Secondary Logo

Journal Logo

Infections and Extracorporeal Membrane Oxygenation: Incidence, Therapy, and Outcome

Haneke, Fabian*; Schildhauer, Thomas A.*; Schlebes, Alexander D.*; Strauch, Justus T.; Swol, Justyna*

doi: 10.1097/MAT.0000000000000308
Clinical Critical Care
Free

The objective is to assess the influence of infections and the microbiological spectrum on the general outcome of patients undergoing therapy with extracorporeal devices (ECDs), extracorporeal membrane oxygenation, extracorporeal life support, and pumpless extracorporeal lung assist. We performed a single-center, retrospective analysis of 99 patients receiving ECD. Infections requiring ECD, nosocomial infections occurring during treatment, the use of guideline-based antiinfective therapies, and patient outcomes were described and statistically analyzed. We analyzed 88 patients—survivors and nonsurvivors—and subdivided the infections into primary and nosocomial infections. The median patient age was 54.0 years, 85.2% were men, and 45 (51.1%) survived. Surviving ECD patients had a higher risk of nosocomial infection because of their prolonged hospital stay. Our results indicated that early, focused, antiinfective therapy was important to avoid severe infection complications. Infections causing sepsis and multiorgan dysfunction were negatively associated with outcome and successful weaning of ECD. The percentages and types of pathogens in the ECD cohort did not differ from the general colonization of intensive care units. Because a significant correlation between pathogens, infections, and outcome was not detected, we recommend focusing on clinical parameters to decide whether patients will benefit from ECD support.

From the Departments of *Surgery and Cardiac and Thoracic Surgery, University BG Hospital Bergmannsheil, Bochum, Germany.

Submitted for consideration May 2015; accepted for publication in revised form October 2015.

Disclosure: The authors have no conflicts of interest to report.

Correspondence: Justyna Swol, Department of Surgery, BG University Hospital Bergmannsheil, Bürkle-de-la Camp Platz 1, 44789 Bochum, Germany. Email: jswol@icloud.com.

Extracorporeal membrane oxygenation (ECMO) is a lifesaving invasive treatment used to prevent acute respiratory failure or life-threatening loss of cardiac performance. In 2009, Peek et al.1 showed that patients with acute respiratory distress syndrome benefit from ECMO therapy.

In modern intensive medical care, the early diagnosis and treatment of sepsis and other infections play an important role. When extracorporeal life support (ECLS) was first introduced, sepsis or severe infections were considered a contraindication because of a lack of knowledge regarding the growth of microorganisms on membranes and cannulas.2 Currently, the imbalance between macrocirculatory and microcirculatory oxygenation and decarboxylation caused by sepsis is one of the indications for extracorporeal treatment.

However, infections are an important complication, and prognosis is often a limiting factor in modern intensive care. This study analyzes the incidence and treatment of infections, the microbiological spectrum, and the outcome of patients undergoing ECMO.

Back to Top | Article Outline

Materials and Methods

Extracorporeal devices (ECDs) include ECMO, ECLS, and pumpless extracorporeal lung assist (pECLA). This study was a retrospective, observational, single-center analysis that included 99 adult patients who received ECD while admitted to the intensive care unit (ICU) at the university hospital with a 13 bed surgical ICU between 2008 and 2014. To avoid misrepresentation with regard to the incidence of nosocomial infections, ECD treatment for less than 48 hours was an exclusion criterion for the analysis.

Back to Top | Article Outline

Demographic and Clinical Data

Demographic data, particularly gender and age, were recorded to describe the patient collective. The Simplified Acute Physiology Score (SAPS) II, SAPS 3 and Sequential Organ Failure Assessment (SOFA) score were calculated to quantify the physical condition of the patients. Furthermore, the length of stay (LOS) in the referring hospital, LOS in hospital, and the LOS in the ICU and hospital after ECLS treatment were analyzed and compared. In-hospital mortality and survival until successful ECMO weaning and hospital discharge were defined as the clinical outcome.

