. Yes ii. No 11) How often is the dressing over the cannulation site changed? (femoral and neck sites) i. Our institution does not use a dressing over the cannulation site ii. Every 24 h iii. Every 48 h iv. Every 72 h v. Only when necessary (i.e., bloody, not adhering well) 12) What antibacterial prophylaxis do you place your ECMO patients on? i. None ii. Ampicillin þ Aminoglycoside (Gentamicin or Amikacin) iii. Ampicillin þ Cephalosporin iv. Vancomycin þ Aminoglycoside (Gentamicin or Amikacin) v. Vancomycin þ Cephalosporin vi. Vancomycin vii. Cephalosporin viii. Other
1. Bizzarro MJ, Conrad SA, Kaufman DA, Rycus P. Infections acquired during extracorporeal membrane oxygenation in neonates, children, and adults. Pediatr Crit Care Med. 2011;12(3): 277. 2. Extracorporeal life support organization: ECLS registry report international summary. Updated January 2014. 3. Annich G, Lynch W, MacLaren G, Wilson J, Bartlett R. ECMO: Extracorporeal cardiopulmonary support in critical care. 4th ed. Ann Arbor, Michigan: Extracorporeal Life Support Organization; 2012. 4. Tse-Chang A, Midodzi W, Joffe AR, Robinson JL. Infections in children receiving extracorporeal life support. Infect Control Hosp Epidemiol. 2011;32(2):115-120. 5. 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. 2009;25(2):169-173. 6. Elerian LF, Sparks JW, Meyer TA, et al. Usefulness of surveillance cultures in neonatal extracorporeal membrane oxygenation. ASAIO J. 2001;47(3):220-223. 7. Odetola FO, Moler FW, Dechert RE, VanDerElzen K, Chenoweth C. Nosocomial catheter-related bloodstream infections in a pediatric intensive care unit: Risk and rates associated with various intravascular technologies. Pediatr Crit Care Med. 2003;4(4): 432-436.
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
6
Journal of Intensive Care Medicine
8. Steiner CK, Stewart DL, Bond SJ, Hornung CA, McKay VJ. Predictors of acquiring a nosocomial bloodstream infection on extracorporeal membrane oxygenation. J Pediatr Surg. 2001;36(3):487-492. 9. 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. 2006;7(6):546-550. 10. Kao LS, Fleming GM, Escamilla RJ, Lew DF, Lally KP. Antimicrobial prophylaxis and infection surveillance in extracorporeal membrane oxygenation patients: a multi-institutional survey of practice patterns. ASAIO J. 2011;57(3):231. 11. Miller MR, Griswold M, Harris JM, et al. Decreasing PICU catheter-associated bloodstream infections: NACHRI’s quality transformation efforts. Pediatrics. 2010;125(2):206-213. 12. Miller MR, Niedner MF, Huskins WC, et al. Reducing PICU central line–associated bloodstream infections: 3-year results. Pediatrics. 2011;128(5):e1077-e1083.
13. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med. 2006;355(26):2725-2732. 14. Rinke ML, Chen AR, Bundy DG, et al. Implementation of a central line maintenance care bundle in hospitalized pediatric oncology patients. Pediatrics. 2012;130(4):e996-e1004. 15. Schulman J, Stricof R, Stevens TP, et al. Statewide NICU centralline-associated bloodstream infection rates decline after bundles and checklists. Pediatrics. 2011;127(3):436-444. 16. Rehder KJ, Turner DA, Cheifetz IM. Extracorporeal membrane oxygenation for neonatal and pediatric respiratory failure: an evidence-based review of the past decade (2002-2012). Pediatr Crit Care Med. 2013;14(9):851-861. 17. Pieri M, Greco T, De Bonis M, et al. Diagnosis of infection in patients undergoing extracorporeal membrane oxygenation: a case-control study. J Thorac Cardiovasc Surg. 2012;143(6): 1411-1416. e1.
Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on February 11, 2015
