When ECMO can no longer bridge: a narrative review and proposed consent template

1 Introduction

The use of extracorporeal membrane oxygenation (ECMO) has expanded over the past decade, providing life-sustaining support for patients with severe cardiac and/or respiratory failure. While ECMO may serve as a bridge to recovery, transplantation, or durable mechanical circulatory support, a subset of patients will not successfully wean from ECMO or will not meaningfully recover. In these circumstances, complex clinical, ethical, and logistical challenges arise surrounding end-of-life decision-making.

There are currently no widely accepted guidelines or standardized approaches for ECMO withdrawal in this patient population. Consequently, practices vary across institutions, and navigating withdrawal of ECMO support or even who to place on ECMO is challenging in the absence of consensus.

The objective of this review is to map and characterize the epidemiology and incidence of ECMO withdrawal, transplant eligibility, and practices surrounding consent and goals-of-care. We finish with an assessment of current practices in ECMO withdrawal and provide an ECMO-specific consent template (Supplementary Table S1) that operationalizes the current state of the literature on this topic.

2 Methods

2.1 Search strategy

A narrative review was conducted to identify literature relevant to the withdrawal of ECMO in adult patients who could not bridge to recovery, transplantation, or durable mechanical support. The aim was to characterize existing evidence, clinical practices, and institutional approaches and protocols related to ECMO withdrawal.

An extensive literature search was performed using PubMed, Embase, Scopus, and JSTOR from ECMO inception through June 2026. Search terms included combinations of the keywords and Medical Subject Headings (MeSH) terms “extracorporeal membrane oxygenation”, “ECMO”, “extracorporeal life support”, “ECLS”, “end-of-life care”, “goals of care”, “bridge to nowhere”, “transplant eligibility”, “withdrawal of life-sustaining treatment”, “compassionate ECMO discontinuation”, “palliative care”, and “withdrawal of care”. Professional society guidelines were also reviewed to identify additional sources.

A meta-analysis was not performed because studies addressing ECMO withdrawal employ heterogenous outcome definitions across distinct populations of VA-ECMO, VV-ECMO, and ECPR, and without randomized comparison data. This precluded meaningful quantitative pooling of the available data. Furthermore, the objective of this narrative review extended beyond isolated quantifiable outcomes to encompass a multi-faceted synthesis of epidemiology, transplant eligibility, communication frameworks, and technical approaches to ECMO discontinuation. A scoping review methodology was also considered and similarly not pursued. The objective of this review was not to catalogue the extent of existing literature, rather to synthesize a fragmented evidence base into clinically useful information for practicing clinicians and an actionable tool—the ECMO Consent Template presented in Supplementary Table S1.

2.2 Eligibility criteria

Sources were selected for inclusion if they addressed ECMO withdrawal, end-of-life decision-making during ECMO support, transplant or durable mechanical circulatory support eligibility, or institutional protocols for ECMO withdrawal. Clinical practice guidelines, review articles, commentaries, registry analyses, and original research studies were included. Sources addressing non-English-language publications, conference abstracts, pediatric-only populations, animal and pre-clinical studies were excluded.

2.3 Study selection

Given the diverse and limited literature specifically addressing ECMO withdrawal, sources were selected non-systematically by the author team based on relevance to the review objectives and clinical significance, rather than through a structured or reproducible screening protocol. This approach allowed for synthesis of clinical practice patterns, institutional protocols, and expert commentary across a the broad evidence base. This does not represent a comprehensive or exhaustive account of all available literature.

