Extracorporeal membrane oxygenation (ECMO) is a temporary technique for providing life support for cardiac dysfunction, pulmonary dysfunction, or both.
The two forms of ECMO, veno-arterial (VA) and veno-venous (VV), are used to support cardiopulmonary and pulmonary dysfunction, respectively. Historically, ECMO was predominantly used in the neonatal and pediatric populations, as early adult studies failed to improve outcomes. ECMO has become far more common in the adult population because of positive results in published case series and clinical trials during the 2009 influenza A(H1N1) pandemic in 2009 to 2010. Advances in technology that make the technique much easier to implement likely fueled the renewed interest. Although exact criteria for ECMO are not available, patients who are good candidates are generally considered to be relatively young and suffering from acute illness that is believed to be reversible or organ dysfunction that is otherwise treatable. With the increase in the use in the adult population, a number of different codes have been generated to better identify the method of support with distinctly different relative value units assigned to each code from a very simple prior coding scheme. To effectively be reimbursed for use of the technique, it is imperative that the clinician understands the new coding scheme and works with payers to determine what is incorporated into each specific code.
Extracorporeal membrane oxygenation (ECMO) is a temporary technique for providing life support for cardiac dysfunction, pulmonary dysfunction, or both until the native organ(s) recover or other definitive therapy is implemented. Although the technique has existed since the 1970s, early failures in randomized clinical trials in adults resulted in a limited number of centers supporting the technology in patients over the age of 18 years.1 Over the past 5 years, there has been a dramatic increase in the number of ECMO cases in the adult population motivated by improved technologies and demonstration of improved outcomes in select populations.2‐4 This article reviews the current indications for ECMO therapy in the adult population and the evidence supporting its use. Furthermore, 2015 updates to the coding and reimbursement for the therapy are discussed.
For decades, severe respiratory and cardiac failure have been associated with heroic efforts, high costs, and frequent mortality. For the most severe cases of ARDS and acute heart failure, mortality can approach or exceed 50%, respectively.5‐8 Although mortality rarely occurs because of the primary insult, patients will frequently develop shock and multisystem organ dysfunction and ultimately die of withdrawal of support in what is perceived to be a futile state. The thought of providing temporary extracorporeal support to allow stabilization of other end organs and allow definitive treatment of the primary organ has been available at a few specialized centers for decades.9‐12 However, with the development of many new technologies, the use of extracorporeal support has continued to increase.2 What was once believed to be a rare salvage therapy is now rapidly becoming a commonly considered, potentially effective option in high-mortality situations.13,14
ECMO is, in essence, cardiopulmonary bypass that has been optimized for weeks rather than hours of operation. A typical circuit (Fig 1) has a venous inflow that draws blood from the patient’s venous circulation into a pump, pushes that blood through an artificial lung (oxygenator), and returns the oxygenated blood to the patient’s venous circulation (veno-venous [VV] ECMO) for circulation to the lungs or the patient’s arterial circulation (veno-arterial [VA] ECMO) for cardiopulmonary support. In the adult patient, VV ECMO support is achieved through cannulae in either the internal jugular and femoral vein or using newer dual-lumen internal jugular cannulae (Fig 2). For VA support, there are a multitude of cannulation options. Typically, percutaneous adult support is achieved via the femoral artery and vein. Central cannulation is used after failure to wean from cardiopulmonary bypass or when sufficient flow cannot be obtained from peripheral cannulation.
Over the past decade, multiple technological innovations, including centrifugal pumps and polymethylpentene oxygenators, have made it easier and safer to implement the technology. There are also a variety of pump-based technologies designed to temporarily support left-sided heart failure exclusively that are useful, but their discussion is beyond the scope of this article.15
The use of adult ECMO has continued to expand for both cardiac and respiratory failure. There has been explosive growth in the use of adult ECMO for respiratory failure likely due to the results of studies and experience from the 2009 influenza A(H1N1) (A[H1N1]) pandemic. According to the Extracorporeal Life Support Organization (ELSO) registry reports from 2004 and 2012, the number of adult ECMO cases has dramatically expanded from around 150 reported runs in 2003 to just under 1,000 reported runs in 2011.2,9 From recent reports, it appears there has been continued acceleration of ECMO use in the adult population.
The original randomized trial of ECMO for respiratory support was published in 1979 by Zapol et al.1The trial, conducted at nine centers, involved patients with extremely severe ARDS randomized to either contemporary ventilator management or VA ECMO. Mortality in both groups exceeded 90%. Supporters of ECMO therapy are quick to identify fundamental flaws of the design, including implementation at inexperienced centers, the use of VA instead of VV ECMO for respiratory failure, the time until initiation of therapy (> 9 days) allowing continued lung injury, and a very high level of anticoagulation and subsequent bleeding.
