fb-pixel Skip to main content

Lifesaving implants complicate end-of-life care

What to do with these devices in terminally ill patients is becoming a growing issue

Professional societies have issued calls for physicians to address with patients the issue of when to deactivate a defibrillator or pacemaker.

More than a decade has passed, but Nathan Goldstein, then a medical resident at Mount Sinai Hospital in New York City, still remembers it clearly.

A man with terminal lung cancer had planned to die at home with hospice care. Instead, the man was lying on a stretcher in the busy emergency room. Every few minutes, his heart received a shock from his internal defibrillator, preventing his heart from stopping.

The implantable cardioverter defibrillator, or ICD, can be a lifesaving device when its charge of electricity jolts the heart out of a potentially lethal rhythm. But when someone is dying from an irreversible disease like cancer, the defribillator’s shocks to a failing heart might only serve to postpone death.


Ultimately, with the family’s consent, Goldstein and his colleagues deactivated the patient’s defibrillator. The shocks stopped. And the patient died, not at home the way he and his family had planned, but in the hospital’s ER.

With more than 100,000 defibrillators and at least twice as many pacemakers implanted in aging patients in the United States each year, what to do with these devices in terminally ill patients is becoming a growing issue. Professional societies have issued calls for physicians to address with patients the issue of when to deactivate a defibrillator or pacemaker — implanted devices that can regularize a too-fast or too-slow heartbeat. However, a recent study found that nearly one-third of patients with defibrillators were still receiving shocks in the last 24 hours of their lives. Additionally, half of patients with a do-not-resuscitate order still had their defibrillators turned on in the final hour before death.

“I love technology,” said Paul Mueller, an internist and ethicist at Mayo Clinic. “When a new technology becomes available, there’s almost an imperative to use it. But it comes at a consequence.”


Mueller and his team recently reviewed the records of more than 100 patients whose devices were deactivated before they died in recent years, and found that only one had included a request for deactivation in an advance directive, often leaving family members of patients too sick to make the decision to struggle with the issue.

“Our goal is not to be critical of the technology — we embrace the technology — but we now need to grapple with the end-of-life concerns that will inevitably arise in these patients,” Mueller said.

That unwanted defibrillator shocks occur at the end of life represents an unintended consequence of the technology’s success. Researchers developed the device — now about the size of a small cellphone and surgically implanted just under the skin in the chest — in the 1980s. The ICD works by continuously monitoring the heart’s rhythm and quickly delivering a shock of electricity if it senses a dangerous rhythm.

At first, the device was only placed in those who had already experienced a potentially lethal heart rhythm. But by the early 2000s, researchers found that defibrillators could also save the lives of certain patients with poor heart function, whose weakened hearts place them at high risk for the sorts of rhythms that lead to sudden death. The pool of potential candidates for defibrillators now included many older, sicker patients.

It was around that time that Goldstein, who is now an associate professor of geriatrics and palliative medicine at Mount Sinai Medical Center, watched the man dying of lung cancer receive shocks from his defibrillator. Goldstein wondered why this man — clearly aware of his impending death — hadn’t had the device deactivated.


Goldstein teamed up with cardiology colleagues to find out whether his patient’s experience was isolated. In theory, deactivating an ICD shouldn’t be hard. The device doesn’t need to be removed. Instead, a cardiologist or a representative from the defibrillator company electronically reprograms it not to deliver shocks. While painless, deactivation does take planning.

To determine whether or not deactivation was happening, the research team interviewed next of kin for 100 patients with ICDs who had died of any cause. Their study results, published in 2004, revealed that up to 20 percent of patients had received a defibrillator shock in the last days to minutes of their lives. Doctors had discussed deactivating devices in a minority of cases and when they did, these conversations often took place just a few days prior to the patient’s death.

These findings drove Goldstein to study the barriers that prevent these discussions. His team interviewed general cardiologists, electrophysiologists — subspecialized cardiologists who are trained to manage irregular heart rhythms — and internists. The doctors described systemwide issues such as the lack of long-term relationships with patients and sufficient time in the office, and discomfort discussing end-of-life.

“When you start talking about . . . turning it off, then you are sort of shutting off the hope,” one physician said. Additionally, doctors pointed to barriers unique to the ICD, specifically that a small, internal device doesn’t provide a clear physical reminder to prompt a conversation.


