A Retrospective Study of Infection in Patients Requiring Extracorporeal Membrane Oxygenation Support

The International Society for Heart and Lung Transplantation (ISHLT): 2024 infection definitions for durable and acute mechanical circulatory support devices

Infections remain a significant concern in patients receiving mechanical circulatory support (MCS), encompassing both durable and acute devices. This consensus manuscript provides updated definitions for infections associated with durable MCS devices and new definitions for infections in acute MCS, integrating a comprehensive review of existing literature and collaborative discussions among multidisciplinary specialists. By establishing consensus definitions, we seek to enhance clinical care, facilitate consistent reporting in research studies, and ultimately improve outcomes for patients receiving MCS.

Characteristics and Outcomes of Prolonged Venoarterial Extracorporeal Membrane Oxygenation After Cardiac Surgery: The Post-Cardiotomy Extracorporeal Life Support (PELS-1) Cohort Study

OBJECTIVES

Most post-cardiotomy (PC) extracorporeal membrane oxygenation (ECMO) runs last less than 7 days. Studies on the outcomes of longer runs have provided conflicting results. This study investigates patient characteristics and short- and long-term outcomes in relation to PC ECMO duration, with a focus on prolonged (> 7 d) ECMO.

DESIGN

Retrospective observational cohort study.

SETTING

Thirty-four centers from 16 countries between January 2000 and December 2020.

PATIENTS

Adults requiring post PC ECMO between 2000 and 2020.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Characteristics, in-hospital, and post-discharge outcomes were compared among patients categorized by ECMO duration. Survivors and nonsurvivors were compared in the subgroup of patients with ECMO duration greater than 7 days. The primary outcome was in-hospital mortality. Two thousand twenty-one patients were included who required PC ECMO for 0-3 days (n = 649 [32.1%]), 4-7 days (n = 776 [38.3%]), 8-10 days (n = 263 [13.0%]), and greater than 10 days (n = 333 [16.5%]). There were no major differences in the investigated preoperative and procedural characteristics among ECMO duration groups. However, the longer ECMO duration category was associated with multiple complications including bleeding, acute kidney injury, arrhythmias, and sepsis. Hospital mortality followed a U-shape curve, with lowest mortality in patients with ECMO duration of 4-7 days (n = 394, 50.8%) and highest in patients with greater than 10 days ECMO support (n = 242, 72.7%). There was no significant difference in post-discharge survival between ECMO duration groups. In patients with ECMO duration greater than 7 days, age, comorbidities, valvular diseases, and complex procedures were associated with nonsurvival.

CONCLUSIONS

Nearly 30% of PC ECMO patients were supported for greater than 7 days. In-hospital mortality increased after 7 days of support, especially in patients undergoing valvular and complex surgery, or who had complications, although the long-term post-discharge prognosis was comparable to PC ECMO patients with shorter support duration.

Incidence, risk factors and outcomes of nosocomial infection in adult patients supported by extracorporeal membrane oxygenation: a systematic review and meta-analysis

BACKGROUND

An increasing number of patients requires extracorporeal membrane oxygenation (ECMO) for life support. This supportive modality is associated with nosocomial infections (NIs). This systematic review and meta-analysis aim to assess the incidence and risk factors of NIs in adult.

METHODS

We searched PubMed, Scopus, Web of Science, and ProQuest databases up to 2022. The primary endpoint was incidence of NI. Secondary endpoints included time to infection, source of infection, ECMO duration, Intensive care and hospital length of stay (LOS), ECMO survival and overall survival. Incidence of NI was reported as pooled proportions and 95% confidence intervals (CIs), while dichotomous outcomes were presented as risk ratios (RR) as the effective index and 95% CIs using a random-effects model.

RESULTS

Among the 4,733 adult patients who received ECMO support in the 30 included studies, 1,249 ECMO-related NIs per 1000 ECMO-days was observed. The pooled incidence of NIs across 18 studies involving 3424 patients was 26% (95% CI 14-38%).Ventilator-associated pneumonia (VAP) and bloodstream infections (BSI) were the most common NI sources. Infected patients had lower ECMO survival and overall survival rates compared to non-infected patients, with risk ratio values of 0.84 (95% CI 0.74-0.96, P = 0.01) and 0.80 (95% CI 0.71-0.90, P < 0.001), respectively.

CONCLUSION

Results showed that 16% and 20% lower of ECMO survival and overall survival in patients with NI than patients without NI, respectively. However, NI increased the risk of in-hospital mortality by 37% in infected patients compared with non-infected patients. In addition, this study identified the significant positive correlation between ECMO duration and ECMO-related NI.