Back to Top | Article Outline

ECMO Data

Cannulation techniques were performed percutaneously. Cannulation date, cannula removal date, days on ECD, and cannulation modus (venovenous, venoarterial) were recorded. Complications and renal replacement therapy were also recorded.

Back to Top | Article Outline

Infections and Microbiological Data

The diagnostic standardization and categorization of infections were based on the Centers for Disease Control and Prevention definitions for nosocomial infections.3 Microbiological isolations were correlated with clinical symptoms and typical inflammatory characteristics in blood samples and radiographic findings. The microbiological data were categorized according to the sepsis classification by Annane et al.4 Documented information in the clinical records from the ICU was analyzed to define the infectious status. Patients who already suffered an infection before hospitalization or those who required ECMO/ECLS because of primary infections, as well as patients for whom an acute nosocomial infection was diagnosed after more than 48 hours of extracorporeal treatment, were considered.

The antiinfective regimen during ECLS therapy (Table 1) and the isolated microbiological species in blood cultures, tracheal secretions, urine, wounds, and others were recorded. As stated in the current Extracorporeal Life Support Organization (ELSO) ECMO guidelines, prophylactic use of antibiotic drugs during ECMO therapy was not performed.5

Table 1

Table 1

Back to Top | Article Outline

Statistical Analysis

The data were compared and calculated using IBM SPSS Statistics version 22 (IBM Corporation, Armonk, NY). Descriptive statistics were used to present the medians (range) for categorical data sets, whereas means and standard deviations (SDs) were calculated to express the amount and variance of continuous variables. Statistical tests were used to analyze the available data. The χ2 test was used to compare categorical variables; continuous variables were processed using the Mann-Whitney U test. The tests were two sided, and a p less than 0.05 was considered statistically significant. The cohort was divided into two groups: survivors and nonsurvivors. Infections were subdivided into primary and nosocomial infections.

Back to Top | Article Outline

Privacy Policy

All patient data were collected in a pseudonymous form using continuous study numbers. The data were saved, using only this code, in an exported database in an SPSS data sheet on an access-controlled PC and then evaluated. Only authorized personnel had access to the original data.

Back to Top | Article Outline

Results

We analyzed the data from 88 patients, 45 survivors and 43 nonsurvivors (Table 2). Eleven patients were excluded because of their short ECMO run for less than 48 hours (Table 3). In total, 85.2% of the patients were men with a median age of 54 years. In the nonsurvivor group, the median age was 10 years higher than in the survivor group (59 vs. 49, p < 0.001). Forty-six patients (52.3%) suffered from major trauma, with a significant difference between the survivors and nonsurvivors (32 survivors vs. 14 nonsurvivors, p = 0.001). In the group with extrapulmonary sepsis, 18 of 23 patients were in the nonsurvivor group, although we did not calculate the significance because of the small number of patients. We repeated this analysis for ECLS during cardiopulmonary resuscitation (n = 6) and other indications (n = 5). Venovenous ECMO was performed most often (76.1%), followed by (pECLA; 19.3%) and ECLS (9.1%); the incidences were similar in each group of survivors and nonsurvivors.

Table 2

Table 2

Table 3

Table 3

We also recorded scores and clinical data for the included patients (Table 4). The median SOFA score was 13 in all patients and in the nonsurvivor group, and surviving patients demonstrated a median score of 12. The median values for SAPS II and SAPS 3 were 72 and 73, respectively (67 vs. 76 and 64 vs. 68 for survivors and nonsurvivors). A total of 16 patients were cannulated at referral hospitals; the majority of these patients (n = 11) did not survive. The differences between survivor and nonsurvivor groups for the admission SOFA score (p = 0.001), SAPS II (p < 0.001), and SAPS 3 (p = 0.001) were significant.

Table 4

Table 4

The median ventilation time was 483.5 hours (survivors 667.0 vs. nonsurvivors 331.0). The median value for pre-ECMO time was 6.4 days (4.6 vs. 7.8), and 57% of patients were successfully weaned from ECMO and stayed in the hospital post-ECMO for a median time of 17.8 days (40.1 vs. 0). The general LOS in the hospital was 3.5–233.4 days (median, 45.0 days; 9.7 vs. 8.7), and 53.2% of all patients were discharged from the ICU after a median time of 9.2 days; most of these discharged patients (95.7%) survived.