In the neonatal population, ECMO therapy has been considered a standard of care because of the results of several randomized controlled trials.16‐19 However, adult data remained lacking. Since the Zapol trial, there have been a variety of case series and nonrandomized trials published on the adult population but nothing with the certainty of the neonatal literature. The frequently cited series by Hemmila et al20 from the University of Michigan detailed the experienced of treating 255 patients with Pao2/Fio2 ratios < 100 and expected mortality of > 80% at the time. The series reported a survival of 52%. There were also multiple case reports and several series detailing successful treatment of patients with specific conditions, including pulmonary embolism and pulmonary contusion.21‐26
After decades of disagreement on the actual efficacy of ECMO, the results of the Conventional Ventilation or ECMO for Severe Adult Respiratory Failure (CESAR) trial were published in 2009.4 In this randomized trial, patients were treated using conventional mechanical ventilation at one of many large hospitals in the United Kingdom or were transported to Glenfield Hospital in Leicester, England. Death or disability in the group transferred was 37% at 6 months vs 53% in the control group (P = .03). There was considerable controversy surrounding what on the surface appears to be a definitive result. Concerns surrounding the lack of a standardized ventilator and management protocol for patients in the control group and the fact the analyses were intention to treat and one-fourth of patients randomized to ECMO did not receive the therapy because of improvement are most commonly cited. Many consider the CESAR trial not to be a trial of ECMO but a trial of transfer of patients to a high-volume ARDS referral center that has ECMO capability.
Around the same time CESAR was published, the initial ECMO experience from the A(H1N1) pandemic was released. The first major manuscript was from The Australia and New Zealand Extracorporeal Membrane Oxygenation (ANZ ECMO) Influenza Investigators.27 Here, the authors compared 68 patients who received ECMO for confirmed or suspected A(H1N1) vs 133 who did not. Survival to discharge in the conventional therapy group was 87%. The patients who received ECMO were younger and had fewer comorbidities but were remarkably more ill, with 57% requiring vasopressors at ICU admission compared with 34% in the conventional therapy group and an average Pao2/Fio2 ratio of 56 on 18 cm H2O of partial end-expiratory pressure. Overall survival for this group was 75%.
Similar experiences for confirmed or suspected A(H1N1) to the ANZ ECMO group were reported by investigators in Great Britain.3 Using sophisticated matching techniques, they demonstrated a survival benefit for 80 patients referred for ECMO vs 195 patients not referred for ECMO. In contrast, the REVA Research Network in France reported their experience as well for 123 patients with influenza A (probable H1N1) receiving ECMO.28 After matching, they demonstrated no benefit to ECMO vs conventional treatment, but their matching algorithm left 51 patients unmatched who were younger with lower Pao2/Fio2 ratios with a high rate of survival.
Although the actual benefit of ECMO therapy for respiratory failure continues to be debated in the adult population, one item that is universally agreed upon is the desire to have meaningful recovery. Criteria tend to vary slightly from center to center, but in general many now use the criteria for enrollment into the CESAR trial to be the base criteria for initiation of ECMO therapy (Table 1).4 The guiding principle is that patients not have too much time, typically between 5 and 10 days, on the ventilator prior to cannulation minimizing additional ventilator-induced lung injury and that their disease state be potentially reversible. As such, patients of advanced age, with significant comorbidities and/or other terminal illness are usually excluded from consideration for therapy.
There are generally three indications for adult VA ECMO therapy: (1) inability to wean from cardiopulmonary bypass; (2) hypoxic, biventricular, or right-sided cardiac failure; and (3) extracorporeal cardiopulmonary resuscitation. Of note, the use of ECMO for isolated left-sided heart failure is an option, but other forms of temporary mechanical circulatory support are potentially better options that require less anticoagulation and/or supervision.15 Despite the increasing number of options, the number of adult VA ECMO cases continues to increase. The reasons for this are probably multifactorial, including acquisition of equipment and ease of use at facilities that previously did not have the technology.
Data supporting the use of ECMO for VA support are sparse. Although there exist some data to suggest that certain patients will expire without extracorporeal support, such as the intraaortic balloon pump score, predictors of survival after VA ECMO in the adult population are lacking.29Overall survival for adult VA support from the 2014 ELSO registry is 40% and from extracorporeal cardiopulmonary resuscitation, 29%.
