Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Sep 28.
Published in final edited form as: ASAIO J. 2022 Sep 28;68(6):859–864. doi: 10.1097/MAT.0000000000001575

Indwelling central venous catheters drive bloodstream infection during veno-venous extracorporeal membrane oxygenation support

Adwaiy Manerikar 1, Satoshi Watanabe 2, Viswajit Kandula 1, Azad Karim 1, Sanket Thakkar 1, Mark Saine 1, Samuel S Kim 1, Rafael Garza-Castillon 1, David D Odell 1, Ankit Bharat 1,2, Chitaru Kurihara 1
PMCID: PMC8958168  NIHMSID: NIHMS1735296  PMID: 34593682

Abstract

Blood stream infection (BSI) is a potentially lethal complication in patients receiving extracorporeal membrane oxygenation (ECMO). It may be particularly common in patients with veno-venous (VV) ECMO due to their long hospitalization in the intensive care unit. Given that these patients have concurrent indwelling central venous catheters (CVC), it is unclear whether the ECMO circuit, CVC, or both, contribute to BSI. This study evaluated the risk factors associated with BSI in patients receiving VV ECMO in a single institution study of 61 patients from 2016 through 2019. All ECMO catheters and the circuit oxygenator fluid were aseptically collected and analyzed for microorganisms at the time of decannulation. New BSI was diagnosed in 15 (24.6%) patients and increased mortality by 3-fold. None of the ECMO catheters or oxygenator fluid were culture positive. BSI increased with CVC use of over 8 days and was significantly lowered when CVC were exchanged by day 8 compared to patients with exchanges at later points (15.0% vs 42.8%, p=0.02). Median length of CVC use in the BSI-negative and BSI-positive group were 6.3±5.0 and 9.4±5.1, respectively (p=0.04). In summary, BSI is a potentially lethal complication in patients receiving ECMO. Indwelling CVC, not the ECMO circuitry, is the likely contributor for BSI, and exchanging CVC by day 8 can reduce the incidence of BSI.

Keywords: artificial organs, circulatory support devices, outcomes

Introduction

Veno-venous extracorporeal membrane oxygenation (VV-ECMO) is a life-support technique that is frequently used for patients with respiratory failure(1). VV-ECMO has been used for the treatment of severe acute respiratory distress syndrome (ARDS) from a variety of causes including coronavirus disease 2019 (COVID-19)(25). While technological advances have allowed for increased adoption of this technology(6), major adverse effects have been associated with ECMO, including nosocomial infections and bloodstream infection (BSI)(7). The Extracorporeal Life Support Organization reported an incidence of 15.4 infections per 1000 ECMO days(8) and other studies have found rates as high as 64%(9). While various studies have described BSI occurring during ECMO, the source remains unclear. Since patients undergoing ECMO frequently have simultaneous indwelling central venous catheters (CVCs), it remains unknown if the origin of bloodstream infection is the ECMO circuit, CVC, or both. It has been shown that longer durations of indwelling catheters are associated with increased risk of BSI. These papers cite lengths of 10, 14, and 15 days as potential cut-offs for increased risk (1012). Current guidelines recommend against scheduled replacement of CVCs and state that they should remain in place until an infection is noted(13). However, this data is based on studies conducted in general ICU patients in whom the impact of BSI and mortality may be different than those supported with ECMO. Furthermore, the total duration of CVC use in the non-ECMO patients is much less(14) and the risks of periodic change of CVC might outweigh the benefits. ECMO patients often require support for longer durations, typically more than 7 days and, as such, these data and recommendations may not be relevant in patients receiving ECMO.

The aim of this study was to analyze the incidence of postoperative complications including acute kidney injury (AKI), gastrointestinal bleeding (GI bleeding), pump malfunction, and oxygenator dysfunction in patients supported with VV ECMO and the impact of BSI on these complications as well as overall mortality. Additionally, this study sought to determine the characteristics and risk factors for patients with BSI associated with VV-ECMO, and evaluated whether CVCs or ECMO circuits played a role in the development of BSI.

Patients and Methods

Study subjects

Patient data was collected retrospectively using the electronic medical record and stored in a database. This study was approved by the Northwestern University Institutional Review Board (STU00207250). However, the need for patient consent for data collection was waived by the IRB as this was a retrospective study. Adult patients undergoing VV-ECMO at Northwestern between January 2015 and February 2019 were included. Thirteen patients were excluded from this study due to potential confounding effects. Reasons for exclusion included patients without culture results from ECMO catheters or circuit oxygenators or patients requiring conversion to veno-arterial ECMO or veno-arterial-veno ECMO. Continuous anticoagulation was not used, and patients were not monitored with bleeding parameters such as ACT or aPTT, consistent with recent reports (15, 16). In accordance with standard ICU protocol for patients’ not undergoing systemic anticoagulation, patients received subcutaneous heparin injections for deep vein thrombosis prophylaxis. To limit potential thrombotic complications in ECMO circuit, flow was maintained at least 3.0–3.5L/min consistent with recent reports demonstrating the feasibility of using VV-ECMO without anticoagulation (15, 17, 18). Transfusions were initiated under the following circumstances: Platelets <50,000/mL, Hemoglobin <7 g/dL, or hemodynamic instability in the setting of active blood loss. Various cannulation strategies [Internal jugular vein – femoral vein cannulation vs ProtekDuo® cannulation (CardiacAssist Inc., Pittsburgh, PA, USA)] were used in patients depending on the surgeon preference. The VV-ECMO circuit included Quadrox iD adult (7.0) oxygenator (MAQUET Holding B.V. & Co. KG, Germany) and Rotaflow pump (MAQUET Holding B.V. & Co. KG, Germany). Circuit components including the tubing and oxygenator were coated with heparin but the cannulas were not.