Fifty-eight patients (29 vs. 29) were referred from an external hospital where they stayed for a median time of 4.5 days. Sixteen patients were cannulated at a referral center, and 11 did not survive; 8 of them died because of severe abdominal sepsis.

Fifty patients received continuous renal replacement therapy (CRRT), and 32 (64%) of these patients died. The mortality in the group of patients who received CRRT was statistically significant (p = 0.001) higher as in the group without CRRT. We recorded 26 complications with an equal rate in both groups. Clotting or replacement was the most common complication (n = 16); four of five recorded patients with low flow during treatment were in the nonsurvivor group. There were three cases of coagulation disorder, and two nonsurvivors suffered from cerebral bleeding. Accidental cannula removal and ischemia of an extremity occurred once in each group.

A total of 218 pathogens were isolated (88 Gram positive, 80 Gram negative, and 50 others); of these, 43% (n = 94) were detected in the survivor group and 57% (n = 124) in the nonsurvivor group. Forty pathogens were detected in blood cultures, 104 in tracheal secretions, 17 in urine, and 57 in all other samples (Table 5).

Table 5

Table 5

Staphylococci were the most commonly isolated microbe in the entire cohort (n = 54). We detected 18 cases of Staphylococcus aureus, of which 22% were methicillin-resistant S. aureus (MRSA). All MRSA species were found in the nonsurvivor group.

Fungi were isolated with the second highest frequency (n = 37), and the majority of these cases were detected in the nonsurvivor group (n = 28). Seventeen samples from the nonsurvivor group tested positive for Enterococcus spp., with a total number of 29 positive samples. Enterobacter spp. (n = 25, 16 in the nonsurvivor group), Klebsiella pneumoniae (n = 17), and Pseudomonas aeruginosa (n = 15) demonstrated high incidences, particularly in tracheal secretions. In 36% of all tracheal secretions, we found typical cellular inflammatory infiltration, although there was no difference between the two groups.

As described in the Materials and Methods section, we categorized the infections into nine groups, with two subgroups each (Table 6). Forty-eight patients (54.5%) suffered from primary infections. Independent of primary infectious status, nosocomial infections were detected in 38 patients (43.2%). We recorded 102 infections, including 48 in the survivor group and 54 in ECMO patients who did not survive. Nosocomial infections (n = 42) demonstrated a higher incidence in survivors (n = 26), and primary infections (n = 60) were more common in the nonsurvivor group (n = 38). Respiratory infections presented the highest incidence (n = 69). A total of 69% of patients with nosocomial respiratory infections during ECD support survived, and 22 patients in the nonsurviving group suffered a primary respiratory infection. Abdominal infections had the second highest incidence (n = 14), and most of these cases were primary infections (n = 12). A total of 75% of patients with a primary infection were in the nonsurvivor group. Urinary tract and skin and tissue infections had incidences of five cases each. All patients (n = 3) with a primary urinary tract infection died, whereas both patients with a nosocomial infection survived. Four of five patients suffering from a skin or other tissue infection did not survive. There were three recorded nosocomial cardiovascular infections in the cohort, especially catheter-related infections, of which one patient survived. Wound infections (n = 1) and bone and joint infections (n = 1) were rare, and we did not record any cerebral infections in the cohort.

Table 6

Table 6

We analyzed 215 antibiotic treatment cycles, including 98 in the survivor group and 117 in the nonsurvivor group (Table 7). We recorded 112 first-line cycles, 71 second runs, 23 third runs, and 9 others.

Table 7

Table 7

Carbapenems (n = 60) were the most commonly prescribed antibiotic class, except in first-line therapy, where β-lactamase inhibitors were administered most often. β-Lactamase inhibitors were more common in the survivor group (n = 23), whereas carbapenems (n = 31), vancomycin (n = 10), and tigecycline (n = 3) were more commonly prescribed in the nonsurvivor group. All other antibiotics were given in equal amounts in both groups. Nine patients (10.2%) received more than three cycles of antibiotic treatment. One patient each was treated with tobramycin, tuberculostatics, and ceftazidime. Three patients received a further cycle with tigecycline and three with cotrimoxazole.