3 Epidemiology and incidence of ECMO withdrawal

The application of ECMO as a tool for bridge to transplant, recovery, or decision, has grown exponentially in the last decade. Yet, survival rates remain modest. In the most recent ELSO Registry International Summary, survival rates range from 60% among those on VV-ECMO support to 49% for those supported by VA-ECMO, with just 31% survival among those cannulated in the setting of ECPR (1, 2). Thus, clinically safe and ethically sound withdrawal from life sustaining treatment (WLST), specifically from ECMO, remains a priority in present and future practice. There have been relatively few studies that have specifically addressed the prevalence and timing of ECMO withdrawal among patient populations. The decision for WLST from ECMO varies widely, and is often specific to a given case, indication, and institutional practices. At the very least, discussion regarding WLST must involve the patient if they have capacity or their surrogate decision makers. What follows is a review of the current literature on the epidemiology of ECMO withdrawal (3).

A 2019 single-center study found that withdrawal from ECMO support was requested in just over 25% of supported patients (4). Of 235 adult patients, 81% received VA-ECMO support and 26% were cannulated in the setting of ECPR. For those patients in which WLST were requested, a majority were placed on ECMO support as a bridge to decision (82%), while a smaller percentage were placed on ECMO as a bridge to transplant or further long term mechanical circulatory support. In all cases, the decision to withdraw ECMO support was made in collaboration with the patient's surrogate decision makers, as none of these patients had decision making capacity at the time of WLST. Among those who underwent WLST, only 19% had clinical documentation suggesting futility for continued support (4).

Carlson et al. 2022 conducted a similar retrospective cohort study evaluating the population characteristics and markers for the WLST in those requiring ECMO support (5). Their study included 150 patients at a single academic center, 80% of which were cannulated for VA-ECMO. Of those studied, nearly half experienced WLST, with withdrawal occurring at a median of day 5 of ECMO therapy. Of those experiencing WLST, nearly 45% underwent withdrawal of ECMO support within the first 5 days (termed “early withdrawal”), with 25% undergoing withdrawal within the first 72 h. WLST was far more frequent among those on VA-ECMO (90.4% of patients undergoing WLST) compared to those cannulated for VV-ECMO. Factors independently associated with WLST included older age, elevated BMI, higher Acute Physiology and Chronic Health Evaluation (APACHE-II) scores, and cannulation for VA-ECMO. Authors further investigated the reasoning behind withdrawal of life sustaining therapies, noting family requests in >90% of cases. Multi-organ failure and concern for primary organ damage determined to be non-survivable or inoperable were other commonly cited reasons for WLST (in 48% and 41% of cases respectively). Neurologic compromise was observed in only 9% of the cohort (5).

Similarly, a 2025 retrospective observational study including over 23,000 patients from the ELSO registry found that older age, higher BMI, cardiac arrest before ECMO initiation, and renal failure were all associated with increased odds of death following liberation from ECMO support. Of the patients who died within their cohort, 43.7% died in the context of withdrawal of ECMO. For this study, withdrawal had a broader definition and included those with poor prognosis followed by death < 72 h after withdrawal, those with ECMO complications followed by death < 24 h after withdrawal, and those who died < 12 h following ECMO withdrawal (6).

A 2019 multicenter international survey of more than 500 physicians, respondents identified patient preferences, underlying comorbidities, and the etiology of respiratory failure as the factors most frequently considered when evaluating withdrawal of VV-ECMO support. Notably, more than 90% of physicians reported involving other clinicians in the decision-making process, highlighting the multidisciplinary nature of ECMO withdrawal discussions (7).

Another retrospective study published in 2022 utilized ELSO registry data and specifically investigated factors associated with early withdrawal of life sustaining therapies, defined as withdrawal <72 h following ECMO cannulation in the setting of ECPR. Between 2007 and 2017, 411 ECPR cases were identified. Of those, > 50% underwent early withdrawal, associated with statistically significant increased rates of early withdrawal for those with low pH at time of ECPR, lower pH at 24 h, increased ventilatory requirements at 24 h, lower MAP, and increased rates for all ECMO related complications (8). Such findings raise concerns that withdrawal may be occurring too early, before proper post-arrest prognostication can be completed.