Patients with respiratory failure were considered for ECMO if they failed to achieve satisfactory gas exchange (PaO2>55 mmHg, Oxygen saturations >88%, pH>7.2, with plateau pressures less than 35) despite lung protective mechanical ventilation and recruitment maneuvers with neuromuscular blockade, consistent with previous reports (19).

Management of Central Venous Catheters

Consistent with the ICU guidelines, patients did not undergo a practice of routine exchange of CVCs. However, after detecting several cases of BSI, programmatic practice changed to exchanging CVCs every 7 days after March 2018. Hence, for a subset of patients, CVCs were replaced every 7 days, even if there were no signs of infection. New CVC replacement was performed in the same vein and not a wire-based replacement but different area or in the new vein. Weekly surveillance cultures for all ECMO patients were used to monitor for BSI simultaneously. These cohorts represent two different time points in the institution’s ECMO program but did not differ in any management strategies except for the use of anticoagulation. In order to determine whether use of anticoagulation affected the BSI in the two cohorts, a Cox analysis was performed, as described below. Also, all the patients were contained in a narrow study period of 3 years and there were no temporal factors that could impact the outcomes. For the remaining patients, CVCs were left in place until an infection or complication necessitated removal. At the conclusion of the study, infection rates were analyzed in conjunction with CVC duration to determine what length of CVC use was associated with increased risk of infection. Pulmonary artery (PA) catheter was not used as routine for patients who were supported VV ECMO. When required, temporary hemodialysis catheter was changed every 7 days.

Culture of ECMO catheters and the circuit oxygenator

Catheter samples were removed aseptically at the time of ECMO decannulation for all patients. After removal, the tips were sent for microbiological examination. At that time, swabs of the ECMO oxygenator and CVC tips were also analyzed for the presence of microorganisms. Catheter-related infection was defined when any one of the following criteria were met: 1) local sign of infection at the insertion site; 2) positive microbiological sample taken from a cannula or CVC tip. Only infections occurring between initiation of ECMO and decannulation were defined as ECMO-related infections.

Definitions of complications

Complications following VV-ECMO cannulation were defined using the following criteria. Pneumonia was defined as newly diagnosed chest radiograph infiltrate with positive bronchial alveolar lavage. Urinary tract infection was defined as a positive urine culture with clinical sings of infection. Bloodstream Infection was defined as a positive microbial culture. Acute kidney injury was defined using the Risk, Failure, Loss of kidney function and End-stage kidney disease (RIFLE) classification (20). Gastrointestinal bleeding was defined as either hematemesis, melena, guiaic-positive stool or active bleeding at the time of endoscopy or colonoscopy. Finally, neurological dysfunction, was defined as a new deficit confirmed via abnormal neuroimaging. This was further divided into ischemic or hemorrhagic causes.

Statistical analysis

Patient demographics, postoperative complications, and outcomes were compared between both the infection and non-infection cohorts and also the CVC-exchange vs no exchange cohorts. Continuous variables were compared using t-test and reported as means while categorical variables were compared using chi-square test and reported as number (percentage). P-values <0.05 were accepted as statistically significant. The Kaplan-Meier method was used to estimate survival and a log-rank test was performed to compare survival between the two groups. Cox proportional hazard regression was used to derive hazard ratios and 95% confidence intervals. To build these models, a univariate analysis was performed and included all predictors if the test had a p-value of 0.2 or less. To assess the overall goodness of fit, Gronnesby and Borgan tests were utilized. Statistical analyses were performed using Stata/MP14 (StataCorp, College Station, TX).