Back to Top | Article Outline

Discussion

Age and ECMO indications, especially trauma, have been shown to be independent predictors of outcome.6,7 In addition, a SOFA score above 11 is associated with a very high mortality up to 95%.8 In our cohort, 45 of 88 patients (51.1%) with a median SOFA score of 13 survived. The results of the SAPS 2 and the SAPS 3 mortality predictions were similar to the SOFA results, but the SOFA score revealed differences in the severity of the disease in the survivor and nonsurvivor groups. The mean SOFA score and SD for the surviving group are 11 ± 3 SD and 14 ± 3 SD for the nonsurviving group.

Ventilation time and LOS were shorter in the nonsurvivor group, likely because of patient death, although we did not find a correlation between ventilation time and ECLS outcome. Previous studies have suggested that ECMO treatment increases the risk of nosocomial infection. Complications, especially bleeding, are considered a statistically significant predictor of mortality,9 although in our cohort, these complications played a minor role. Nosocomial infections had a higher incidence in survivors, which demonstrates the increasing risk of infections because of prolonged in-hospital stay.10

In the literature, infection rates in the ICU range from 13.9% to greater than 44%.10,11 In the this study, nosocomial infections, independent of the primary infectious status, were detected in 38 patients (43.2%). Forty-eight patients (54.5%) suffered from primary infections.

In 1995, Vincent et al. analyzed 1,417 ICUs in Europe to determine the incidence of pathogens and infections. Pneumonia and other lower respiratory infections (64.7%) were the most common diseases, followed by urinary tract infections (17.6%) and bloodstream infections (12%).10,12,13 In our ECMO cohort, we found similar incidences for respiratory infections (64.7%) but lower rates of urinary tract infections (5%). The incidence of urinary tract infections was too low to identify significance. Existing studies have found that these infections increase the morbidity of ICU patients but not their mortality.10,14 The literature states that a cause-and-effect relationship between organisms detected in blood cultures and the appearance of bloodstream infections cannot be proven, but the fact that almost every fifth microbe in our cohort was isolated from the blood suggests that ECMO patients carry a risk of developing bacteremia. This results in increased risk of provoking a systemic inflammatory response in the host in response to pathogens.15 Most positive blood cultures were observed in the nonsurvivor group, and further research is necessary to determine whether these ECMO patients have higher bacterial exposure than non-ECMO patients who do not survive the ICU. Therefore, a data comparison of bacteremia in blood cultures before and after ECMO treatment is necessary. Blood samples were not collected before ECMO cannulation. This represents an additional limitation of this study.

In general, our microbiological findings did not differ conspicuously from published incidences in the literature. Large-scale international studies show that S. aureus and other Staphylococcus spp. (38%) are the most common pathogens causing bloodstream infections, followed by Escherichia coli (24%). In addition, Candida spp. are often isolated in the ICU and can cause pneumonia and bloodstream infections with a mortality rate of up to 47%.10,16–18 Current figures state that the balance is shifting from multiresistant staphylococci to highly resistant classes of Acinetobacter spp. and P. aeruginosa,19 and the microbiological findings in our cohort support this trend.

The treatment of infected patients in our study was based on common national and international guidelines.20,21 β-Lactamase inhibitors are used as the initial therapy for respiratory infections.22 Imipenem is a first-line antibiotic for abdominal infections or second-line treatment option.23

Tigecycline was administered more often in the nonsurvivor group because it was used as a rescue treatment for severe infections caused by multidrug-resistant bacteria. Contemporary studies have shown that higher doses of antibiotics than advised may improve therapeutic outcome.24

Ferrer et al.25 demonstrated that every hour of delay between sepsis diagnosis to therapy reduces the survival rate, and the most feared complication of sepsis or septic shock resulting from an infection is multiorgan dysfunction.26 Extracorporeal membrane oxygenation enables oxygenation and decarboxylation, and the additional application of CRRT can be a safe and effective method27 and serve as a reference point for the severity of sepsis. Once present at high levels, microbiological colonization of the oxygenator is possible.28,29 Our results suggest that noncontrollable complications, rather than the infection itself, influence the success of ECLS.

