Veno-venous Extracorporeal Membrane Oxygenation for COVID-19: A Call For System-Wide Checks to Ensure Equitable Delivery For All

Abstract

Extracorporeal membrane oxygenation (ECMO) has emerged in the COVID-19 pandemic as a potentially beneficial yet scare resource for treating critically ill patients, with variable allocation across the United States. The existing literature has not addressed barriers patients may face in access to ECMO as a result of healthcare inequity. We present a novel patient-centered framework of ECMO access, providing evidence for potential bias and opportunities to mitigate this bias at every stage between a marginalized patient’s initial presentation to treatment with ECMO. While equitable access to ECMO support is a global challenge, this piece focuses primarily on patients in the United States with severe COVID-19-associated ARDS to draw from current literature on VV-ECMO for ARDS and does not address issues that affect ECMO access on a more international scale.

Social, economic, and health inequities disproportionately affect minoritized patients and are key drivers of persistent and widening disparities in the United States and globally.^1,2 The COVID-19 pandemic has thrown these inequities into sharp relief, with minoritized and marginalized communities experiencing a disproportionate burden of disease, inadequate access to high-quality care, and poorer outcomes.³ As evidence of these disparities, the COVID-19-related mortality rate among Black people in the United States is more than three times higher than among White people.⁴

Acute respiratory distress syndrome (ARDS) after infection with SARS-CoV-2 has been the leading cause of morbidity and mortality during the pandemic. Data from randomized, controlled trials suggests that outcomes for patients with severe ARDS are improved by timely referral to centers that specialize in the care of ARDS patients and have the capacity to deliver extracorporeal membrane oxygenation (ECMO).⁵ The COVID-19 pandemic has created surges of patients with severe ARDS, during which ECMO has become a truly scarce resource and, thus, one at risk for inequitable delivery. Recent literature has focused on operational challenges associated with the effective and ethical distribution of VV-ECMO for COVID-19 but has not addressed the potential for bias-related inequities in ECMO.^6,7

For a patient to go from a severe illness in the community, receive treatment at an ECMO center, and achieve a functional return to that community necessitates navigating a path through several levels of awareness, recognition, diagnosis, treatment, and referral. Each interface has opportunities for both implicit and explicit bias to influence care and decision-making. Similar to the ‘chain of survival’ necessary for a patient to survive after out-of-hospital cardiac arrest, one must complete a ‘circuit’ of interactions between communities, patients, providers, institutions, and ultimately back to the community upon discharge to successfully manage a patient with severe COVID-19-associated ARDS using ECMO (Figure 1). Therefore, we define and describe the stages of this circuit, highlight potential sources of bias, and offer ways to ‘check the circuit’ for opportunities to mitigate bias. By doing so, ECMO programs and systems can examine their practices and implement strategies to ensure more equitable delivery of ECMO.

Figure 1.: Stages of the circuity of equity framework for patients receiving veno-venous extracorporeal membrane oxygenation for COVID-19 associated ARDS.

Circuit of Equity

Individual- and Community-level Risks for Contracting SARS-Cov-2 and Developing COVID-19

SARS-CoV-2 is a highly transmissible virus that requires physical distancing, isolation of infected individuals, and masking to mitigate exposure. The ability to distance and isolate, however, is not equitably distributed. Crowding within minoritized households and workplaces is associated with an increased incidence of contracting COVID-19.⁸ Once contracted, preexisting conditions such as hypertension, obesity, and diabetes mellitus significantly increase the risk of developing severe COVID-19 pneumonia.⁹ These conditions are more prevalent in low-income and minoritized populations, due to differences in insurance status, access to high-quality preventative and primary care services, and other inequities.¹⁰ A meta-analysis of 72 reports found the incidence of death due to COVID-19 to be 15% higher in Hispanic populations and 105% higher in Black populations compared to White populations.¹¹ Finally, although vaccination is the most effective intervention for the prevention of severe COVID-19, disparities in vaccine access and vaccine hesitancy persist among low-income and minoritized communities due to widespread misinformation and preexisting mistrust of health systems.¹²

