During hospital rounds, Kenneth Covinsky, MD, MPH, a professor of medicine and clinician researcher in the Division of Geriatrics at the University of California San Francisco, sometimes meets families at a wrenching crossroads: a loved one with advanced dementia is no longer eating, slowly losing weight because they no longer are hungry.

They ask the question almost every time — can’t something be done?

Technically, yes. A feeding tube can deliver nutrition. But the treatment can also bring pain, restraints to keep the tube in place, and a heightened risk for aspiration pneumonia. What seems like a simple fix is rarely simple at all. Instead, Covinsky recommends feeding patients food they love.

“A lot of times in geriatrics, they’re going to do as well if I treat them light as if I treat them heavy, but if I treat them heavy, I’m going to cause a lot of discomfort and pain, and I might even make their outcomes worse,” said Covinsky.

Clinicians routinely must decide to provide life-prolonging treatments unlikely to improve outcomes or quality of life. In US hospitals, where aggressive care for terminally ill patients is often the default, those choices become more complicated.

“The vast majority of clinicians feel they are required by ethics or by law or by policy to provide therapies that they worry are not beneficial,” said Jason Batten, MD, a clinical instructor in the Department of Anesthesiology and Perioperative Medicine at the University of California, Los Angeles. 

Mismatch With Reality

Batten recently led a study published in JAMA Network Open of over 100 physicians, nurses, and other clinicians to find out how they make such decisions. Clinicians worked in emergency departments, hospitals, ICUs, and palliative care settings.

Professional societies have issued guidance giving clinicians two options for withholding or withdrawing interventions they believe will not help and may cause harm: reach a shared decision with the patient and family, or use an institutional process when family wants the care.

But the study found many clinicians are not doing either of these — they will not mention a potential intervention, like dialysis, or tell families they will not offer the treatment they think will not benefit the patient.

“There’s a lot of potential harm, emotionally, mentally, that we can actually inflict on a family if we’re asking them to make decisions that really aren’t a choice. The patient is going to pass away,” said Elizabeth Dzeng, MD, PhD, MPH, an associate professor in residence in the Division of Hospital Medicine at the University of California San Francisco, who helped lead the study. “Why are we asking them about treatments that just are completely not going to work?”

One clinician commented that they would not offer cardiopulmonary resuscitation in a specific case because doing so would just prolong inevitable death.

“I think a lot of physicians experience a lot of moral distress around this, feeling like they aren’t able to act ethically according to what they believe is right,” Dzeng said.

The dissonance between policy and reality can lead to variations in end-of-life care “simply based on which hospital a patient happens to be admitted to and which physician happens to be on service at that time,” Batten said.

“By ensuring professional guidance explicitly addresses the ethical issues surrounding alternate approaches that physicians routinely encounter in end-of-life care, physicians can be more supported and empowered to practice in ways that are consistent with guidelines and also, importantly, consistent from physician to physician and hospital to hospital,” he said.

Batten noted that these alternative approaches already occur in other specialties, such as when a surgeon decides a patient is not a candidate for a procedure.

Aligning

Holly Prigerson, PhD, director of the Cornell Center for Research on End-of-Life Care, Cornell University in New York City, said she and her colleagues have found that family members almost uniformly believe their loved ones are going to improve because of aggressive care at the end of life.

While clinicians want to “save lives and give people hope,” Prigerson said a persistent “moral quandary” occurs when patient expectations become a “mismatch for the reality of what these treatments can offer.”

This misalignment often translates into high-intensity interventions during a patient’s final days, which is reflected in data that show up to 25% of Medicare payments go to beneficiaries in their last 12 months of life, with the highest spending in acute care hospitals.

Ciera Sears, MD, a geriatrician and internist at Cleveland Clinic in Avon, Ohio, said she usually focuses on learning about the patient to align her recommendations for medical care with their hopes and desires. If a patient no longer has brain function, she encourages families to think about what the patient might have wanted.

Sometimes clinicians must also make decisions about care settings that may not improve a patient’s condition. Covinsky often finds himself helping navigate families through the choice of transferring an older patient from the hospital floor to an ICU.

While doing so might improve a patient’s chances of living longer, an ICU stay also increases the risk of developing delirium. He advises that if treatments are not working on the hospital floor, they will likely not work in the ICU, and patients should begin receiving hospice care.

Prigerson and her colleagues developed a communication technique for oncologists to help patients understand their prognosis and make sound decisions on treatment. Rather than describing the size of a tumor, clinicians are instructed to tell a patient that their cancer is growing, and that treatment is not stopping its spread. Clinicians can also inform patients of the benefits and harms of various treatments.

Prigerson said this approach is intended to correct the bias of both oncologists and patients toward overtreatment.

Not all clinicians view these discordant expectations as a source of distress, however. Robert Arnold, MD, vice chair of professional development for the Brookdale Department of Geriatrics and Palliative Medicine at the Mount Sinai School of Medicine in New York City, said he is comfortable when patients and families choose not to follow his recommendations.

“I think we should just be more humble about what constitutes the right decision,” he said. “Are there cases in which families ask for things that I’m pretty sure won’t achieve their goals? The answer is ‘yes, sometimes.’ And yet, it’s the patient who’s living the consequence, and family who’s living the consequence.”

Arnold said clinicians need more training in how to have conversations about nonbeneficial care.

“I think most of the time, if physicians are straightforward and empathic and clear, patients and families make the best decisions they can, and they’re decisions that I think in the end, are good decisions,” he said.

Batten reported receiving a research training grant funded by the National Institutes of Health. Dzeng reported receiving grant funding from the California Department of Public Health Alzheimer’s Disease Research Award, the National Institute on Aging, and the National Palliative Care Research Center. Covinsky reported that his research program is funded by the National Institute on Aging. Arnold reported being chair in the board of VitalTalk. Sears and Prigerson reported having no financial conflicts of interest.

Brenda Sandburg is a freelance journalist for Medscape Medical News. She has written about the biopharmaceutical industry and legal issues for the Pink Sheet and American Lawyer Media.