Results

Study population and Clinical Characteristics of Infections

During the study period, 61 patients were placed on VV-ECMO. Overall 23 patients had BSI and 15 out of 23 (65.2%) were CVC related. Table 1 shows pre VV-ECMO characteristics of the study cohort. Fifteen patients developed BSI while undergoing VV ECMO and 46 patients remained infection-free. The BSI+ve group had a longer duration of ECMO support [BSI-free (10.4 ± 12.7 days) vs BSI+ve (34 ± 24.8 days), p<0.001]. The BSI+ve group also had lower baseline platelets (p=0.03), RESP score (p=0.02), and PaO2 (p<0.01). The BSI+ve group had a higher incidence of interstitial lung disease (p-0.04, Supplemental Table1). The mean time to documented infection was 15.4 days. In terms of post-cannulation complications between the BSI+ve and BSI-ve groups, the BSI+ve group had a greater incidence of GI bleeding (p=0.001), hemodialysis use (p=0.05), and anticoagulation (p=0.04) (Table 2). Importantly, the BSI+ve group had worse survival rate compared to the BSI-ve group (p=0.03, Figure 1). The causative microorganisms of bloodstream infection are listed in Table 3. The predominant organisms were Candida spp. (n=5, 26.3%), Staphylococcus, Enterococcus, and non-fermenting gram-negative bacilli such as Pseudomonas aeruginosa and Stenotrophomas maltophila. At the time of BSI, no other sources of infection were found in these patients, including pneumonia, skin infection, and urinary tract infections. None of the ECMO cannula tips or oxygenator fluid swabs were found to be culture positive. Median length of CVC use in the BSI-ve and BSI+ve group were 6.3±5.0 and 9.4±5.1, respectively (p=0.04). BSI+ve group did demonstrate higher incidence of new pneumonia during their ICU course compared to BSI-ve free group (60.0% vs 26.0%, p=0.01), but urinary tract infection was not significantly different (2.1% vs 13.3%, p=0.08). In patients who were diagnosed with BSI, the CVC was replaced but in no patients the ECMO circuit was changed. All patients converted to culture-negative with antimicrobials and CVC replacement even through the ECMO circuit was the same.

Table I.

Baseline characteristics of patients developing infected vs infection free patients

Variable Infection Free (n=46) Infection (n=15) P value
ECMO Support days 10.4 ± 12.7 34 ± 24.8 <0.01
Age, years 48.1 ± 15.7 45.5 ± 14.7 0.56
BMI, kg/m2 28.9 ± 9.4 31 ± 11.2 0.46
BSA, m2 1.9 ± 0.3 1.9 ± 0.5 1.00
Hypertension 14 (30.4%) 2 (13.3%) 0.31
Diabetes 7 (15.2%) 1 (6.7%) 0.67
Chronic kidney disease 11 (23.9%) 1 (6.7%) 0.26
Smoking 12 (26.1%) 4 (26.7%) 1.00
Inotrope 16 (34.8%) 4 (26.7%) 0.75
Pressors 26 (56.5%) 8 (53.3%) 0.13
Nitric oxide 15 (32.6%) 1 (6.7%) 0.09
RESP Score 0.8 ± 3.3 −1.4 ± 3.1 0.02
Laboratory
Hemoglobin, g/dL 10.6 ± 2.8 11.3 ± 2.2 0.32
WBC, 1,000/mm3 13.6 ± 6.4 16.6 ± 12.5 0.38
Platelets, 1,000/mm3 231.4 ± 128 161.3 ± 94.2 0.03
Sodium, mEq/L 137.1 ± 8.5 138.4 ± 5.5 0.50
Creatinine, mg/dL 1.5 ± 2.4 1.9 ± 2.6 0.60
BUN, mg/dL 28.3 ± 19.2 32.3 ± 21.9 0.53
AST, U/L 71.1 ± 145.3 75.4 ± 107.7 0.90
ALT, U/L 57.3 ± 124.1 59 ± 86 0.95
Total bilirubin, mg/dL 2.1 ± 6.6 1 ± 0.8 0.27
Albumin, g/dL 2.9 ± 0.8 2.9 ± 0.4 1.00
INR 1.4 ± 0.7 1.4 ± 0.3 1.00
ABG (at cannulation)
pH 7.3 ± 0.1 7.3 ± 0.1 1.00
PaCO2 55.8 ± 21.9 57.5 ± 14.2 0.73
PaO2 116 ± 89.5 71.4 ± 25.1 <0.01
HCO3 26.5 ± 6.9 26.5 ± 6.5 1.00
Lactate 3.4 ± 3.4 4 ± 3.4 0.56

Continuous data are shown as means ± standard deviation (SD). Categorical variables are presented as number and percentage. BMI, body mass index; BSA, body surface area; RESP, Respiratory ECMO Survival Prediction; WBC, white blood cell; BUN, blood urea nitrogen; AST, aspartate aminotransferase; ALT, alanine aminotransferase; INR, international normalized ratio; PT, prothrombin time; PTT partial thromboplastin time

Table II.