Metagenomic Next-generation Sequencing for Pathogen Identification in Bronchoalveolar Lavage Fluid From Neonates Receiving Extracorporeal Membrane Oxygenation

BACKGROUND

Neonates on extracorporeal membrane oxygenation (ECMO) are at high risk of infection. Rapid and accurate identification of pathogens is essential to improve the prognosis of children on ECMO. Metagenome next-generation sequencing (mNGS) has been used in recent years to detect pathogenic bacteria, but evidence for its use in neonates on ECMO is lacking.

METHODS

This retrospective study was conducted using an electronic medical record system. We analyzed the results of mNGS and conventional microbiological tests (CMTs) in bronchoalveolar lavage fluid of neonates receiving ECMO support with pulmonary infections in our hospital from July 2021 to January 2023.

RESULTS

We screened 18 ECMO-supported neonates with pneumonia for inclusion in the study. The median age of the included children was 2 (1-4) days, the median gestational age was 38.3 (33-40 +4 ) weeks, and the median weight was 3.3 (2.2-4.8) kg. The detection rate of mNGS was 77.8% (14/18), higher than the 44.4% (8/18) of CMT ( P = 0.04). A total of 20 pathogens were detected in mNGS, with the top 3 most common pathogens being Klebsiella pneumoniae , Acinetobacter baumannii and Escherichia coli . Mixed infections were found in 14 cases (77.8%), including 13 cases (72.2%) with mixed infections detected by mNGS and 7 cases (27.8%) with mixed infections detected by CMT. A total of 9 children underwent treatment changes based on mNGS results and all of them experienced relief of symptoms.

CONCLUSION

Compared with CMT, mNGS can detect pathogens earlier and more sensitively, and may play an important role in ECMO-supported neonatal pneumonia pathogen detection and optimization of antibiotic therapy.

A nomogram to predict nosocomial infection in patients on venoarterial extracorporeal membrane oxygenation after cardiac surgery

INTRODUCTION

After cardiac surgery, patients on venoarterial extracorporeal membrane oxygenation (VA-ECMO) have a higher risk of nosocomial infection in the intensive care unit (ICU). We aimed to establish an intuitive nomogram to predict the probability of nosocomial infection in patients on VA-ECMO after cardiac surgery.

METHODS

We included patients on VA-ECMO after cardiac surgery between January 2011 and December 2020 at a single center. We developed a nomogram based on independent predictors identified using univariate and multivariate logistic regression analyses. We selected the optimal model and assessed its performance through internal validation and decision-curve analyses.

RESULTS

Overall, 503 patients were included; 363 and 140 patients were randomly divided into development and validation sets, respectively. Independent predictors derived from the development set to predict nosocomial infection included older age, white blood cell (WBC) count abnormality, ECMO environment in the ICU, and mechanical ventilation (MV) duration, which were entered into the model to create the nomogram. The model showed good discrimination, with areas under the curve (95% confidence interval) of 0.743 (0.692-0.794) in the development set and 0.732 (0.643-0.820) in the validation set. The optimal cutoff probability of the model was 0.457 in the development set (sensitivity, 0.683; specificity, 0.719). The model showed qualified calibration in both the development and validation sets (Hosmer-Lemeshow test, p > .05). The threshold probabilities ranged from 0.20 to 0.70.

CONCLUSIONS

For adult patients receiving VA-ECMO treatment after cardiac surgery, a nomogram-monitoring tool could be used in clinical practice to identify patients with high-risk nosocomial infections and provide an early warning.

Nosocomial infections in adult patients receiving extracorporeal membrane oxygenation in China: A retrospective cohort study

BACKGROUND

Extracorporeal membrane oxygenation (ECMO) has been increasingly used in China, but nosocomial infections (NI) in patients receiving ECMO remain poorly characterized. Thus, this study aimed to investigate the incidence rate, causative was pathogens, and risk factors of NIs in ECMO patients.

METHODS

A retrospective cohort study of patients receiving ECMO between January 2015 and October 2021 was conducted in a tertiary hospital. General demographics and clinical data of the included patients were collected from the electronic medical record system and the real-time NI surveillance system.

RESULTS

A total of 86 infected patients with 110 episodes of NIs were identified in 196 patients receiving ECMO. The incidence of NI was 59.2/1000 ECMO days. The median time for the first NI in ECMO patients was 5 days (interquartile range: 2-8 days). Hospital-acquired pneumonia and bloodstream infections were common types of NIs in ECMO patients, and the main pathogens were gram-negative bacteria. Pre-ECMO invasive mechanical ventilation (OR = 2.40, 95% CI:1.12-5.15) and prolonged duration of ECMO (OR = 1.26, 95% CI:1.15-1.39) were risk factors for NIs during ECMO support.

DISCUSSION

This study identified the main infection sites and pathogens of NIs in ECMO patients. Although NIs may not affect successful ECMO weaning, additional measures should be implemented to reduce the incidence of NI during ECMO support.