It was previously reported that ECD changes the levels of drugs in the blood. For instance, lipophilic substances may become absorbed in the circuit.30 In 2012, Shekar et al.31 developed a sheet model to record the pharmacokinetics of study drugs during ECMO to serve as a basis for further research. In our analysis, we did not uncover evidence to support this hypothesis, although additional work is required to determine whether membranes and tube systems influence the pharmacokinetics of antiinfective drugs such that alterations in dosing are necessary.

Back to Top | Article Outline

Study Limitations

Because of the various indications for ECMO support and the broad spectrum of patients included, it was difficult to obtain significant findings. Most infections and pathogens in our cohort had a low incidence (between n = 1 and n = 5). Significance could be calculated for all parameters, but p values were not meaningful because of the multiple use of parameters in some groups or, in part, a very low number of cases. For a better overview, p values are shown in a table, and those that are important and meaningful are commented on in the results.

An additional limitation is that we do not have data for the pre-ECMO incidence of bacteremia because blood samples were not collected before ECMO cannulation in every patient. Therefore, a data comparison of bacteremia in blood cultures before and after ECMO treatment could reveal the true risk of bloodstream infections.

Back to Top | Article Outline

Conclusion

Respiratory infections were the most common type of complication in our cohort. Moreover, the high incidence of pathogens in blood cultures suggests that the risk of developing a bloodstream infection in ECMO patients is high. However, there was no clear correlation between infections, ECMO therapy, and outcome, so we recommend focusing on clinical parameters to weigh therapeutic decision making for patients with ECLS. Regardless, an early and focused antiinfective regimen is important to avoid further complications of infections, such as dysfunction of more than one organ system.