Incidence of Adverse Events in Infection and Infection-Free Groups

Event Infection-Free (n=46; 425 days) EPPD Infection (n=15; 482 days) EPPD P value
AKI 16 (34.8%) - 9 (64.3%) - 0.13
Dialysis 11 (23.9%) - 8 (53.3%) - 0.05
Neurological dysfunction 4 (8.7%) 0.009 1 (6.7%) 0.002 1.00
IND 1 (2.2%) 0.002 0 (0%) 0.00 1.00
HND 3 (6.5%) 0.007 1 (6.7%) 0.002 1.00
GI bleeding 4 (8.7%) 0.009 9 (60.0%) 0.02 0.001
Anticoagulation 22 (47.8%) - 12 (80.0%) - 0.04

EPPD, event per patient per day; AKI, acute kidney injury; IND, ischemic neurological dysfunction; HND, hemorrhagic neurological dysfunction; GI bleeding; gastrointestinal bleeding;

Figure 1.

Figure 1.

Survival of patients supporting veno-venous extracorporeal membrane oxygenation: infection vs. non-infection.

Table III.

Microorganisms Isolated from Bloodstream

Microorganism Patients
Non-fermenting Gram-negative bacilli, n (%) 3 (15.8%)
Pseudomonas Aeruginosa, n (%) 2 (10.5%)
Stenotrophomas maltophilia, n (%) 1 (5.3%)
Enterobacteriaceae 2 (10.5%)
Escherichia coli, n (%) 1 (5.3%)
Enterobacter cloacae, n (%) 1 (5.3%)
Staphylococcus spp. n (%) 3 (15.8%)
Staphylococcus aureus. n (%) 1 (5.3%)
Staphylococcus epidermidis, n (%) 2 (10.5%)
Streptococcus spp. n (%) 1 (5.3%)
Candida spp. n (%) 5 (26.3%)
Candida parapsilosis, n (%) 1 (5.3%)
Candida albicans, n (%) 1 (5.3%)
Candida dubliniensis, n (%) 1 (5.3%)
Candida tropicalis, n (%) 1 (5.3%)
Candida Auris, n (%) 1 (5.3%)
Enterococcus faecalis, n (%) 4 (21.1%)
Enterococcus faecium, n (%) 1 (5.3%)

Cox multivariable logistic regression analysis of association between BSI and mortality in patients supported with VV-ECMO

The patients in the BSI+ve group had a higher mortality. However, these patients also had a longer duration of ECMO support and lower RESP and the mortality may be unrelated to the BSI. Therefore, to determine predictors of mortality, all variables were placed in a univariate analysis. This study included variables that had a p-value of 0.2 or less for a multiple Cox proportional hazards model with postoperative survival as the outcome (Supplemental Table 2). This study found that body surface area (BSA) (p<0.01) of the patient, RESP score (p<0.01), and low platelet counts (p<0.01) prior to initiation of ECMO were independent predictors of postoperative survival (Table 4). Furthermore, bloodstream infection while undergoing ECMO was indeed an independent predictor of postoperative mortality (p=0.03) (Table 4).

Table IV.

Cox Multivariable Logistic Regression Analysis: Predictors of Postoperative Mortality

Variable HR P value 95% CI
Infection 1.43 0.03 1.15
Age, years 1.01 0.27 0.98–1.05
BSA, m2 0.04 <0.001 0.01–0.03
RESP Score 0.75 <0.001 0.62–0.91
Laboratory
Platelets, 1,000/mm3 0.99 <0.001 0.98–1.00
ALT, U/L 1.00 0.71 0.99–1.00
Albumin, g/dL 0.57 0.18 0.25–1.29

BSA, body surface area; RESP; Respiratory ECMO Survival Prediction; ALT, Alanine aminotransferase;

Clinical Characteristics of CVC Exchange Cohorts

In no patients, replacement of CVC was associated with an iatrogenic complications. A new central line was placed when patients developed BSI. However, when limited venous access precluded a new line, the original line was replaced. There were no cases of a second infection after the original episode. A CVC use of greater than 8 days was found to be a cut-off point for increasing rate of infection (Table 5). Forty patients had their CVC replaced within 8 days while 21 patients had a CVC in place for greater than 8 days (Table 6). Indeed, indwelling CVC for greater than 8 days was associated with higher rate of infection (15.0% vs 42.8%, p=0.02). There were no other significant differences in patient characteristics between the two groups, except for lactate (p=0.02). However, the group that underwent CVC replacement longer than 8 days had a significantly longer duration of ECMO support (8.6 ± 14.7 days vs 26.9 ± 21.5 days, p<0.001). There was no significant difference in the etiology of lung failure between the two groups (Supplemental Table3).

Table V.

Incidence of Infection Based on Days Prior To CVC Exchange (61 patients)

Days before CVC Exchange Infection rate, line changed before cut-off Infection rate, line changed after cut-off P value
7 16.20% 37.5% 0.07
8 15% 42.9% 0.03
9 14.60% 45.0% 0.02
above 10 15.20% 53.3% 0.01

Table VI.