Antifungal Dosing in Critically Ill Patients on Extracorporeal Membrane Oxygenation

Extracorporeal membrane oxygenation (ECMO) is an established advanced life support system, providing temporary cardiac and/or respiratory support in critically ill patients. Fungal infections are associated with increased mortality in patients on ECMO. Antifungal drug dosing for critically ill patients is highly challenging because of altered pharmacokinetics (PK). PK changes during critical illness; in particular, the drug volume of distribution (Vd) and clearance can be exacerbated by ECMO. This article discusses the available literature to inform adequate dosing of antifungals in this patient population. The number of antifungal PK studies in critically ill patients on ECMO is growing; currently available literature consists of case reports and studies with small sample sizes providing inconsistent findings, with scant or no data for some antifungals. Current data are insufficient to provide definitive empirical drug dosing guidance and use of dosing strategies derived from critically patients not on ECMO is reasonable. However, due to high PK variability, therapeutic drug monitoring should be considered where available in critically ill patients receiving ECMO to prevent subtherapeutic or toxic antifungal exposures.

Impact of Nosocomial Infection on in-Hospital Mortality Rate in Adult Patients Under Venoarterial Extracorporeal Membrane Oxygenation After Cardiac Surgery

Objective

There was no consensus on the impact of nosocomial infection on In-hospital mortality rate in patients receiving ECMO. This study aimed to investigate the impact of nosocomial infection (NI) on In-hospital mortality rate in adult patients receiving venoarterial extracorporeal membrane oxygenation (VA-ECMO) after cardiac surgery.

Materials and Methods

This retrospective study included 503 adult patients who underwent VA-ECMO after cardiac surgery. The impact of time-dependent NIs on In-hospital mortality rate within 28 days of ECMO initiation was investigated using a Cox regression model. The cumulative incidence function for death was compared between patients with NIs and those without NIs using a competing risk model.

Results

Within 28 days after ECMO initiation, 206 (41.0%) patients developed NIs, and 220 (43.7%) patients died. The prevalence rates of NIs were 27.8% and 20.3% during and after ECMO therapy, respectively. The incidence rates of NIs during and after ECMO therapy were 49‰ and 25‰, respectively. Time-dependent NI was an independent risk factor for predicting death (hazard ratio = 1.05, 95% confidence interval = 1.00-1.11). The cumulative incidence of death in patients with NI was significantly higher than that in patients without NI at each time point within 28 days of ECMO initiation. (Z = 5.816, P = 0.0159).

Conclusion

NI was a common complication in adult patients who received VA-ECMO after cardiac surgery, and time-dependent NI was an independent risk factor for predicting mortality in these patients. Using a competing risk model, we confirmed that NIs increased the risk of In-hospital mortality rate in these patients.

Clearance of blood stream infections in patients receiving extracorporeal membrane oxygenation: a retrospective single-center cohort study

BACKGROUND

There are limited data on the treatment of blood stream infections (BSIs) in patients receiving extracorporeal membrane oxygenation (ECMO). Current guidance recommends documenting clearance only in fungal and Gram-positive BSIs. This study investigates the incidence and clinical significance of blood stream infections with positive repeat cultures (BSIPRC) in ECMO as well as clinical factors that may predict positive repeat cultures.

METHODS

All BSIs in patients receiving ECMO at Brooke Army Medical Center between September 2012 and October 2021 were included in this study. BSIPRC was defined as re-isolation of the same organism on repeat blood cultures following an initial positive blood culture.

RESULTS

A total of 60 patients developed 87 BSI (38.5 BSI per 1000 ECMO days). Of the 80 (92%) BSIs who had repeat blood cultures drawn, patients had BSIPRC in 35 (44%) of cases. Fever, leukocytosis, and vasopressor requirement on day of repeat culture were not associated with persistent positivity. There was no difference in survival to discharge for patients with BSIPRC as compared to single day BSI (58% vs. 63%, p = 0.78). 19% of patients with Gram-negative bacteremia had BSIPRC, and gram-negative bacteremia in general was associated with an 83% morality.

CONCLUSIONS

There were no clinical findings that differentiated patients with BSIPRC from those who had a single day of positivity. BSI was associated with high mortality in patients with Gram-negative bacteremia. Given high incidence of positive repeat cultures being seen in Gram-negative BSIs, repeat blood cultures have utility for all BSIs in patients receiving ECMO.

No increase of device associated infections in German intensive care units during the start of the COVID-19 pandemic in 2020

Background

The COVID-19 pandemic may have had a substantial impact on the incidence of device-associated healthcare-associated infections (HAI), in particular in intensive care units (ICU). A significant increase of HAI was reported by US hospitals when comparing incidence rates from 2019 and 2020. The objective of this study was to investigate the development of the most relevant device-associated HAI in German ICUs during the year 2020 as compared to 2019.