Back to Top | Article Outline

References

1. 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
2. McCune S, Short BL, Miller MK, Lotze A, Anderson KD. Extracorporeal membrane oxygenation therapy in neonates with septic shock. J Pediatr Surg. 1990;25:479–482
3. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control. 1988;16:128–140
4. Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet. 2005;365:63–78
5. ELSO Guidelines for Cardiopulmonary Extracorporeal Life Support. Extracorporeal Life Support Organization. Ann Arbor, MI, USA Version 1.3 November 2013 Available at: www.elsonet.org
6. Hemmila MR, Rowe SA, Boules TN, et al. Extracorporeal life support for severe acute respiratory distress syndrome in adults. Ann Surg. 2004;240:595–605; discussion 605
7. Ried M, Bein T, Philipp A, et al. Extracorporeal lung support in trauma patients with severe chest injury and acute lung failure: A 10-year institutional experience. Crit Care. 2013;17:R110
8. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22:707–710
9. Aubron C, Cheng AC, Pilcher D, et al. Factors associated with outcomes of patients on extracorporeal membrane oxygenation support: A 5-year cohort study. Crit Care. 2013;17:R73
10. Vincent JL, Bihari DJ, Suter PM, et al. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. EPIC International Advisory Committee. JAMA. 1995;274:639–644
11. Maillet JM, Guérot E, Novara A, et al. Comparison of intensive-care-unit-acquired infections and their outcomes among patients over and under 80 years of age. J Hosp Infect. 2014;87:152–158
12. Rosenthal VD, Bijie H, Maki DG, et al. International Nosocomial Infection Control Consortium (INICC) report, data summary of 36 countries, for 2004–2009. Am J Infect Control. 2012;40:396–407
13. O’Neill JM, Schutze GE, Heulitt MJ, Simpson PM, Taylor BJ. Nosocomial infections during extracorporeal membrane oxygenation. Intensive Care Med. 2001;27:1247–1253
14. Bagshaw SM, Laupland KB. Epidemiology of intensive care unit-acquired urinary tract infections. Curr Opin Infect Dis. 2006;19:67–71
15. Klein Klouwenberg PM, Ong DS, Bonten MJ, Cremer OL. Classification of sepsis, severe sepsis and septic shock: The impact of minor variations in data capture and definition of SIRS criteria. Intensive Care Med. 2012;38:811–819
16. Anderson DJ, Moehring RW, Sloane R, et al. Bloodstream infections in community hospitals in the 21st century: A multicenter cohort study. PLoS One. 2014;9:e91713
17. Pappas PG, Kauffman CA, Andes D, et al. Clinical practice guidelines for the management of candidiasis: 2009 update by the Infectious Diseases Society of America. Clin Infect Dis. 2009;48:503–535
18. Morgan J, Meltzer MI, Plikaytis BD, et al. Excess mortality, hospital stay, and cost due to candidemia: A case-control study using data from population-based candidemia surveillance. Infect Control Hosp Epidemiol. 2005;26:540–547
19. Doyle JS, Buising KL, Thursky KA, Worth LJ, Richards MJ. Epidemiology of infections acquired in intensive care units. Semin Respir Crit Care Med. 2011;32:115–138
20. Owens RC Jr. Antimicrobial stewardship: Concepts and strategies in the 21st century. Diagn Microbiol Infect Dis. 2008;61:110–128
21. Höffken G, Lorenz J, Kern W, et al.Paul-Ehrlich-Society of Chemotherapy; German Respiratory Diseases Society; German Infectious Diseases Society; Competence Network CAPNETZ for the Management of Lower Respiratory Tract Infections and Community-acquired Pneumonia. Guidelines of the Paul-Ehrlich-Society of Chemotherapy, the German Respiratory Diseases Society, the German Infectious Diseases Society and of the Competence Network CAPNETZ for the Management of Lower Respiratory Tract Infections and Community-acquired Pneumonia. Pneumologie. 2010;64:149–154
22. File TM Jr. Recommendations for treatment of hospital-acquired and ventilator-associated pneumonia: Review of recent international guidelines. Clin Infect Dis. 2010;51(suppl 1):S42–S47
23. Balfour JA, Bryson HM, Brogden RN. Imipenem/cilastatin: An update of its antibacterial activity, pharmacokinetics and therapeutic efficacy in the treatment of serious infections. Drugs. 1996;51:99–136
24. De Pascale G, Montini L, Pennisi M, et al. High dose tigecycline in critically ill patients with severe infections due to multidrug-resistant bacteria. Crit Care. 2014;18:R90
25. Ferrer R, Martin-Loeches I, Phillips G, et al. Empiric antibiotic treatment reduces mortality in severe sepsis and septic shock from the first hour: Results from a guideline-based performance improvement program. Crit Care Med. 2014;42:1749–1755
26. Zangrillo A, Landoni G, Biondi-Zoccai G, et al. A meta-analysis of complications and mortality of extracorporeal membrane oxygenation. Crit Care Resusc. 2013;3:172–178
27. Santiago MJ, Sánchez A, López-Herce J, et al. The use of continuous renal replacement therapy in series with extracorporeal membrane oxygenation. Kidney Int. 2009;76:1289–1292
28. Kuehn C, Orszag P, Burgwitz K, et al. Microbial adhesion on membrane oxygenators in patients requiring extracorporeal life support detected by a universal rDNA PCR test. ASAIO J. 2013;59:368–373
29. Müller T, Lubnow M, Philipp A, et al. Risk of circuit infection in septic patients on extracorporeal membrane oxygenation: A preliminary study. Artif Organs. 2011;35:E84–E90
30. Wildschut ED, Ahsman MJ, Allegaert K, Mathot RA, Tibboel D. Determinants of drug absorption in different ECMO circuits. Intensive Care Med. 2010;36:2109–2116
31. Shekar K, Roberts JA, Welch S, et al. ASAP ECMO: Antibiotic, sedative and analgesic pharmacokinetics during extracorporeal membrane oxygenation: A multi-centre study to optimise drug therapy during ECMO. BMC Anesthesiol. 2012;12:29
Keywords:

extracorporeal life support (ECLS); extracorporeal membrane oxygenation (ECMO); infections; antibiotics; outcome

Copyright © 2016 by the American Society for Artificial Internal Organs