Baseline characteristics of patients with CVC exchange before and after 8 days

Variable ≤ 8 Days (n=40) >8 Days (n=21) P value
Infection 6 (15.0%) 9(42.8%) 0.02
ECMO Support days 8.6 ± 14.7 26.9 ± 21.5 0.001
Age, years 46.3 ± 15.1 51.1 ± 17.6 0.29
BMI, kg/m2 28.8 ± 14.6 30.7 ± 10.6 0.56
BSA, m2 1.9 ± 0.5 1.9 ± 0.3 1.00
Hypertension 11 (27.5%) 5 (23.8%) 1.00
Diabetes mellitus 5 (12.5%) 3 (14.3%) 1.00
Chronic kidney disease 9 (22.5%) 3 (14.3%) 0.52
Smoking 11 (27.5%) 5 (23.8%) 1.00
Intrope 15 (27.5%) 5 (23.8%) 0.39
Pressors 22 (55.0%) 12 (57.1%) 1.00
Nitric oxide 10 (25.0%) 6 (28.6%) 0.77
RESP Score 0.6 ± 3 −0.4 ± 2.6 0.18
Laboratory
Hemoglobin, g/dL 11.2 ± 3.7 10 ± 1.8 0.09
WBC, 1,000/mm3 13.3 ± 6.5 16.2 ± 10.2 0.24
Platelets, 1,000/mm3 212.6 ± 116.9 217.3 ± 113.1 0.88
Sodium, mEq/L 138.3 ± 35.3 135.8 ± 10.5 0.68
Creatinine, mg/dL 1.7 ± 0.9 1.5 ± 2.2 0.69
BUN, mg/dL 31.3 ± 23.2 25.5 ± 18 0.28
AST, U/L 91.8 ± 222.1 37.8 ± 26.8 0.13
ALT, U/L 71.6 ± 196.3 33.4 ± 17.8 0.23
Total bilirubin, mg/dL 1.1 ± 1.7 3.2 ± 9.4 0.31
Albumin, g/dL 3 ± 0.8 2.7 ± 0.6 0.10
INR 1.3 ± 0.4 1.6 ± 1 0.19
ABG (at cannulation)
pH 7.3 ± 1.9 7.3 ± 0.1 1.00
PaCO2 56.1 ± 28.2 56.4 ± 23.2 0.96
PaO2 111.7 ± 53.9 91.7 ± 66 0.24
HCO3 25.5 ± 9.2 28.2 ± 6.8 0.20
Lactate 4.2 ± 3.1 2.5 ± 2.3 0.02

Continuous data are shown as means ± standard deviation (SD). Categorical variables are presented as number and percentage. BMI, body mass index; BSA, body surface area; RESP, Respiratory ECMO Survival Prediction; WBC, white blood cell; BUN, blood urea nitrogen; AST, aspartate aminotransferase; ALT, alanine aminotransferase; INR, international normalized ratio; PT, prothrombin time; PTT partial thromboplastin time

Complication rates and mortality of CVC exchange groups

Next, a comparison was made between post-cannulation complications between patients undergoing CVC replacement within 8 days and those who did not (Table 7). After VV-ECMO initiation, 13 patients had GI bleeding; the incidence in the replacement group was 10% whereas the non-replacement group had an incidence of 42.9% (p<0.01). There were no significant differences between the two groups in the incidence of AKI (45.0% vs 33.3%, p=0.42), hemodialysis use (32.5% vs 28.6%, p=1.00), and neurological dysfunction (10.0% vs 4.8%, p=0.65). In the group that did not undergo CVC replacement, there was higher incidence of urinary tract infection (14.2% vs 0%, p=0.01), but new pneumonia during VV-ECMO supports was not significantly different (42.8% vs 30.0%, p=0.32). In addition, infection rates (event per patient day) were significantly greater in the CVC>8 days (0.0076 vs 0.0043, p=0.03).

Table VII.

Incidence of Adverse Events in ≤ 8 and >8 Day Exchange Groups

Event ≤ 8 Days (n=40; 343 days) EPPD >8 days (n=21; 564 days) EPPD P value
AKI 18 (45.0%) - 7 (33.3%) - 0.42
Dialysis 13 (32.5%) - 6 (28.6%) - 1.00
Neurological dysfunction 4 (10.0%) 0.01 1 (4.8%) 0.002 0.65
IND 1 (2.5%) 0.002 1 (4.8%) 0.002 1.00
HND 3 (7.5%) 0.009 0 (0.0%) 0.00 0.54
GI bleeding 4 (10.0%) 0.01 9 (42.9%) 0.016 0.01
Anticoagulation 17 (42.5%) - 17 (81.0%) - 0.01

EPPD, event per patient per day; AKI, acute kidney injury; IND, ischemic neurological dysfunction; HND, hemorrhagic neurological dysfunction; GI bleeding; gastrointestinal bleeding;

Cox multivariable logistic regression analysis for development of bloodstream infection

To determine predictors of BSI, all variables were placed in a univariate analysis and included all predictors with a p-value of 0.2 or less in a multiple Cox proportional hazards model with postoperative survival as the outcome (Supplemental Table 4). This study found that length of CVC use (p=0.04) was the only predictor for development of BSI while undergoing ECMO (Table 8).

