Families stall end of life talks, say doctors. True?

Doctors and nurses said patients and their families created the largest obstacles to end-of-life decision making in the ICU, in a large survey published in JAMA Internal Medicine.

About 1,300 staff at 13 academic hospitals in Canada rated barriers to end-of-life goals of care on a 1-7 scale. Doctors and nurses considered the largest barriers to end-of-life decision making to be family members' and patients' reluctance to accept a poor prognosis, along with their poor understanding of the limitations and complications of life-sustaining treatments (rated about 6 on the 7 point scale).

Unsurprisingly, nurses and doctors considered their own communication skills as non-factors in obstructing end-of-life discussions. However, respondents acknowledged a lack of training in these discussions: only 29% of attendings, 35% of residents, and 10% of nurses reported receiving training on facilitating end-of-life decision making.

Surveys of patients and families show that 70% of people say they want to die at home; 80% with chronic disease say they want to avoid aggressive care at the end of life. But the reality is that half of Americans die in the hospital, and 20% die in ICUs, the vast majority of them with acute complications of chronic diseases.

Why the disconnect?

Prognoses Uncertain, Emotions Unexplored

While early discussions of end-of-life preferences should be encouraged, the truth is that even among the tiny minority of people who prepare advance directives, few grasp what they're deciding. Most advance directives only stipulate that people would not want to be kept alive with machines and drugs if there is no hope for recovery. This isn't helpful in the dynamic, probabilistic world of the ICU.

What if you're critically ill on ventilator day 6 with a 60% likelihood of survival, but with a 90% chance of moderate functional impairment and a 50% chance of severe impairment? What if there's an 80% chance of survival, but a 95% chance of frequent re-hospitalizations over the next 6-12 months, stays in nursing homes and long-term acute care hospitals in between, and virtually no chance of getting home, or even taking a walk outside? What if you'll survive 3 more years but in a dependent state, and every dollar you've saved will be spent at custodial care facilities, leaving nothing from your life's work for your spouse or children?

Most families have to make decisions for their loved ones even without the benefit of these predictions, because accurate prognosis is rarely possible in the ICU. That leaves doctors to be either vague (humbly, truthfully, but unhelpfully), or over-confidently optimistic or pessimistic, depending on their own personalities, biases, and willingness to engage emotionally.

Add to the mix family members' feelings of love, regret, guilt, and fear of their own deaths, and asking for or expecting "rational" decision-making seems almost autistically obtuse. It's their Mom, after all. Until we accord respect to the emotional black box at the center of all these discussions -- the unknowability of death; the visceral terror of losing a parent, spouse, child, sibling; each discussant's own fear of dying; the guilt of surviving -- we haven't even begun to talk.

What Should be Done?

Although they're gaining ground, palliative care and end-of-life discussions have always gotten short shrift in the revenue-obsessed U.S. health care system. Much could be done, with immeasurable returns in the currency of our own humanity, but also with likely reductions in ICU and overall health care costs:

• More and better studies on short- and long-term outcomes after critical illness to allow physicians to provide realistic likely outcomes after critical illness more confidently. An excellent concise clinical review in PulmCCM Journal by Tessa Damm and Jayshil Patel provides a survey of what we know today.

• More and better training in end-of-life discussions in internal medicine residencies and critical care fellowships (I received about 15 minutes' worth, as I recall). Less than half of surveyed critical care programs consider their training to be good; 28% felt it was fair or marginal. Training should be based in evidence from outcomes studies, but more importantly should teach communications skills on how to have difficult conversations. Google "end of life discussions training for physicians": given its importance, the lack of resources is a bit embarrassing and sad.

• Promotion to the public of the IHI's The Conversation Project for end-of-life planning. If attention from A-list journalists and columnists continues (see Ellen Goodman's recent op-ed in the New York Times), public awareness will increase and "The Conversation" will get easier for everyone.

• Media campaigns to counteract TV and film portrayals of overly positive outcomes after critical illness and "health care miracle" narratives that drive people's unrealistic beliefs in the restorative powers of medical care during life's inevitable final trajectory of decline.

I know how wide the gap can be between the "rational" perspective of doctors and that of "irrational" patients and their families, because I crossed it in 2014 during my Dad's end-of-life journey after a decade of progressive dementia. Despite an intensivist and a Ph.D. nursing instructor in the family, he still somehow got a PEG tube a month before his death.

I hope that experience made me a better doctor, but I know it made me a more empathetic person, including when I happen to be talking to my patients' families who are in similar trying circumstances.

Families certainly do slow down end-of-life discussions, or derail them entirely. Their delays and decisions may seem inexplicable, but they're not. Most often, we physicians just haven't expended the time or the energy to understand them. It's not only because we're too busy (although that's certainly part of it): we were never taught how. Before we blame families when a patient's care plan takes a seemingly irrational turn, we should first look hard at ourselves, our own fears and failings.

Read more:

John J. You MD et al.  Barriers to Goals of Care Discussions With Seriously Ill Hospitalized Patients and Their Families: A Multicenter Survey of Clinicians. JAMA Intern Med. 2015;175(4):549-556.