Read our COVID-19 research and news.

A team in the University of Utah Health eCPR Program puts a patient on the life support machinery. A consortium at the university is gathering data on how people who have cardiac arrests outside of the hospital fare after being hooked up to it.

Scott Youngquist & Joseph Tonna

Cardiac arrest kills most victims outside the hospital. Could an artificial heart-lung machine help?

Real-life medical theater doesn't get much more dramatic: A 66-year-old tourist collapsed in the Louvre in Paris, in front of a painting by Eugène Delacroix. Firefighters stationed at the museum administered CPR. Then, a French "mobile intensive care unit"—delayed by rush hour traffic and arriving 19 minutes later—swooped in. Surrounded by majestic artwork, the medical team hooked up the patient to a high-tech life support system: Liters of his blood were routed outside his body, infused with oxygen, and pumped back in.

The man, treated several years ago, died within 24 hours of arriving at a hospital. But the strategy to try to save him, called extracorporeal cardiopulmonary resuscitation (ECPR), is sparking excitement in medical circles—and some anxiety. The machinery for ECPR is already widely used to support patients in heart surgery and sometimes to rescue those who suffer cardiac arrest in the hospital. It also treats infants and children teetering near death from heart or lung failure. (In pediatrics, it's known as extracorporeal membrane oxygenation, or ECMO.)

Now, efforts are spreading to apply ECPR to adults who, like the tourist in Paris, suffer cardiac arrest outside the hospital. Although Paris has deployed ECPR in museums and on subway platforms, most cities limit it to emergency rooms or catheterization labs.

"We all really want it to work," says Clifton Callaway, an emergency medicine physician at the University of Pittsburgh in Pennsylvania. But he and others are awaiting a verdict from clinical trials now underway. One worry is that people rescued by ECPR may have a poor quality of life afterward, for example, because of cognitive impairment. Another question is whether the technology, which requires extensive training, an overhaul of paramedic practices, and tens or hundreds of thousands of dollars per patient to implement, is worth whatever benefit it confers.

Still, doctors crave a better treatment for cardiac arrest, which can result from blocked arteries, drug overdoses, hypothermia, and other causes. Only about 10% of the 350,000 or so adults in the United States whose hearts stop outside a hospital each year survive. The chest compressions of CPR offer, "at best, 25% of normal" blood flow, says Steven Brooks, an emergency medicine physician at Queen's University in Kingston, Canada. With ECPR that soars to 100%.

Japan was the first to publish case studies of ECPR in out-of-hospital cardiac arrests, beginning in the 1980s. In 2015, an Australian team wrote in Resuscitation that 14 of 24 patients survived without neurological problems after having a cardiac arrest inside or outside of the hospital and receiving treatment with ECPR and other interventions. The University of Minnesota (UM) in Minneapolis has treated more patients with ECPR than anywhere else in the United States, more than 200 in the past 3 years. The team published late last year in Resuscitation that of 100 people, 40 survived and fared well. Some had minimal deficits, such as mild short-term memory loss, but were expected to improve with time, UM cardiologist and critical care physician Jason Bartos says. "You have the sickest patients … [and] an opportunity to provide a big benefit," he adds.

Just how big that opportunity is remains an open question. Doctors like Bartos and Callaway—whose hospital uses ECPR on about five people brought in with cardiac arrest each year—know that medicine is littered with tales of phenomenal treatments that falter during randomized trials. One worry about the observational research, such as the reports from Japan and Minnesota, is that the patients receiving ECPR may have already had a better chance of survival than most, skewing the results.

With that in mind, investigators have launched randomized trials to compare ECPR to standard CPR. The largest, in the Czech Republic, is slated to report results next year. "If you don't do a trial early on, then it will be implemented very widely, and it's very difficult to step back and re-evaluate whether it actually is useful," says Marcel van de Poll, a critical care physician at Maastricht University Medical Center in the Netherlands. His is one of six Dutch hospitals that enrolled the first of a planned 110 people in May 2017. Like many trials in emergency medicine, it doesn't obtain consent at the time of enrollment because the patient is unconscious and time is too precious to seek consent from family members. A third trial, in Minneapolis, expects to launch this summer.

The ethics of ECPR weigh heavily and are one reason doctors consider trials so important. ECPR is a desperate measure and an invasive one, including what one doctor describes as "garden hose–size catheters" inserted into the groin. "We have no idea whether these patients would like to be under such a type of care," says Jan Bělohlávek, a cardiologist at Charles University in Prague who is leading the Czech trial. He enrolled the first patient in 2013 and the 185th last week. The majority of those on ECPR still die, Bělohlávek points out, and it is "a very bad dying": Unlike a cardiac arrest followed by sudden death, with ECPR, patients can endure slow-motion organ failure, coupled with anguish among loved ones watching the decline.

Whereas guidelines govern when to stop CPR, with ECPR, "all of a sudden you're supporting heart and lung function, you can support [that] indefinitely, which may create conundrums," says Brian Grunau, an emergency medicine physician at St. Paul's Hospital in Vancouver, Canada. In January, he and his colleagues described in Circulation treating a young person with ECPR who sustained severe brain damage, whose family initially resisted removing life support.

One of the biggest fears has been that ECPR might rescue patients only to leave them in dismal shape. Centers performing ECPR say this happens only rarely. But it does happen. In Minneapolis, six patients of 100 were left with severe brain damage; all died of infections within a few months. In Prague, four patients who initially survived died within 6 months of heart failure, sepsis, or pneumonia.

Whether to implement ECPR across whole cities and regions won't be an easy decision, many believe. Even if it works well, it may be best for a small subset of cardiac arrest patients: people who are relatively young, have few other health issues, and experienced a fixable problem that caused their heart to stop, such as a blocked artery. And yet, "That's why it's so enticing," Brooks says. "The people who it could help are those who are in the prime of their life."