The Washington Post Column

How We Die: Choice and Chance

No matter how you ask the question, most people — right-wing, left-wing, atheist, religious — will tell you that they don’t want to die like Terri Schiavo. That is, they don’t want to spend their final days in a hospital, tied up to a machine, unable to feed themselves, unable to speak. Nine out of 10 Americans have told Gallup pollsters that they don’t want to die in an institution. Another poll found that 82 percent of Americans would, upon being told they had very little time to live, prefer simply to go home: They want to die in their own beds, surrounded by family, in a setting that feels natural.

The vast majority, of course, will not die at home. Perhaps they won’t be on a feeding tube, but they will almost certainly be in a hospital or a nursing home. Perhaps their relatives won’t be battling one another in the national media, but they are statistically more likely to be at odds with one another, or with their doctors, than ever before. Many will be surrounded by strangers and hooked up to machines, even if they’ve written living wills that say they don’t want to be. Many will die, in other words, in a setting that doesn’t feel natural at all.

Along with all of its unexpected political implications, the Schiavo case has had the effect of exposing the enormous gap between what Americans imagine death should be like and what death actually is like for most people in the 21st century. A hundred years ago, when average life expectancy was 47, people who got sick either recovered or died quickly. Now that life expectancy is 75, most Americans will spend at least two years of their lives too disabled, one way or another, to care for themselves without help.

Yet, although we see video images of death all the time — movie shootouts, scenes of faraway warfare — we don’t much like dwelling on the medicalized environments in which most people in our society actually pass away, and we don’t like thinking about the murky ethical dilemmas that their deaths often present. In some sense the Schiavo case has attracted so much attention precisely because it brings, almost for the first time, a very common, very painful, but usually very private dilemma into the public sphere. This case is unusual only in that the family has such a relatively clear set of choices, or at least knows with some degree of certainty what will happen when the tube is removed. More often, the decisions are extraordinarily delicate medical and ethical judgment calls. When is someone’s health so frail that the risks of an operation — or the price of an operation — become too great? Who decides when the pain that will result from a particular cancer treatment is going to outweigh the benefits?

But because we don’t dwell on it, and because we haven’t thought about it, the system that has sprung up to care for the elderly and the terminally ill is neither medically nor ethically consistent. In different regions of the country — even among hospitals rated the best in the country — there are huge variations in the kind of treatment given to dying patients. In one study — carried out over several years, in five different hospitals in five different regions — nurses were trained to conduct constant conversations with the ill and dying, trained to respect their wishes, and trained to allow those who wanted to die at home to do so. Even so, there was a wide range of outcomes. In one region, a third of the patients died in a hospital despite their wishes. In another, the figure was 66 percent.

These variations couldn’t be explained by medical necessity or patient preferences. On the contrary, it seems that the most important factor in determining whether a particular region has a high rate of medical intervention on behalf of the terminally ill — risky operations, respirators, artificial feeding — is not local religious practices but the local availability of hospital beds and the number of local doctors. There are, for example, a lot of both in Miami. As a result, Medicare spending on a patient in the last six months of life was twice as high in Miami as in Minneapolis. Dying patients in Miami spent four times as many days in intensive care units and saw at least twice as many specialists, too. Yet another study, conducted by the Dartmouth medical school, found no correlation between the amount of treatment given to dying patients in particular regions and that region’s overall mortality rate: Just because you see more specialists doesn’t mean you’ll live longer.

In this sense, the Schiavo case is not only unusual but actually misleading. All of the commentary makes it sound as though these momentous decisions are not only crystal clear but are ultimately made on moral grounds, as if there were something important at stake: the sanctity of human life vs. the right to die, or the wishes of the dying person vs. the wishes of the family. But in practice, we’ve designed a health care system in which the fulfillment of one’s wishes on this matter depends on serendipity. You will die at home — or not — because of where you live. You will be kept on a respirator — or not — because that’s how your local hospital does it. And until we and our politicians are able to focus more thoughtfully on the realities of 21st century death, serendipity, not ethical debate, will remain the ruling principle.