Few studies exist that specifically investigate withdrawal of temporary support among those with decision making capacity. A small, single center retrospective case series published in 2022 reviewed cases of temporary MCS, including VA-ECMO, RVAD, LVAD, or BiVAD therapies over a 13-year period. Of 796 patients included in the study, 178 patients underwent WLST, and only 6 of these cases involved patients with decision making capacity at the time of WLST. Of these cases, uncertainty of path to recovery, progressive disability, desire to no longer be supported by machines, and decline in overall quality of life were factors that appeared to drive these decisions. In a majority of these cases, the decision to WLST occurred after they were found not to be a candidate for long term LVAD therapy or transplant (9).

In nearly all studies described above, the decision to withdraw therapies were primarily driven by family or surrogate decision makers. Overall, WLST will continue to occur in those supported by temporary MCS, specifically VA-ECMO support.

4 Transplant eligibility

A critical inflection point in the ECMO course occurs when a patient is determined ineligible for transplantation or durable MCS. At this moment, ECMO transitions from a bridge therapy to a device providing futile care. Understanding the criteria that define this transition is therefore essential to the withdrawal discussion. Eligibility criteria for solid organ transplants and durable mechanical circulatory support in the form of left ventricular devices or total artificial hearts vary across professional society guidelines and transplant centers. Although guidelines delineate both absolute and relative contraindications, specific criteria and the interpretation vary by institution, particularly for lung and heart transplantation. Furthermore, consideration for transplant eligibility requires comprehensive pre-transplant screening, including laboratory tests, imaging, frailty assessment, and multidisciplinary review (10).

Lung transplantation eligibility follows guidelines outlined by the International Society for Heart and Lung Transplantation (ISHLT), American Thoracic Society (ATS), and the European Respiratory Society (ERS). Eligibility assessment incorporates contraindications, frailty, functional status to estimate waitlist mortality and post-transplant survival. Overall, transplantation is considered if the patient has an estimated >50% chance of dying within 2 years without transplant and >80% likelihood of survival of 5-year survival post-transplant. Absolute contraindications include lack of patient willingness, malignancy with high recurrence risk, severe organ dysfunction or acute organ failure, disseminated infection, tuberculosis, uncontrolled HIV with a detectable viral load, and septic shock, active substance use, or other medical condition expected to limit survival post-transplant (11).

For heart transplantation, the American College of Cardiology identifies systemic illness with a life expectancy <2 years and irreversible pulmonary hypertension as absolute contraindications. Relative contraindications include diabetes with end-organ damage, severe peripheral vascular or cerebrovascular disease, morbid obesity or cachexia, pulmonary, renal, and hepatic dysfunction. Drugs, tobacco, and alcohol use within 6 months, mental illness or psychosocial instability are all listed as relative contraindications (12, 13).

However, there is rarely one factor determining patient ineligibility, rather a combination of patient comorbidities, functional status, psychosocial factors, and predicted survival in the absence of transplantation and with transplantation (14). Eligibility is not a static, one-time determination made prior to cannulation. A substantial portion of ECMO cannulations occur under circumstances that preclude prior transplant eligibility assessment; in these cases, ECMO is leveraged as a bridge-to-decision specifically to preserve organ function while allowing time for transplant eligibility assessment (4). Even among patients confirmed eligible at cannulation, eligibility may be lost during the clinical course as ECMO-bridged transplant patients experience substantial morbidity and mortality such as stroke, multi-organ failure, uncontrolled infection, or prolonged deconditioning (15–18). WLST following a failed bridge-to-transplant is a common and clinically significant scenario.

5 Consent, communication, and goals of care

Framing ECMO as a bridge therapy with defined exit criteria at the outset helps delineate goals-of-care and the circumstances under which withdrawal may be medically or ethically appropriate. A recurring theme in the literature is that downstream conflict regarding goals-of-care, futility and withdrawal are rooted in upstream failures of communication at the time of consent (7, 12, 19). Abrams et al. 2014 first drew attention to preparing the patient and families upfront, i.e., at the time of cannulation, for scenarios in which ECMO may fail to achieve its intended goals (20).