Methods

We utilized the data of the ICU component of the German National Reference Center for Surveillance of Nosocomial Infections (KISS = Krankenhaus-Infektions-Surveillance-System) for the period 2019–2020. We focused on central line-associated bloodstream infections (CLABSI), catheter-associated urinary tract infections (CAUTI), ventilator-associated lower respiratory infections (VALRTI) and bloodstream infections associated with the use of Extracorporeal-Life-Support-Systems (ECLSABSI). Device use was defined as the number device days per 100 patient days; device-associated infection rates as the number of device-associated infections per 1000 device days. To compare the pooled means between the years and quarters we calculated rate ratios of device-associated infection rates with 95% confidence intervals by Poisson regression models.

Results

The number of participating ICUs in the surveillance system decreased from 982 in 2019 to 921 in 2020 (6.2%). Device utilization rates increased significantly for central lines and ventilator use. VALRTI rates and CAUTI rates decreased in 2020 compared with 2019, however, no increase was shown for CLABSI or ECLSABSI. This result was also confirmed when the corresponding quarters per year were analyzed.

Conclusions

The lack of an increase in device-associated healthcare associated infections (HAI) in German ICUs may be due to the lower overall incidence of COVID-19 cases in Germany in 2020 compared with US, to a very high availability of ICU beds per 100,000 inhabitants compared with many other countries, and a change in the ICU patient mix due to numerous elective procedures that were postponed during the first two waves. The primary reason seems to be that only 7% of all ICU patients in Germany in 2020 were COVID-19 patients.

Nosocomial infections during extracorporeal membrane oxygenation

PURPOSE OF THE REVIEW

The aim of this review is to present the latest evidence regarding the prevention, diagnosis and treatment of nosocomial infections during extracorporeal membrane oxygenation (ECMO) support.

RECENT FINDINGS

New descriptive data from the Extracorporeal Life Support Organisation (ELSO) registry and single centre studies have been published. In parallel, there is an increase in the availability of evidence about the diagnostic accuracy of infection markers, yield of routine cultures, effectivity of antibiotic prophylaxis and other preventive measures.

SUMMARY

ECMO is a rescue therapy for severe hemodynamic or respiratory failure. Nosocomial infections on ECMO support are frequent (infection rate ranging between 20.5% to more than 50% of ECMO runs) and have impact in survival, with reported increases in the risk of death up to 63% in infected patients. However, diagnosis and treatment are challenging, as the unique relationship between patient and circuit may act as a confounder for infection and exacerbate the variability of antibiotic pharmacokinetics in critical illness. Clinical practice regarding antibiotic treatment and infection prevention is not yet supported by high-quality evidence.

Clinical and microbiological characteristics of and risk factors for bloodstream infections among patients with extracorporeal membrane oxygenation: a single-center retrospective cohort study

Extracorporeal membrane oxygenation (ECMO) provides hemodynamic and oxygenation support to critically ill patients. Due to multiple catheter cannulations, patients on ECMO are vulnerable to bloodstream infections (BSIs). We aimed to investigate the incidence, clinical characteristics, risk factors, and microorganisms associated with BSIs during ECMO. This single-center retrospective cohort study was conducted between January 2015 and May 2021. Patients aged 18 years or older with an ECMO duration of > 48 h for cardiogenic or respiratory support were included in the study. Patients who developed bacteremia or candidemia from 12 h after ECMO cannulation to 7 days after de-cannulation were included. The clinical factors between non-BSI and BSI were compared, along with an analysis of the risk factors associated with BSI during ECMO. A total of 480 patients underwent ECMO for cardiogenic shock (n = 267, 55.6%) or respiratory failure (n = 213, 44.4%) during the study period. The incidence was 20.0 episodes per 1000 ECMO-days. Approximately 20.2% (97/480) and 5.4% (26/480) of the patients developed bacteremia and candidemia, respectively. The median numbers of days of BSI development were 8.00 days for bacteremia and 11.0 days for candidemia. The most common pathogens were methicillin-resistant coagulase-negative staphylococci (n = 24), followed by vancomycin-resistant Enterococcus (n = 21). Multivariable logistic analysis demonstrated that hemodialysis (odds ratio [OR] 2.647, p < 0.001), veno-arterial-venous mode (OR 1.911, p = 0.030), and total ECMO duration (OR 1.030, p = 0.007) were significant risk factors for bacteremia. The total ECMO duration was the only risk factor associated with candidemia (OR 1.035, p = 0.010). The mortality rate was significantly higher in the bacteremia (57.7%) and candidemia (69.2%) groups than that in the non-BSI group (43.6%). BSI is a common complication of patients receiving ECMO support and is associated with poor clinical outcomes. Determining the type of frequently isolated organisms and the median onset time of BSI would help in the selection of appropriate prophylactic antibiotics or antifungal agents.