Improving use of ECMO

Extracorporeal membrane oxygenation (ECMO) can be a lifesaver for critically ill children, but it comes with a burden of morbidity and mortality. Two clinicians at Boston Children’s Hospital hope to reduce the risks for patients on ECMO. Since 2006, Ravi Thiagarajan, MD, MPH, chief of the Division of Cardiac Critical Care has conducted research to improve ECMO outcomes. In 2013, he was joined by Peta Alexander, MBBS. Together, they have worked on various initiatives to improve the way clinicians use ECMO to support patients.

Ravi Thiagarajan, MD, MPH, chief, Division of Cardiac Critical Care
Ravi Thiagarajan, MD, MPH, Chief, Division of Cardiac Critical Care
Peta Alexander, MBBS
Peta Alexander, MBBS, Division of Cardiac Critical Care

Improving the anticoagulation process

“Over time, we’ve tried to learn more about outcomes for specific procedures and conditions,” says Thiagarajan. “That led us to look at processes, one of which is anticoagulation. We need to use blood thinners so clots don’t form, but it comes with a risk of bleeding, so we are working to improve that process.” Using Standardized Clinical Assessment and Management Plans (SCAMPs) methodology, they are gaining a better understanding of ECMO and anticoagulation in ECMO. “One issue with ECMO research is that we have only 30-40 patients a year on ECMO,” says Alexander. “So it takes a long time to gather a large amount of data.” But since 2008, with SCAMPs, they have about 200 patients in a database where they have captured every aspect of coagulation and every decision made based on the patient’s clinical condition and results of laboratory testing. Thiagarajan says they have used this data to make changes in the ECMO protocol over four iterations. “We’ve learned there are many things that contribute to bleeding and clotting risks, and that one specific test doesn’t help us understand that,” explains Thiagarajan. “We started by using the activated clotting time (ACT) to manage anticoagulants in the ECMO circuit, and learned that test was not enough.” They switched to a heparin anti-Xa assay, but results showed that even this test alone was inadequate to facilitate decision-making around patient anticoagulation management. “Our second protocol standardized how we approached bleeding in ECMO patients, and the third protocol targeted reduced anti-thrombin use — a medication which potentiates the effect of heparin, but is associated with increased bleeding,” says Thiagarajan. They also found that the anti-Xa level was not the only test that helped them provide optimal anticoagulation in patients. “We also discovered that many ECMO patients had natural levels of anticoagulation and sometimes we didn’t need to use a medication to augment that because it increased their bleeding risk,” he says. The fourth iteration of the protocol incorporates thromboelastography — a whole blood test that provides information on factors that facilitate clotting, such as platelet function, and to use that as an additional level of testing. Within the first three iterations, we’ve learned a lot about the coagulation system and the tests we need to reduce bleeding complications, and we’ve demonstrated that the use of some high-cost interventions, for example anti-thrombin, do not contribute to better patient outcomes.” Thiagarajan says that they may continue to make adjustments to the protocol as they learn more. “The SCAMP protocol has standardized the anticoagulation management of these patients and facilitated detailed clinical data collection,” says Alexander. “We can use this data to polish our protocols and gain a better understanding of the coagulation system and the affects anticoagulation has on the ECMO circuit and the patient.” They will be presenting their work on anticoagulation at the EuroELSO 2019 conference in April. The ECMO SCAMP continues to be supported by the Clinical Pathways Program within Program for Patient Safety and Quality.