The barriers to improving equity at the community level are enormous and complex. The role of an ECMO center at this stage is likely limited to supporting regional advocacy efforts and developing anti-racism policies and health promotion initiatives. Nonetheless, as we describe the other stages and connections between them, it is important to remember that we are describing a circuit rather than a one-way path. Patients come from and potentially return to communities affected by factors that must be considered when developing health services. ECMO centers should support post-intensive care clinic as well as rehabilitation and other services that promote return to work and community reintegration.

Primary and Secondary Sites of Diagnosis and Treatment

The next stage of the circuit is the site(s) of treatment for patients with COVID-19 severe enough to require hospitalization (e.g., community and regional hospitals without ECMO capabilities). Compared to nonminoritized or higher-income patients, underrepresented minorities are more likely to live in areas of the country with limited access to specialized facilities and, as a result, are more likely to have their initial point of care be a smaller institution with limited capacity to support patients with severe respiratory failure.¹³

As in the previous stage of the circuit, an ECMO center is somewhat limited in its ability to mitigate bias at the primary or secondary sites of treatment. However, there are opportunities for ECMO centers to reduce bias at this stage through targeted educational outreach. This could, for example, include periodic educational conferences or symposia where providers at these institutions share best practices and provide continuing medical education, joint quality improvement initiatives, and structured case reviews. Examples of such regional and national initiatives include the Pacific Northwest Sepsis Conference, Seattle Area Chest Grand Rounds, Project ECHO (Extension for Community Healthcare Outcomes), Western New England Acute Cardiology Conference, and Veteran’s Affairs Specialty Care Access Network (SCAN) ECHO.

Referral to an ECMO Center

Patients who develop severe ARDS with refractory hypoxemia, severely reduced compliance or marked ventilatory insufficiency might benefit from transfer to an ECMO center.¹⁴ It is critical for the treating provider to recognize when these indications are present and place the initial referral to an ECMO center. An ECMO center, for our framework, is defined as any hospital with the ability to cannulate patients for VV-ECMO and continue managing the circuit and the patient’s underlying disease.

To the extent patients from underrepresented backgrounds present to hospitals with limited critical care resources, they are subject to the bias of being cared for by providers who may have less familiarity with the indications for referral to an ECMO center.^15,16 Disseminating updated criteria and protocols, providing feedback to referring facilities, and maintaining a regional ECMO capacity dashboard are strategies to support timely referral to a center with a current capacity to care for patients who ultimately require ECMO. When ECMO centers reach full capacity, as has been common during the pandemic, a centralized system for managing capacity and referrals would alleviate the burden many providers face in having to place multiple calls to multiple ECMO centers before an accepting center is identified.

This is important as phone calls between referring facility and ECMO center providers are high-risk points for bias.¹⁷ Regardless of whether the patient is accepted for transfer to an ECMO center, subsequent management may be strongly influenced by the presentation of their case by a provider at the referring facility or the questions posed by providers at the ECMO center. Implicit biases on the part of both the referring and ECMO providers may lead to bias-prone communication.^18,19 Opportunities to mitigate bias at this stage include blinding information such as race, ethnicity, employment, payor, and location of origin. For example, disclosing patients’ payor status has been shown to bias transfer decisions in other referral systems.^20,21 ECMO centers should use a standardized system and form to collate necessary information while reducing the potential for bias in decision-making (Table 1).