Table VIII.

Cox Multivariable Logistic Regression Analysis: Predictors of Infection

Variable HR P value 95% CI
Days of central line 1.01 0.04 1.00–1.02
Diabetes mellitus 1.19 0.12 0.95–1.48
Preoperative NO 0.30 0.33 0.26–3.48
RESP Score 0.84 0.33 0.26–3.48
Laboratory
Hemoglobin, g/dL 1.12 0.64 0.67–1.89
Platelets, 1,000/mm3 0.99 0.17 0.98–1.00
AST, U/L 0.99 0.46 0.97–1.01
ALT, U/L 1.01 0.22 0.98–1.04
ABG (at cannulation)
pH 0.18 0.70 0.14–1.49
Lactate 1.19 0.12 0.95–1.48

NO, nitric oxide; RESP; Respiratory ECMO Survival Prediction; AST, aspartate aminotransferase; ALT, Alanine aminotransferase

Discussion

In this study, it was found that BSI during VV-ECMO support for adult ARDS patients is associated with mortality. This study also found that for adult patients supported with VV-ECMO, use of a CVC for greater than 8 days was associated with increased risk of bloodstream infection. Patients with a CVC in place for greater than 8 days also experienced higher rates of adverse events such as GI bleeding. Additionally, bloodstream infection while on VV-ECMO was associated with increased rates of GI bleeding as well. These findings were consistent with previous papers that suggest infection is associated with prolonged VV-ECMO support days (21, 22), as patients in this study who developed infection had nearly triple the VV-ECMO support days. This data suggested that ECMO cannulas are unlikely sources of BSI as none of culture from ECMO cannula became positive result. However, using CVC greater than 8 days could be a source of infection. This study postulates that the high flow rates in the ECMO circuit and the cannulas prevents colonization of bacteria. Summarizing above, high infection rate with the culture positivity of CVC tip and not of ECMO components indicates that the longer duration of CVC is the risk factor of blood stream infection.

Based on existing literature, this is the first study conducted with VV-ECMO patients that examines the effect of regularly scheduled CVC replacement on infection rates. This study’s data is consistent with Yoshida, et al who reported that use of a CVC for greater than 10 days for intensive care unit patients may be a risk factor for infection (10). This suggests that scheduled replacement of CVCs in patient ECMO patients may provide a benefit in terms of both reducing infection rates and improving outcomes.

Furthermore, reports have shown that the ECMO circuit is often colonized by microorganisms, and may play a role in worsening outcomes(23). Kim et al. evaluated the frequency of cannula colonization among 47 patients who underwent ECMO. Cannula colonization was found in 6 (46.2%) of 13 patients who developed bloodstream infection, whereas cannula colonization was found in only 3 patients (8.8%) of 34 patients without BSI (24). However, it is unclear whether the cannula colonization caused or followed the bloodstream infection. On the other hand, this study did not encounter any positive cultures of either the ECMO circuit or the ECMO cannula tip. Alternatively, while not within the scope of the current study, BSI may result from other indwelling catheters(25). Therefore, these findings warrant further investigation into the source of BSI for patients supported via VV-ECMO.

The risk of catheter infection depends on the type of catheter, the insertion technique, the site of insertion, the sterility of the insertion procedure, the purpose of catheter use, site care, number of manipulations, and specific host factors(26). However, in patients supported on VV-ECMO, it has been argued that more routine sampling from CVC may artificially increase the incidence of CVC-associated BSI compared to the ECMO circuit components. It has also been suggested that high ECMO circuit flow and large cannula size reduces stasis which may result in reduced pathogen colonization. In contrast, blood stasis can result around CVC catheters. ECMO circuits have also been shown to induce homeostatic imbalance of the coagulation system that can paradoxically inhibit pathogen adhesion to the blood contact surface. Additionally, the inhibition of platelet aggregation by the contemporary ECMO circuit can also reduce the risk of bacterial colonization. Nevertheless, future studies are necessary to examine other approaches to reduce infection, including use of chlorhexidine gluconate, antimicrobial impregnation and coating of ECMO cannula tips (2729).

This study has some limitations. It studied a small number of patients at a single center which may limit the generalizability of conclusions. Furthermore, this study was conducted retrospectively and was not a randomized controlled trial. Nevertheless, this data indicates that for patients supported with VV-ECMO, scheduled replacement of CVCs prior to 8 days may be associated with decreased risk of bloodstream infection. Given that infectious complications are one of the main contributors to morbidity and mortality associated with VV-ECMO, this approach may improve outcomes associated with mechanical life support.

Supplementary Material

Supplemental Table 1
Supplemental Table 2
Supplemental Table 3
Supplemental Table 4

Acknowledgments

The authors would like to thank Ms. Elena Susan and Ms. Colleen McNulty for administrative assistance in the submission of this manuscript.

Funding:

AB is supported by National Institutes of Health HL145478, HL147290, and HL147575.