Although the number of ECMO centers has continued to expand over the past 5 years, there are data to suggest that high-volume centers have superior outcomes. Data presented at the 2014 ELSO conference suggest that centers performing more than 30 cases/y were consistently associated with better survival. This is likely because of the sophisticated nature of the technology requiring numerous team members to execute highly specialized tasks. Protocolized management of patients is probably of benefit in addition to having emergency procedures in place. Some high-volume centers have transport programs to receive patients who have been placed on support with the plan for referral. Survival in this situation is around 60%.12
The existing current procedural terminology (CPT) codes and relative value units have been heavily revised for 2015 and reflect the increased use of ECMO in the adult population. Prior CPT codes 36822 (cannulation for ECMO), 33960 (first day of management for prolonged ECMO), and 33961 (subsequent day management) have been replaced with a complex set of codes that much better define the care provided. The new codes for adult management and cannulation are shown in Table 2.30
The most important changes in the codes include differentiation in the form of support being provided, VV or VA. Initiation and daily management of VV support is recorded with codes 33946 and 33948, respectively. VA support initiation and daily management are identified with codes 33947 and 33949. Additional changes in the codes help to better identify cannulation techniques and the age of the patient. Codes 33951 to 33956 help identify percutaneous cannulation, open, and central techniques in addition to the age of patients. For the adult population, codes 33952 (percutaneous cannulation for ages ≥ 6 years), 33954 (open cannulation for ages ≥ 6 years), and 33956 (central cannulation for ages ≥ 6 years) will be most commonly used.
There are also new codes to support repositioning of cannulae (33958, percutaneous; 33962, open; and 33964, central; for ages ≥ 6 years), which are to only be used on days after initial cannulae placement. New codes for decannulation have also been provided (33966, percutaneous; 33984, open; and 33986, central; for ages ≥ 6 years). Finally, there have been new codes submitted for left-sided heart vent placement (33988) and removal (33989) when placed through sternotomy or thoracotomy.
Although the new codes better define service, they also are designed to offer better clarity on what is being covered. Prior codes, depending on the state, may have incorporated critical care time as part of the management of ECMO. It is the intention of the new codes that all additional services will be billed separately. Each code has been assigned a zero-day global period. Critical care time or other appropriate evaluation and management services should be billed separately provided there is no time overlap and should be accompanied by the -25 modifier. During the first 24 h of support, daily management and repositioning of the cannulae should not be charged, as they are considered part of the initiation and cannulation codes.
Individual payers may determine when ECMO is an appropriate therapy. Centers for Medicare & Medicaid Services provide coverage for ECMO in the adult population; however, certain payers may decide ECMO therapy is only supported where there is clinical evidence it improves outcomes. Other payers may consider ECMO to be experimental and not eligible for reimbursement. Hence, one should attempt to clarify a payer’s ECMO policy and adhere to their coverage guidelines. With the issuance of the new CPT codes, it is important for payers to update their policies with the new codes. Many policies by payers are limited in scope and/or have not been updated for considerable periods of time. For example, Blue Cross Blue Shield policy in Montana greatly differs in scope from the policy in Texas, and neither has been updated with new codes.31,32 If a payer has no policy or a policy that has not been updated, a conversation with the payer should be documented with the participants in the conversation, date, and time. Each question asked should be documented along with the payer’s representative’s response. This document should be retained on file.
It is preferable to consent patients or their surrogate(s) for ECMO prior to initiation of therapy unless emergent circumstances prohibit such interaction. Discussion should include a realistic assessment of ultimate outcome. Patients’ decision-makers should be aware of complications, including bleeding, infection, machine failure, and stroke (both hemorrhagic and embolic). The consent should also list conventional therapy as an option that was declined.
With the new codes, cannulation and initiation of therapy are best documented in separate notes and may be billed by different providers. On a daily basis, a separate ECMO note should be created documenting the management of ECMO. Other critical care assessments and plans should be documented in a separate critical care note.
There has been continued growth of ECMO for both respiratory and cardiopulmonary support in the adult population. Although evidence continues to expand supporting the efficacy of ECMO, it continues to be seen by some as a salvage therapy. It is important for providers to understand the management and have institutional support to provide therapy or consider transfer of patients receiving ECMO to centers prepared to provide such care. Prior to initiation of therapy, a realistic discussion of outcome and risks should be documented. The new ECMO codes provide greater clarity in the services provided, and clear documentation and coding is required to receive appropriate reimbursement for therapy.
Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.
2009 influenza A(H1N1)
Conventional Ventilation or ECMO for Severe Adult Respiratory Failure
current procedural terminology
extracorporeal membrane oxygenation
Extracorporeal Life Support Organization