As Abrams et al. noted “the greater the effort given to obtaining informed consent before ECMO, the easier it may be to reach a consensus on withdrawal later if ECMO is no longer achieving its intended goals” (20). Despite this early call to action, a follow-up international survey by Abrams et al. 2019 found 45% of respondents sometimes, rarely or never discussed the possibility of ECMO withdrawal (7). The “Extracorporeal Life Support Statement of Purpose” published by Adams et al. represents one of the first of such documents stating the goals and consequences of ECMO (20).

However, ECMO cannulation frequently occurs under conditions that preclude a traditional informed consent discussion. DeMartino et al. 2019 reported that among patients from whom ECMO was ultimately withdrawn, 82% of patients were cannulated as a “bridge to decision”, and 26% were cannulated during active CPR (4). Notably, no patient who underwent withdrawal had decisional capacity at the time of the withdrawal request. This finding is consistent with the broader observation that ECMO is often instituted without adequate time for detailed discussion, resulting in unclear perceptions of goals, poor understandings of limitations, and ethically challenging situations when withdrawal becomes advisable (12).

The clinical reality of ECMO initiation necessitates a reconceptualization of consent for ECMO use, not as a single pre-procedural event, but as a longitudinal, iterative process of communication, prognostic update, and shared decision-making that begins at or shortly after cannulation and continues throughout the ECMO run. Institutions should consider developing an ECMO-specific consent form to standardize the information conveyed to patients and families (12). Moynihan et al. 2021 proposed a structured conversation guide for longitudinal consent and communication during pediatric ECMO with elements directly applicable to adult care, organizing communication into Day 0, Day 1, and Day 2 + conversations (21). When the point of futility is reached, the Compassionate ECMO Discontinuation framework described by Machado et al. in 2022 provides a structured interdisciplinary approach including end-of-life discussion, clinical preparation, technical components, and bereavement support (19). The recently published 2025 ENSURE survey revealed significant regional and cultural variability in end-of-life practices, further underscoring that this process is best guided by institution-specific guidelines adaptable to local legal, ethical, and institutional contexts (22).

For urgent ECMO cannulations, such as a bridge to transplant or bridge to recovery in a decompensating but stable patient, the full scope of informed consent should be addressed with the patient and/or surrogate prior to cannulation. In 2023, Childress et al. argued that the failure to establish clear goals and exit criteria at the outset is what transforms ECMO from a bridge therapy into a de facto destination therapy, creating ethically intractable situations when capacitated patients later refuse withdrawal (23). Prospective goal setting at the time of informed consent should include a discussion of what constitutes an unacceptable outcome and under what circumstances ECMO support be discontinued (21).

For emergent cannulations, such as ECPR, failure to wean from bypass post cardiotomy, or acute cardiogenic shock, meaningful pre-procedural consent is impossible. For these scenarios we recommend a two-phase consent model to discuss the withdrawal of ECMO support. Phase one occurs at the time of cannulation and frames ECMO as a time-limited intervention initiated to allow prognostic clarification. Phase two occurs within 24–48 h of cannulation and consists of a formal goals-of-care conversation once the patient has stabilized and the clinical trajectory more apparent. This approach is supported by DeMartino et al. 2019, who argued that offering a trial of therapy to patients with uncertain prognoses, where some hope of recovery exists, is nevertheless preferable to withholding the therapy altogether. Additionally, withdrawing a treatment initiated under emergency conditions, once a poor prognosis has become evident, is better than not having offered that treatment if there was hope of clinical benefit (4). The 2020 EACTS/ELSO/STS/AATS Expert Consensus similarly endorses discontinuation of ECLS when it fails as a means to recovery, transplant, or transition to durable mechanical support (24).