Footnotes

This work was previously presented as poster presentation at Annual Scientific Meeting of American Society of Artificial Internal Organs (Chicago, June 10th, 2020)

Conflict of Interest: The authors have no conflict of interest to declare.

REFERENCES

  • 1.Raman L, Dalton HJ. Year in Review 2015: Extracorporeal Membrane Oxygenation. Respir Care 2016; 61: 986–991. [DOI] [PubMed] [Google Scholar]
  • 2.Ramanathan K, Antognini D, Combes A, Paden M, Zakhary B, Ogino M, MacLaren G, Brodie D, Shekar K. Planning and provision of ECMO services for severe ARDS during the COVID-19 pandemic and other outbreaks of emerging infectious diseases. The Lancet Respiratory Medicine 2020; 8: 518–526. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Huang L, Zhang W, Yang Y, Wu W, Lu W, Xue H, Zhao H, Wu Y, Shang J, Cai L, Liu L, Liu D, Wang Y, Cao B, Zhan Q, Wang C. Application of extracorporeal membrane oxygenation in patients with severe acute respiratory distress syndrome induced by avian influenza A (H7N9) viral pneumonia: national data from the Chinese multicentre collaboration. BMC Infect Dis 2018; 18: 23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Li X, Guo Z, Li B, Zhang X, Tian R, Wu W, Zhang Z, Lu Y, Chen N, Clifford SP, Huang J. Extracorporeal Membrane Oxygenation for Coronavirus Disease 2019 in Shanghai, China. Asaio j 2020; 66: 475–481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Barbaro RP, MacLaren G, Boonstra PS, Iwashyna TJ, Slutsky AS, Fan E, Bartlett RH, Tonna JE, Hyslop R, Fanning JJ, Rycus PT, Hyer SJ, Anders MM, Agerstrand CL, Hryniewicz K, Diaz R, Lorusso R, Combes A, Brodie D. Extracorporeal membrane oxygenation support in COVID-19: an international cohort study of the Extracorporeal Life Support Organization registry. Lancet 2020; 396: 1071–1078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Smith M, Vukomanovic A, Brodie D, Thiagarajan R, Rycus P, Buscher H. Duration of veno-arterial extracorporeal life support (VA ECMO) and outcome: an analysis of the Extracorporeal Life Support Organization (ELSO) registry. Crit Care 2017; 21: 45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Fan E, Gattinoni L, Combes A, Schmidt M, Peek G, Brodie D, Muller T, Morelli A, Ranieri VM, Pesenti A, Brochard L, Hodgson C, Van Kiersbilck C, Roch A, Quintel M, Papazian L. Venovenous extracorporeal membrane oxygenation for acute respiratory failure. Intensive Care Medicine 2016; 42: 712–724. [DOI] [PubMed] [Google Scholar]
  • 8.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: 277–281. [DOI] [PubMed] [Google Scholar]
  • 9.Schmidt M, Brechot N, Hariri S, Guiguet M, Luyt CE, Makri R, Leprince P, Trouillet JL, Pavie A, Chastre J, Combes A. Nosocomial infections in adult cardiogenic shock patients supported by venoarterial extracorporeal membrane oxygenation. Clin Infect Dis 2012; 55: 1633–1641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Yoshida J, Ishimaru T, Kikuchi T, Matsubara N, Asano I. Association between risk of bloodstream infection and duration of use of totally implantable access ports and central lines: a 24-month study. Am J Infect Control 2011; 39: e39–43. [DOI] [PubMed] [Google Scholar]
  • 11.Mer M, Duse AG, Galpin JS, Richards GA. Central venous catheterization: a prospective, randomized, double-blind study. Clinical and Applied Thrombosis/Hemostasis 2009; 15: 19–26. [DOI] [PubMed] [Google Scholar]
  • 12.Wylie MC, Graham DA, Potter-Bynoe G, Kleinman ME, Randolph AG, Costello JM, Sandora TJ. Risk factors for central line–associated bloodstream infection in pediatric intensive care units. Infection Control & Hospital Epidemiology 2010; 31: 1049–1056. [DOI] [PubMed] [Google Scholar]
  • 13.O’Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, Lipsett PA, Masur H, Mermel LA, Pearson ML, Raad II, Randolph AG, Rupp ME, Saint S, the Healthcare Infection Control Practices Advisory C. Guidelines for the Prevention of Intravascular Catheter-related Infections. Clinical Infectious Diseases 2011; 52: e162–e193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Cobb DK, High KP, Sawyer RG, Sable CA, Adams RB, Lindley DA, Pruett TL, Schwenzer KJ, Farr BM. A controlled trial of scheduled replacement of central venous and pulmonary-artery catheters. N Engl J Med 1992; 327: 1062–1068. [DOI] [PubMed] [Google Scholar]