Table 1. - Example of Referral Call Template

Introductions:
•Role in caring for patient
•Role in decision for transfer
Dates of:
•Presentation to hospital
•Diagnosis of COVID-19
•Initiation of non-invasive ventilation or high-flow nasal cannula
•Intubation
Current Clinical Data:
•Vitals: heart rate, blood pressure, oxygen saturation, temperature
•Vasopressor support: Yes/No; if yes, dose & agent
•Ventilator Parameters: mode, positive end-expiratory pressure, fraction of inspired oxygen, tidal volume, inspiratory time, respiratory rate, etc.
• Blood gas oximetry (with concurrent ventilator parameters)
•Barotrauma: Yes/No
Medical Therapies, Yes/No:
• Prone positioning (If yes, length of time)
•Neuromuscular blockade
•Inhaled pulmonary vasodilators
Pause for synthesis & specific clarifying questions (e.g. what type of vasodilator)
Patient Demographics:
•Age
•Sex assigned at birth
•Height
•Weight
Other Active Conditions, Yes/No:
• Other organ dysfunction (renal, liver, coagulation, cardiac, etc.)
•Sepsis
•Other active life-limiting condition
Patient Chronic Medical Conditions, Yes/No:
•Hypertension
•Cardiovascular disease
•Diabetes
•Immunosuppression
•Chronic lung disease
•Other life-limiting condition
Pause for synthesis, clarifying questions & discussion
Initial Impression:
•Current capacity, ICU and ECMO
•Recommendation for transfer, yes/no/defer for group conference
•Current management suggestions
•Plan for safe retrieval

Acceptance to an ECMO Center

Each ECMO center has its acceptance criteria based on the best available evidence, system logistics, and institutional capacity. Commonly used clinical criteria include objective information such as age, P_aO₂:F_IO₂ ratio, ventilation parameters, time on mechanical ventilation, and the presence or absence of other organ failures. Subjective criteria may also be applied such as judgment about the reversibility of the patient’s underlying condition and likelihood of achieving a good outcome with a finite period of ECMO support. These subjective criteria may produce bias as available data demonstrate a relationship between race, implicit bias, and physician’s assessment of illness severity.^22,23 The effect of this bias may be challenging to mitigate as both over- and underestimation of the likelihood of a poor outcome could affect the decision to accept a patient for transport owing to the narrow window of opportunity for benefit from ECMO support.

ECMO centers can address this by having a stakeholder group discussion (i.e., physicians, nurses, ECMO operators or perfusionists, respiratory therapists, and cultural mediators) about the decision to accept the patient once the intake call and information are collated. The ability to hold such multidisciplinary discussions may be limited by the urgency of the referral or the time of day at which it occurs. This problem can be mitigated by establishing a large group message (e.g., Microsoft teams group) or video conferencing that can be activated at all times as a way to achieve a quorum on short notice. VV-ECMO cannulation in ARDS is often a process coordinated over hours and the benefit of a group discussion can be worth the time needed to gather this discussion. During such group discussions, a blinded template that eliminates phrases or information such as insurance status that might introduce bias should be used to guide the presentation of information to the participating members. Antibias training should be required of any staff who are involved in the care and transfer processes.²⁴

The availability of critical care transport is a limiting factor in the transfer of a patient to the ECMO center. As described above, minoritized patients are more likely to live in communities with higher rates of COVID-19, which can challenge EMS capacity in the area.²⁵ Early assessment of transport capacity should be incorporated into regional dashboards and standardized referral processes.

Decision to Cannulate for ECMO

While many patients accepted for transfer are cannulated before transport or immediately after arrival at the accepting institution, there is a subset of those transferred to an ECMO center who are never cannulated, as they may either improve with advanced ARDS management strategies or deteriorate and are no longer suitable candidates. The decision to forego cannulation after arrival to follow an individual’s clinical course is another point at which bias can affect decision-making.

In other life-threatening conditions with an option to intervene aggressively such as thrombolysis or percutaneous coronary intervention after myocardial infarction, minoritized patients are less likely to receive aggressive interventions for the same health conditions as non-minoritized patients.^22,26 Minoritized patients with COVID-19 are also more likely to present with more severe disease, which could shorten their time window during which cannulation may be appropriate.²⁷ Resource scarcity during the pandemic may lead providers to consider stricter inclusion criteria for cannulation, exacerbating these disparities.⁶ Institutional cannulation criteria, demographic blinding, and multidisciplinary discussions may be useful to guide cannulation decision-making.

