Medical Examiner

Letting Go

Gradual withdrawal of care may make end-of-life decisions easier.

Doctors often fall back on a strategy of presenting the objective data and letting patients’ loved ones make the hard calls.

Alexander Raths / iStock

Mr. R was a patient I took care of in the intensive care unit early in my residency. He was clearly dying. A month earlier he had presented with an infection after spinal surgery. Since then he had developed respiratory failure, requiring a ventilator; septic shock; and renal failure, requiring dialysis. He had little chance of surviving out of the ICU despite our best efforts—and the entire team knew it.

Every day we offered his wife, who was his health care proxy, bleak updates. And every day she reiterated her desire to continue all aggressive measures with hope that Mr. R would recover. After the fifth day of his decline, I wondered how our medical team would convince his family that further aggressive measures were futile. 

Such thoughts made me feel guilty. How, as physicians and nurses, could we convince Mrs. R to accept her husband’s death? Wasn’t our role to be advisers, not persuaders? One of my attending physicians felt similarly when I discussed Mr. R with him on morning rounds: “We aren’t supposed to ‘convince’ the family of anything,” he said. “We present facts, answer questions, and let the family decide.”

Mrs. R and the rest of the family continued to refuse to withdraw care. Mr. R’s blood pressure continued to drop. We were nowhere close to taking him off the ventilator. And treating Mr. R meant that we were denying a precious ICU bed and ventilator to someone in the hospital who had a better chance of surviving.

Was it our job to convince the family to let Mr. R die?

Situations like these are some of the most ethically challenging in any hospital. In the ICU, our team sensed when someone might not make it out of the hospital—and most times we were right. In these instances, our team regularly spent rounds discussing how aggressive care might hurt the patient while not improving her chance of survival.

But discussing medical opinions among medical professionals is very different from recommending that a family withdraw life support for a loved one. And so we frequently reverted to an often fruitless strategy of describing the patient’s condition, presenting data—without offering a recommendation—and waiting on the family to eventually decide to withdraw care.

As a recent Institute of Medicine report titled “Dying in America” suggests, such situations are all too common. The report notes that most people nearing the end of life are not “physically, mentally, or cognitively able to make their own decisions about care.” Such individuals—especially the 70 percent who are in a hospital within 90 days of dying—and their families need guidance, not just facts, at the end of life.  

On Mr. R’s 10th day in the ICU, we gathered the family members again and updated them on his progress. As the meeting concluded, Mrs. R whispered, “He wouldn’t want this.” The statement did not lead to any change that day. But this was the first sign that Mrs. R was realizing the futility of her husband’s care.

Our team met with her the next day. Mrs. R reiterated that she didn’t want her husband to die that day. However, she knew that he would not make it out of the ICU. She wanted more time with him without the guilt of giving up too soon.

We proposed removing interventions slowly, allowing the family to spend ample time with him to say goodbye. That day Mrs. R agreed to stop all tube feedings. Soon after, she agreed to remove a key medication that supported his blood pressure. She agreed to withdraw the ventilator the next day, and Mr. R passed away soon after. We felt that we had done right by both the patient and his family.  

Maybe, then, it wasn’t our job to convince Mrs. R of why death was the right option, but rather to devise a solution that respected both the family’s wishes and the patient’s mortality.

Gradual withdrawal of care was that middle ground. There is evidence in favor of removing life support measures over days rather than all at once. In patients with long ICU stays, a study in the American Journal of Respiratory and Critical Care Medicine showed that sequential withdrawal of care increases family satisfaction with death.

We can prevent many difficult decisions at the end of life. The Institute of Medicine recommends that doctors and patients begin advance care planning (creating advance directives, designating health care proxies, etc.) at any age and readdress these decisions regularly. A clear set of preferences might prevent much of the uncertainty around the care of patients like Mr. R.

But difficult end-of-life conversations will still occur among doctors, patients, and families. When they do, health providers should not be afraid to recommend withdrawing care. But they should do so while aligning death with patients’ and families’ goals of care. By doing so, we might learn to present letting go as not just a last resort, but sometimes the best option for patients and family members.