Who decides when to turn off lifesaving devices?

MDs should discuss when to turn off ICDs

Implantable cardioverter defibrillators (ICDs) are lifesaving devices, as demonstrated by a 2004 study that showed ICDs reduced death by 23% in people with moderate heart failure and poor pumping function, compared to patients who did not receive ICDs.1 But what if that lifesaving device outstays its welcome and prolongs death because its users haven’t discussed when their ICDs should be deactivated?

Nathan Goldstein, MD, assistant professor in the Brookdale Department of Geriatrics and Adult Development, Mount Sinai Medical Center, NY, says a study he and several colleagues conducted on patient deaths and ICDs indicates that as many as one-quarter of patients whose families participated in the study had received shocks from the ICD in the last month of life, and some of those received shocks in their final moments of life, when defibrillation shocks cause discomfort, anxiety, and a prolonged death. Defibrillators are surgically placed inside a patient’s chest to provide ongoing heart rhythm monitoring and standby intervention, and they automatically shock the heart during a cardiac arrest or irregular heartbeat.

"It was incredibly distressing to the families," says Goldstein, who helmed the study, "Management of Implantable Cardioverter Defibrillators in End-of-Life Care," published in the December 2004 Annals of Internal Medicine. "The patient may have been essentially unconscious; but when the shock is delivered, [had] grimaced, and this was very distressing to the patient’s family."

Goldstein and his fellow researchers contacted the next of kin of people who had had ICDs implanted at Yale University but who are now deceased.

"We did a retrospective study, and what we wanted to find out were details about the patient’s death; particularly about ICD," says Goldstein. "We called the next of kin, and asked them details about the patient’s death."

The questions included "Did the patient get the shock at the end of life?" and "Was there ever a discussion of turning off the lifesaving equipment?"

Goldstein says his team’s first realization was that patients with implantable defibrillators were not having discussions with their physicians about turning off the defibrillator, and when.

Goldstein says the study found that discussions between physicians and patients about deactivating defibrillators took place in only 27 of 100 cases of terminally ill patients at Yale-New Haven (CT) Hospital. Even among patients with do-not-resuscitate orders, discussions about continued use of the device occurred in fewer than 45% of the cases.

Doctor-patient discussion needs work

What surprised Goldstein, he says, was the number of people he and his colleagues found who were not aware it is possible to deactivate ICDs. Those who did know still were not likely to have talked about it with the family member who had the ICD.

Of the 27 (out of 100) patients who did have discussions about deactivating their ICDs, 21 elected to turn off the devices. But Goldstein and his team found the discussions about deactivating the defibrillators were not conducted well in advance of death, but in the last few days or hours — even minutes — of the patients’ lives, after some had received shocks that family members described as "distressing."

The authors also report that 27 people they talked with stated that their next of kin received a shock within the last month of their lives. Only nine of these patients had subsequent discussions with their clinician about deactivating their ICDs, and six of them elected to turn off their devices.

"I think if you glance at the study, the main takeaway point is that people should know they can turn off their ICDs, but some people never knew it was even an option," Goldstein says.

"Why don’t we talk about them?" Goldstein asks. "One reason is that physicians in general don’t have lot of experience in talking about this."

One resource physicians can call on is a palliative care team, with members who have the expertise to help patients and families weigh the question of quality of life vs. quantity.

Goldstein said the issue is a difficult one, because it involves deactivating a device the patients and physicians may have been relying upon heavily to keep the patient alive for the previous months or years since it was implanted.

"The device is not bad," Goldstein points out. "It is doing what it is supposed to be doing [in delivering shocks to restart or regulate heartbeats]."

What’s the optimal death?

Talking with a family member or other loved one about deactivating an ICD when his or her health has reached a point at which "restarting" the heart may not be in the patient’s best interest is not an isolated discussion, but one that should be part of a larger discussion of quality of life vs. prolonging life, Goldstein says.

When a patient has an ICD implanted, it typically is with the anticipation that he or she has years to live with a good quality of life. But what if, after a while, the patient is diagnosed with a disease that will result in death, with a significantly diminished (by the patient’s standards) quality of life. Is there a point at which arrhythmia may be a better mode of dying than death from other causes?

"What is the optimal death? That can only be answered on an individual basis," Goldstein says. "This is an option — the patient can continue to live out his or her life as long as possible. But another option is maybe that dying suddenly might be a better outcome for some people.

"Would they rather die sooner, or live longer with more suffering or symptoms? That is a question that only each individual can answer."

But to answer it, people with ICDs need to know their options, he points out.

Finding the correct time for a physician to discuss the options of deactivating the device can be difficult. Some experts advocate making it part of patient education at the time the defibrillator is implanted. But others say the devices are so complicated, and the patient has to learn so much about how it works; implantation is not a good time to bring up deactivation.

Some institutions have included ICD deactivation as one of the subjects in its formal consent for care agreement; other patients learn of the options during discussions of do-not-resuscitate orders. However, Goldstein found that fewer than 50% of patients who had do-not-resuscitate orders had talked about deactivating their ICDs.

He says he plans a future study to determine some optimal times for bringing up the issue of turning off ICDs.

Goldstein says his study is not meant to indicate patients should be counseled to deactivate their ICDs should it become clear that death is imminent.

"There were families who said they did have discussions about deactivating the devices, but the patients decided to leave them on, and that was a good outcome because there was discussion and the patients’ wishes were followed," he says.


1. Bardy GH, Lee KL, Mark DB, et al. Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT). Presented at the 53rd annual meeting of the American College of Cardiology. New Orleans; March 2004.


  • Nathan Goldstein, MD, Assistant Professor, Brookdale Department of Geriatrics and Adult Development, Mount Sinai Medical Center, Box 1070, One Gustave L. Levy Place, New York, NY 10029. Phone: (212) 241-4641. E-mail: nathan.goldstein@mssm.edu.