“The defibrillator discussion is really hard. You have to get a feel for what a patient’s level of understanding is and where they’re at, because if you’re not nuanced about it you can really mess people up,” said John Mandrola, a Kentucky electrophysiologist and columnist for theHeart.org, who has seen older and sicker patients referred to him for ICD placement in recent years. “We are doing better in terms of how we’re implanting these devices. But we’re not doing better in terms of getting people to understand what to expect from them.”

In the absence of information, patients hesitate. In a separate study of patients with ICDs, participants were not aware that deactivation was an option. When made aware, they did not want to discuss deactivating their ICDs. One patient in the study described deactivation as “like an act of suicide.” When it comes to turning off a pacemaker, which could lead to death even more quickly for someone whose heart is dependent on the pacemaker’s function to beat, some physicians balk, too.

This was the landscape that faced Carol Filak when her father, Joseph Hoffman, then 81, was diagnosed with advanced cancer in 2008 and told he likely wouldn’t live longer than six months. Hoffman, a long-term smoker, had suffered poor health for decades, and he’d had a defibrillator implanted years before.

Hoffman took the news of his cancer “in stride,” his daughter recalled. He declined treatment, opting to prioritize the quality of his remaining time. When Filak asked her father about his defibrillator, his answer was simple: “Turn it off.”


Filak asked her father’s cardiologist about turning off the defibrillator. But the doctors told her to wait and see. “Wait for what?” Filak thought. She followed the advice and months later, growing sicker, Hoffman entered hospice with his defibrillator still active.

Filak remembers growing increasingly frantic; she’d tried to plan ahead, but here her father was, in the precise situation she’d dreaded. Each doctor she called seemed to think her request should be someone else’s responsibility.

Finally, a representative from the defibrillator company was dispatched to the hospice to turn the machine off. Hoffman died quietly the next day.

“He died on his own terms,” Filak said. “He went peacefully, but the process angered me. If I hadn’t brought this up, it would have never been turned off. There were no plans for it at all.”

Now that researchers have realized that turning off the devices is a problem, they’re working on how to fix it.

Julie Shea, a cardiology nurse practitioner at Brigham and Women’s Hospital, has witnessed a shift over the past two decades. Now, she says, she sees between six and 12 requests to turn off cardiac devices each year. Ideally, it would happen in a clinic appointment, after a careful conversation between doctor and patient — although what more commonly occurs is that the request comes when a patient is at home, receiving hospice care to prioritize comfort in the final days rather than prolonging life. In this case, Shea works with a representative from the device company to turn the device off in the patient’s home.

Shea says that the need to occasionally replace the battery, which requires an oupatient procedure, offers a “natural stopping point” to have a conversation with aging patients about whether having a cardiac device aligns with their goals. After all, Shea said, patients “aren’t dying from arrhythmias any more, but from heart failure or cancer. We’ve exchanged the reasons these patients are dying, and so we’ve had to become more sensitive to how these patients are dying and what role their device plays.”

For Goldstein, the challenges that continue to plague cardiac device deactivation — coordination of care between specialists and generalists, discomfort surrounding when to initiate conversations about serious illness and who should have these conversations — are a microcosm of the general problems that plague end-of-life care.

Goldstein has secured funding from the National Heart, Lung, and Blood Institute for a study across multiple hospitals to improve the frequency and quality of conversations about ICD deactivation. Cardiologists enrolled in the study will attend a training session on how best to communicate with patients. These doctors will receive reminders when a patient coming to the clinic might benefit from advance planning. Researchers will monitor how many conversations occur, the number of ICDs deactivated, and whether this intervention improves mental health outcomes for caregivers.

“Thinking about someone shocked by an ICD at the end of their life is quite graphic. It makes you stop and say, ‘that’s horrible.’ But it’s not just about the ICD,” Goldstein said. “That might be what grabs our attention. But this is about how we improve communication and change clinician behavior in general. That is what we’re struggling with.”

Dr. Daniela J. Lamas can be reached at danielalamasmd@gmail.com.

Correction: An earlier version of this story contained a graphic that incorrectly identified the left and right sides of the heart.