Once cannulated, goals of care become an iterative process rather than a single decision point. The 2025 AHA scientific statement on palliative care in the cardiovascular ICU advocates for early and regular family meetings for high-risk patients, with inclusion of a “family meeting plan” in the daily ICU bundle (25). Frequent family meetings should start within one week or earlier of ECMO initiation and occur weekly thereafter (26, 27). Normalizing patients and families to regular meetings from the outset helps elicit values, goals, and preferences, particularly when a patient is unlikely to survive the hospitalization (26, 28). Setting expectations for acute MCS before its initiation, or as soon as possible thereafter, diminishes suffering, lowers decisional regret, and limits disagreement over eventual withdrawal of support should the situation become futile (28).

When the goals of ECMO can no longer be met, the groundwork laid during the initial consent process provides the ethical and relational foundation for a compassionate transition to comfort-focused care. We propose that the ECMO consent process be reframed not merely as a procedural authorization but as a goals-of-care discussion encompassing: When the elements above are absent, clinicians and families are at risk of decisional incongruence. To this end, we have operationalized these key elements in a recommended ECMO Consent Template provided in Supplementary Table S1.

• Articulation of the intended bridge strategy which might include recovery, transplant, durable mechanical support, or decision (12, 20).
• Prognostication, including the possibility that ECMO may not achieve its goals (4, 21).
• Definition of conditions under which ECMO would be considered non-beneficial and withdrawal recommended (20, 23).
• Discussion of expected complications of ongoing ECMO support (12, 21).
• Identification of a surrogate decision-maker and documentation of patient's values regarding quality of life and acceptable outcomes.

6 Discontinuation of ECMO

Once a “bridge to nowhere” situation has been identified, withdrawal from ECMO presents a deeply personal and emotional process for the patient, family, and bedside staff, and a technically sophisticated and complex procedure. Historically, literature has focused on whether withdrawal is appropriate rather than how it should be carried out (20, 29). Filling this void is institutional culture and individual clinical judgment utilizing frameworks adapted from broader guidelines on WLST (30, 31).

In 2022, Machado et al. advanced this discussion with an ECMO-specific Compassionate ECMO Discontinuation (CED) checklist (19). To our knowledge, this is the only structured, interdisciplinary and published tool designed specifically for ECMO withdrawal. This framework organizes the withdrawal process into four domains: end-of-life discussions, clinical preparation and pharmacological management, technical circuit discontinuation, and bereavement support.

Beyond this tool, the literature consists of case series, ethical commentaries and survey data (21, 32, 33). The subsections that follow use the CED framework for structure and integrate the broader critical care literature to address: pre-withdrawal preparation, pharmacological management, technical approach, and post-withdrawal care.

6.1 Pre-withdrawal preparation: confirming consensus amongst team and family

The core issue prior to withdrawal is that of achieving consensus amongst clinical team, patient and family (19, 30). However, achieving a consensus often requires time for multidisciplinary clinical evaluation, family meetings, and logistical planning. These components improve the quality of a patient's death and reduce moral distress amongst staff and family. Families may be asked to absorb prognostic information, make goals-of-care decisions, and prepare for a loved one's death within a compressed time frame. This is a recipe for complicated grief and PTSD (19). Due to the uniqueness of each ECMO case, there is no published evidence on the optimal timing, rather broad recommendations that the possibility of futility and withdrawal be introduced in the peri-cannulation phase (12). Additionally, futility may be defined differently amongst clinicians and the staff caring for the patient. Thus clinical teams often manage this tension without a validated framework.

In order to alleviate conflict over ECMO withdrawal, preparation should be multidisciplinary. A consensus of attending physician, bedside nursing, ECMO specialist, palliative care, not necessarily the former instructing the latter. Each discipline contributes material insight: the patient's symptom burden, the family's emotional readiness, the circuit configuration and function, and the presence of lack of anticipatory guidance (19, 30).