Bleeding and transfusion

The team has also done some retrospective research on bleeding and transfusion of patients on ECMO in parallel with the anticoagulation SCAMP. “One thing we learned was that if a patient is bleeding on ECMO you actually have to stop their anticoagulation,” says Alexander. “When this was first proposed locally in 2015, it was a very novel idea. We’ve been able to show that the ECMO circuits will tolerate patients being off anticoagulation. Overall, the patients do better if they’re not being anticoagulated at the same time they are actively bleeding, because no matter how we replace blood products to patients, large blood product transfusions result in an inflammatory process.” Thiagarajan says another aspect of these projects is to use patient and transfusion data to understand more about the benefits and risks of transfusing patients on ECMO. “The amount of hemoglobin in an ECMO circuit is standardized at 35 percent,” Thiagarajan explains. “When we looked at the transfusion process, we learned that most of our transfusions were related to maintaining this arbitrary number in the ECMO circuit rather than managing any clinical concerns for the patient.” Alexander adds, “The belief has been that you need 35 percent hemoglobin to adequately deliver oxygen to end organs and tissue. But oxygen delivery is facilitated by adequate cardiac output. On ECMO, we can optimize cardiac output, that’s part of ECMO support. So theoretically, when you’ve got preserved cardiac output delivered by the ECMO circuit, it should be a time when we can accept a much lower hematocrit, particularly for patients who may have gone on ECMO in very low cardiac output state.” Their next step is to find a way to reduce patients’ exposure to blood transfusions. “Lots of research shows exposure to blood transfusions actually reduces a patient’s chances for survival,” says Thiagarajan. “We need more research, but we think that lowering the 35 percent may work just as well, and be beneficial to patients. We’re excited to learn what that number should be and to reduce blood product exposure.” This is a clinical question that may best be answered by a prospective randomized clinical trial, which we are considering commencing locally.

Work with other hospital programs

Another important area of research and program development for the two has been working with the Medical and Surgical Intensive Care Units (MSICU) at Boston Children’s to report ECMO outcomes for patients. The ECMO Program is overseen by the Department of Anesthesia, with clinical care delivered in the MSICU and Cardiac ICU. In addition to shared ECMO specialists, equipment and protocols, they have combined with researchers in the MSICU to explore two research initiatives supported by the Cardiovascular Critical Care Research (CVCC) Cluster.

Researching long-term ECMO outcomes

The two are also are very interested in looking at long-term outcomes for children who were supported on ECMO. While the most often reported metric is survival to hospital discharge, Thiagarajan and Alexander hope to expand on this timeline. “We want to know how our patients are doing, well after hospital discharge. But gathering data has been difficult, since people come to our center for cardiac care from all over the world, and even getting local patients back for neurodevelopmental follow-up can be challenging,” says Alexander. To better track long-term outcomes, they are piloting a remote application originally developed by the Emergency Department, and then applied to the Cardiac Surgery Program. “We adapted the app for ECMO follow up,” says Alexander. “We ask some simple resource use questions — if they’ve been back in the hospital or had more procedures — and we also ask about their functional status and quality of life, such as how they’re feeling, if they’re back at school and do activities of daily living.” They rolled out a two-year pilot program in August 2018. Parents are contacted with a link to the app upon enrollment, and are asked a few questions about what the child was like before hospitalization. Then they are asked to report quality of life after ECMO at time of discharge and at one month, six months, one year and two years. “That will allow us to see what life actually looks like for them, in addition to what we know about their in-patient course,” says Alexander. “We’re really excited to get information that’s meaningful to us without putting families out of their way.”

Professional organization participation

ECMO, even in large centers like Boston Children’s, is still a small volume event. As such, it can take many years to accumulate enough patient care experience to make informed inferences for aspects of care to change for better outcomes. International researchers have collaborated to combine patient experiences that can then be reviewed for quality and research initiatives. “Part of that work comes from our involvement in the Extracorporeal Life Support Organization (ELSO) registry, an international data registry, in which both of us have had leadership roles,” says Thiagarajan. He says they, and other researchers, use data from that international registry to learn what they can do better and how to make ECMO safer and more efficient.

A multi-disciplinary team

Important clinical research on ECMO use in critically ill patients benefits from insights from bedside nursing and physician teams, as well as ECMO specialists who manage ECMO circuits by the patient’s bedside and collate clinical features, laboratory results and medication changes. Both Alexander and Thiagarajan acknowledge that local ECMO research could not continue without their input and support. In addition, mentoring interested local residents and trainees in cardiology and critical care medicine has been an enduring element of the ECMO research program. Many prior projects were supported by the Rosalie Rose CICU Research Fund, and moving forward, they anticipate undertaking a prospective pilot clinical trial with the support of the Callahan Family Chair endowment.