The incorporation of ethics committee representatives and cultural mediators may also be useful, but the time-sensitive nature of cannulation decisions may sometimes preclude the involvement of such groups.

Care During and After ECMO Support

After cannulation, patients are still subject to differential service delivery. Differential use of and referral for mechanical right heart support, renal replacement therapy, transplantation, rehabilitation, or comfort-focused care for minoritized patients have been reported.^6,28–32 Such inequities in the allocation of intensive therapies take on heightened significance in times of resource scarcity. A recent survey showed that during the pandemic, intensivists were more likely to consider discontinuing ECMO from patients considered to have a poor chance of survival if a better candidate was referred to their center.⁶

Daily critical care and ECMO rounding checklists and automatic consultation of key services can be used to help mitigate differential delivery of care. Hosting scheduled goals of care discussions, with the inclusion of interpreters and cultural mediators when suitable, can ensure that key information is transferred despite differences in language or health literacy. Early involvement of social work and palliative care for all patients and cultural mediators when suitable can improve communication, patient- and family-centered goal setting, and advocacy more broadly.^33,34

Ideally, patients who recover can return to their community. In this best-case scenario, the patient lives still in a community subject to bias and potentially limited access to care; this is why we conceptualize care of an ECMO patient as a circuit, not a one-way path. Evidence supports the use of interventions such as rehabilitation services, postintensive care or post-COVID clinics, and patient navigation outreach to improve postdischarge care.³⁵ Additionally, referral to vocational rehabilitation and psychosocial support services can facilitate community reintegration.^36,37

Checking the Circuit: Evaluating Equity in ECMO

The circuit of equity involves multiple stages from the onset of infection with SARS-CoV-2 that people face in their communities, differences in service delivery for ARDS, application of ECMO, and, ideally, community reintegration. At each stage of the circuit, patients are vulnerable to biases that may reduce the chances of equitable access to, treatment with, and recovery after ECMO for ARDS caused by COVID-19.

The challenges associated with equitable application of ECMO have been recognized elsewhere,^38,39 and we acknowledge that the definition of equity in healthcare is not universally agreed upon.⁴⁰ However, it should be clear that differences in service delivery should not be stratified by sex, race, ethnicity, religion, political alignments, views on vaccination or insurance status. Furthermore, although we describe the circuit in the most current context of the COVID-19 pandemic, this framework may apply to any condition which drives ECMO scarcity in a community. We anticipate the potential for bias in ECMO allocation to only grow as clinical indications expand and new diseases emerge. We acknowledge too that our framework is U.S.-centric, and that other sources of bias resulting from complex political, social and cultural realities may affect access to ECMO in other areas of the world. Time and space considerations prevent an adequate discussion of these important and complicated issues.

The circuit and mitigation measures described above are not meant as a prescription for use at an ECMO center alone and are, instead, intended to increase awareness of many points at which patients are vulnerable to bias and the opportunities to mitigate these risks. By addressing these challenges, ECMO centers can improve health outcomes and generate more confidence in the system by populations who continue to be harmed by explicit and implicit bias.⁴¹

Further research is warranted to explore the manners in which health disparities impact the use of ECMO and test interventions designed to promote equity (Table 1, Supplemental Digital Content, https://links.lww.com/ASAIO/A868). We recommend starting locally before embarking on system-wide or regional initiatives. ECMO centers can, for example, perform periodic queries of transfer center and EHR data to review internal cases of ECMO for COVID-19-associated ARDS, including referrals, acceptances, declines, cannulations, services received, and outcomes. Second, ECMO centers can query the region’s ECMO system, including working to identify differences in sociodemographic information, care, and parameters for those who were and were not referred for advanced COVID-19-associated ARDS care. It is by investigating these issues in detail that improvements can be made in the system of ECMO allocation to ensure equitable distribution of this limited resource for all.