Family meetings should focus on a shared understanding of prognosis rather than documenting consent for withdrawal. The available research on family experience following ECMO withdrawal identifies the following sources of trauma: 1) shock at the speed of death, 2) confusion about the withdrawal sequence, and 3) a sense that a decision was made without full comprehension (19). Pre-withdrawal family meetings should prioritize anticipatory guidance and answer questions families are likely to have such as: What will the withdrawal process look like? How long is the patient expected to survive? What is the anticipated time from circuit discontinuation to death? What will the patient appear to experience and what will be done to prevent suffering? Which other LST will be discontinued and when? Logistical preparation should favor scheduling withdrawal during daytime hours where possible, silencing of alarms, removing non-essential equipment, and honoring cultural and religious practices (19, 30).

6.2 Pharmacological management

ECMO poses specific pharmacological circumstances which include: configuration of cannulae, centrifugal pumps, oxygenator membranes, mechanical ventilation, and the sudden reduction or loss of cardiac output. It is therefore imperative that symptom management during ECMO withdrawal be anticipatory rather than reactive (34). Nerness et al. recommends comfort medications be administered at least 10 min prior to circuit disruption and allow time to circulate before flow is reduced (35). Vasoactive medications should be weaned concurrently with or prior to flow reduction (19). Nurses should have standing orders with clear titration parameters. Opioids and benzodiazepines are the cornerstone of WLST.

The bedside team must balance the obligation to prevent suffering with the ethical imperative not to hasten death. The principle of the “double effect” provides robust ethical support for opioid and sedative use with the intent of relieving dyspnea and pain (30, 36).

6.3 Technical approaches to circuit discontinuation

The technical approach to circuit disconnection differs meaningfully between VV and VA configurations, and between patients with and without native function. Withdrawal from VV-ECMO can be accomplished by: 1) Sequential reduction in sweep gas flow, reducing the delivery of assisted ventilation and oxygenation; 2) Simultaneous cessation of both pump flow and sweep. The circuit may remain clamped and in place until death (19). For fully supported patients on VA-ECMO, hemodynamic collapse is expected and may be rapid. The central question is whether to gradually reduce flow over 15–30 min or more rapidly. Gradual flow reduction allows time for concurrent titration of comfort medications and time for family preparation. A more rapid de-escalation accelerates death but may further compress the window for effective symptom management. The approach will be individualized and explicitly discussed with family prior to withdrawal (19).

6.4 Post-withdrawal: bereavement

Post-withdrawal care encompasses both the patient's family and clinical teams. Family should be focused on allowing families uninterrupted time at the bedside. Cultural and religious practices explored in advance should be honored. A bereavement follow up contact, whether it be a phone call or letter acknowledges the significance of the care and provides families with an opportunity to raise questions they may not have asked in the acute setting (19, 30).

Additionally, staff debriefings should be structured and formalized within 24–72 h of patient death (29, 37). These should be interdisciplinary, voluntary, and led by trained facilitators. Moral distress from clinical staff associated with ECMO withdrawal is well-documented (37, 38). It is cited in 82% of ECMO ethics consultations in one series (39) and is independently associated with burnout among critical care providers (37). The ERAS Cardiac Society consensus and the ENSURE survey both emphasized the need for structured interventions to mitigate moral distress in ECMO end-of-life decision-making (22, 40).

7 Gaps, controversies and future directions

In an ideal state, extracorporeal life support would only be offered to individuals with clear indications that include organ recovery or bridge to transplantation or durable mechanical support. This would be considered in concert with a detailed understanding of that individual's goals of care. These decisions would be based on well-vetted criteria that include expected recovery or progression of disease and other patient-related factors such as co-morbidities, age, frailty, other acute organ dysfunction, overall prognosis and potential for rehabilitation (41). Further, this would be considered in the context of the ability to predict and mitigate the risk associated with prolonged mechanical support for the specific ECLS candidate. Such an approach would seek to maximize the likelihood of meaningful recovery and limit the number of patients placed on ECLS resulting in a futility, thereby decreasing the need for the often-difficult decision making involving WLST. The current state is not congruent with this idealized vision due to limitations in knowledge, lack of standard definitions, controversies involving differences of opinion during uncertain clinical circumstances, culture, religion and other ethical considerations. These gaps, however, provide a robust starting point for future exploration.

