An Insightful Perspective on End-of-Life Decision Processes


As the son of a mother who, mercifully died suddenly as a result of an automobile accident when she was in the early stages of Alzheimer’s disease, and a father who deteriorated over an 15-month period, all of which was spent in a rehabilitation center and a nursing home after he broke his hip at age 82, I have thought a lot about end-of-life issues.

As a result, I was gratified to see an incredibly incisive and thoughtful article on this subject by Dr. Atul Guwande of Harvard Medical School in the August 2, 2010, issue of The New Yorker . The subject of end-of-life care for individuals with terminal illnesses or diseases is not a new one, but Guwande brings new insight to it.

For example, he points out that many people have the mistaken belief that more intensive hospital-based care generally prolongs life, whereas hospice care shortens it. Hence, many people make the decision to employ all possible life-prolonging measures for themselves or their loved ones, believing that, in so doing, they are buying time for something else to work on their behalf. Yet Guwande states:

“Like many people, I had believed that hospice care hastens death, because patients forego hospital treatments and are allowed high-dose narcotics to combat pain. But studies suggest otherwise. In one, researchers followed 4,493 Medicare patients with either terminal cancer or congestive heart failure. They found no difference in survival time between hospice and non-hospice patients with breast cancer, prostate cancer, and colon cancer. Curiously, hospice care seemed to extend life for some patients; those with pancreatic cancer gained six weeks, those with lung cancer gained six weeks, and those with congestive heart failure gained three months. The lesson seems almost Zen: you live longer only when you stop trying to live longer.”

The second point Guwande makes is that there is often a trade-off between extending life and being mentally alert: those who receive hospice care are often able to manage their affairs without losing mental alertness, and are able to make plans for themselves and others with competent professional assistance. Those who receive hospital care are often in situations in which they are removed from loved ones, lapse into unconsciousness, and experience completely debilitating pain and discomfort.

We tend to think of end-of-life issues mostly in terms of the elderly, but Guwande’s story focused on the terminal illness of a woman who was delivering her first child. What none of us can know is whether, by opting for more aggressive treatments, she and her husband missed opportunities to discuss longer-term questions about the dying mother’s preferences as to how her daughter would be raised. I lost a wonderful cousin to breast cancer 22 years ago when she was 41 years old. She left behind a husband and a 4-year-old daughter when she died. However, she had ample time and alertness to have many discussions with her husband that enabled him to gain the value of her insights on raising a daughter to adulthood as a single parent. The value of those conversations was incalculably large, but, in many instances, aggressive hospital care makes these conversations almost impossible to have.

The question Guwande’s observation begs is: why, if the more aggressive treatments shorten life, reduce the quality of life, and reduce the ability of patients to spend valuable time with loved ones, would patients choose more aggressive treatments? He gives two answers:

  • Every prognosis for a terminal disease contains a traditional bell-curve distribution with an average life expectancy for the patient, but with wide variations from a few weeks to years or even decades. Although the size of the curve for those with the disease living a very long life indicates that few people survive for a long time, many people convince themselves that they will be the exception, rather than the rule. When they opt for aggressive, expensive treatment, they are effectively buying a lottery ticket for the biggest jackpot of all, a long life, but often with the odds associated with buying a lottery ticket for a $200 million lottery prize.
  • Physicians, either out of desire to keep up hope for patients, or out of the desire to keep patients from going elsewhere for treatment, or because they simply do not know the appropriate life expectancy, routinely overestimate survival times, often by over 500%.

Guwande pointed out that people seem willing to explore hospice options, as long as more aggressive treatment options are not foreclosed to them. He cites a Johns Hopkins pilot program that reduced hospital care and costs for patients who opted for hospice care, but knew that they could receive hospital care at any time. He recommended that Medicare and other health plans eliminate the irrevocable either-or decision for terminally ill patients.

He also has an interesting set of recommendations for changing the decision criteria and process for patients opting for end-of-life decisions. Clearly, getting individuals to discuss their options with physicians and loved ones in advance is preferable to getting decisions made when a person is already terminally ill. That recommendation is not new.

However, the insight he brings to it is that the dialogue is as much about giving reassurance to the loved ones as it is to getting a particular decision from the patient. Even loved ones who want to respect the wishes of the patient will feel guilty using less than the most aggressive treatments. The discussion that, in advance, addresses their deepest concerns of guilt is extremely productive. Similarly, a discussion between the patient and the physician, which helps the physician understand when being overly optimistic or recommending the most aggressive treatments is not helpful to the patient, is also desirable.

How do policymakers avoid getting caught up in the “death panel” trap? The simplest answer is to broaden the dialogue beyond end of life treatment discussions. There are many circumstances in which individuals are incapacitated and unable to register their preferences when a situation is not life threatening, but decisions have to be made quickly. For example, the physician often encounters something unexpected during a surgical procedure when the patient is under a general anesthetic.

There are also many situations in which all treatment options are imperfect, and in which there is no ability to get a better answer even after all medical risks and probabilities are assessed. This is the current situation with prostate cancer treatment options. The discussion about the patient’s broader values and preferences is probably one that cannot be handled fully at one time or in one circumstance. It would be most helpful for health plans and Medicare to cover an annual discussion that is free flowing and that simply enables a physician to get to know the patients far better than he or she can in the context of individual office visits designed to diagnose, treat, or decide on a treatment option.

Although many people legitimately criticized those who raised the “death panel” argument, the “death panel” advocates’ concerns could have been addressed, and the right kind of dialogue could have taken place if it had been framed as a way of increasing patient empowerment, as opposed to a piece of a larger program to increase government control over the health care system.

Nevertheless, the biggest conceptual breakthrough toward which Guwande takes us is that more aggressive care is not better care, and, in many end-of-life situations, it may actually shorten life and worsen the quality of the life that is led immediately before death. Moreover, because of the inability of patients to engage with loved ones on transitional issues, it may have negative long term consequences.