  • 15.Kurihara C, Walter JM, Karim A, Thakkar S, Saine M, Odell DD, Kim S, Tomic R, Wunderink R, Budinger GRS, Bharat A. Feasibility of veno-venous extracorporeal membrane oxygenation without systemic anticoagulation. The Annals of Thoracic Surgery 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Wood KL, Ayers B, Gosev I, Kumar N, Melvin AL, Barrus B, Prasad S. Venoarterial-Extracorporeal Membrane Oxygenation Without Routine Systemic Anticoagulation Decreases Adverse Events. Ann Thorac Surg 2020; 109: 1458–1466. [DOI] [PubMed] [Google Scholar]
  • 17.Tomasko J, Prasad SM, Dell DO, DeCamp MM, Bharat A. Therapeutic anticoagulation-free extracorporeal membrane oxygenation as a bridge to lung transplantation. J Heart Lung Transplant 2016; 35: 947–948. [DOI] [PubMed] [Google Scholar]
  • 18.Kurihara C, Walter JM, Singer BD, Cajigas H, Shayan S, Al-Qamari A, DeCamp MM, Wunderink R, Budinger GRS, Bharat A. Extracorporeal Membrane Oxygenation Can Successfully Support Patients With Severe Acute Respiratory Distress Syndrome in Lieu of Mechanical Ventilation. Crit Care Med 2018; 46: e1070–e1073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kurihara C, Walter JM, Karim A, Thakkar S, Saine M, Odell DD, Kim S, Tomic R, Wunderink RG, Budinger GRS, Bharat A. Feasibility of Venovenous Extracorporeal Membrane Oxygenation Without Systemic Anticoagulation. Ann Thorac Surg 2020; 110: 1209–1215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure - definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004; 8: R204–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Grasselli G, Scaravilli V, Di Bella S, Biffi S, Bombino M, Patroniti N, Bisi L, Peri AM, Pesenti A, Gori A, Alagna L. Nosocomial Infections During Extracorporeal Membrane Oxygenation: Incidence, Etiology, and Impact on Patients’ Outcome. Crit Care Med 2017; 45: 1726–1733. [DOI] [PubMed] [Google Scholar]
  • 22.Aubron C, Cheng AC, Pilcher D, Leong T, Magrin G, Cooper DJ, Scheinkestel C, Pellegrino V. Factors associated with outcomes of patients on extracorporeal membrane oxygenation support: a 5-year cohort study. Crit Care 2013; 17: R73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Thomas G, Hraiech S, Cassir N, Lehingue S, Rambaud R, Wiramus S, Guervilly C, Klasen F, Adda M, Dizier S, Roch A, Papazian L, Forel J-M. Venovenous extracorporeal membrane oxygenation devices-related colonisations and infections. Annals of intensive care 2017; 7: 111–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Kim DW, Yeo HJ, Yoon SH, Lee SE, Lee SJ, Cho WH, Jeon DS, Kim YS, Son BS, Kim do H. Impact of bloodstream infections on catheter colonization during extracorporeal membrane oxygenation. J Artif Organs 2016; 19: 128–133. [DOI] [PubMed] [Google Scholar]
  • 25.Na SJ, Chung CR, Choi HJ, Cho YH, Yang JH, Suh GY, Jeon K. Blood Stream Infection in Patients on Venovenous Extracorporeal Membrane Oxygenation for Respiratory Failure. Infection control and hospital epidemiology 2018; 39: 871–874. [DOI] [PubMed] [Google Scholar]
  • 26.O’Grady NP, Alexander M, Dellinger EP, Gerberding JL, Heard SO, Maki DG, Masur H, McCormick RD, Mermel LA, Pearson ML, Raad II, Randolph A, Weinstein RA. Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control 2002; 30: 476–489. [DOI] [PubMed] [Google Scholar]
  • 27.Afonso E, Blot K, Blot S. Prevention of hospital-acquired bloodstream infections through chlorhexidine gluconate-impregnated washcloth bathing in intensive care units: a systematic review and meta-analysis of randomised crossover trials. Euro Surveill 2016; 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Swan JT, Ashton CM, Bui LN, Pham VP, Shirkey BA, Blackshear JE, Bersamin JB, Pomer RM, Johnson ML, Magtoto AD, Butler MO, Tran SK, Sanchez LR, Patel JG, Ochoa RA Jr., Hai SA, Denison KI, Graviss EA, Wray NP. Effect of Chlorhexidine Bathing Every Other Day on Prevention of Hospital-Acquired Infections in the Surgical ICU: A Single-Center, Randomized Controlled Trial. Crit Care Med 2016; 44: 1822–1832. [DOI] [PubMed] [Google Scholar]
  • 29.Lai NM, Chaiyakunapruk N, Lai NA, O’Riordan E, Pau WS, Saint S. Catheter impregnation, coating or bonding for reducing central venous catheter-related infections in adults. Cochrane Database Syst Rev 2016; 3: Cd007878. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Table 1
Supplemental Table 2
Supplemental Table 3
Supplemental Table 4

RESOURCES