Depending on the clinical scenario, ECLS may be initiated in controlled situations following multidisciplinary discussion that includes shared decision-making with the patient or surrogate, or it may occur as part of an extremely time pressured scenario with limited information about the patient, their goals, or elucidation of a precise etiology of the patient's decompensation. There are clearly many clinical scenarios that lie in between. What they all share to varying degrees is prognostic uncertainty. Deficiencies in accurate prognostication were identified as one of four key causes of ethical and moral distress in ECMO decision making in a national survey of experienced ECMO team members (42). Future research directed at clearly defining criteria associated with meaningful recovery or adverse outcome is an obvious priority. While many prediction models have been published for both VV and VA ECMO, most included patients already on ECMO limiting ability to predict pre-cannulation outcomes (43). Artificial intelligence and machine learning methods are under active investigation and have potential to contribute in this area (44). It is essential to understand what variables are associated with favorable outcomes in the selection of ECLS candidates including the development of tools that can be effectively applied with established validity in time-sensitive situations. While application of prediction models is unlikely to fully inform any specific decision to place a patient on ECLS, such tools may aid in the communication and understanding of expectations with patients, their families and the care team and may be helpful in decreasing the need for WLST through more selective use of ECLS. Regardless of the methodology employed, any prediction model requires rigorous validation and determination of how to best apply the information to specific clinical scenarios.

The concept of futility of care is also a key contributor to moral distress (42). However, there is no clear consensus on what defines futility. This is a particularly controversial area owing to cultural, religious and clinical differences of opinion where disagreement may result from hope for unexpected or unlikely outcomes (45). Complicating this issue is the fact that no timelines are universally accepted to limit duration of therapy or further escalation in care such as limitation of oxygenator exchanges (46). Additionally, some centers have successfully supported the occasional patient with exceptionally long ECMO runs that have resulted in good outcomes. These experiences have the potential to influence decision-making around WLST.

Issues around effective communication are another source of controversy. These often involve concern over a lack of understanding of patient wishes and goals of care and disconnects between expected outcomes (33, 42). Integration of palliative care teams early in the course of care with regular scheduled multidisciplinary meetings have been proposed as a process improvement and has been recommended in organizational statements based on expert consensus opinion (25). While empirically valuable, it is unclear if these interventions lead to easier resolution of conflict and decreased moral distress. Some organizations utilize mandatory palliative care consultation for any patient supported with ECLS, but this is far from standard (42). There remains concern that palliative care consultation carries a negative connotation for families and the care team with some interpreting palliative care involvement as a sign of giving up hope for survival. Future work should seek to define measurable outcomes related to enhanced communication, the best techniques for communication and team composition to facilitate the most effective communication.

8 Conclusions

The withdrawal of ECMO in patients who cannot be bridged to recovery, transplantation, or durable mechanical circulatory support remains a complex and understudied area of critical care. Existing evidence emphasizes the importance of multidisciplinary collaboration, ongoing goals-of-care discussions, proactive symptom management, and support for patients, families, and healthcare professionals throughout the withdrawal process.

In the absence of standardized national or international guidelines, significant variability exists across centers, and important knowledge gaps remain regarding communication, technical withdrawal practices, and bereavement support. Until more robust evidence is available, ECMO centers should consider developing institution-specific policies to guide the management of patients for whom ECMO has become a futile life-sustaining therapy promoting consistent, patient-centered care while reducing moral distress and improving end-of-life outcomes. To support this, we provide a consent template (Supplementary Table S1) operationalizing the elements outlined in Section 5.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fanes.2026.1944591/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Fabrizio Monaco, University of Bologna, Italy

Reviewed by: Patrick Scheiermann, LMU Munich University Hospital, Germany

Eva Alvarez Torres, University Hospital October 12, Spain