The Fake 'Mid-Death Crisis'

If readers of The Fiscal Times had somehow missed the last three or four decades of arguments that "end of life care" was going to bankrupt the nation, then David Ewing Duncan’s five-part series on "Life at All Costs" provided them with a good introduction to that line of commentary. He reports all the standard claims about how people are receiving extremely expensive care that they do not even want, and that this is a major cause of U.S. health care costs. He also, in reporting objections to these arguments, does provide some contrary evidence.

Yet this series, like much of the commentary on which it reports, combines mistaken facts and inferences with a set of opinions that masquerade as ethical imperatives. Claiming to be experts on "ethics," commentators like Dan Callahan have long made a series of claims that reduce to the following propositions:

1) A large part of health care dollars are spent on care during the last few months or year of life;
2) This care is in some sense "wasted" because, after all, it didn’t work;
3) It could be avoided if physicians properly informed patients that they were going to die, and if patients were willing to listen;
4) Much of this care is painful and dehumanizing, and people don’t really want it;
5) We know they don’t want it because, when they’re not near death, some people say they wouldn’t want it;
6) Therefore we could save a great deal of money and treat people with more dignity if we didn’t force them to have so much care when they are about to die anyway. Thus (abracadabra!) there is a way to save money and be morally superior at the same time;
7) Americans have to learn to accept limits, in spite of their irrational desire for miracles. Other countries save money by rationing, and we will have to bite that bullet. They have a solution that we should copy.

This argument begins from a truth: lots of money is spent on people who are near the end of life. But even that is a deceptive truth, and much of the rest of the argument is either factually wrong or pure opinion.

(1) & (2). Much of the commentary about excessive costs at the end of life begins from reports, as Mr. Duncan writes, that around a quarter of Medicare dollars are spent on patients in their last year of life. This should be no surprise: a large portion of health care spending goes to people who are very sick, and death is one indicator of severe illness. So what does the fact mean?

Not much. First, as Mr. Duncan points out, the share of Medicare costs spent on patients who die within a year has not risen over time. This means that spending on everything else rises just as quickly as spending on people who die. This means, in turn, that spending on people who die is not a distinctive driver of increased health care costs. The arguments that Mr. Duncan reports as somehow news have been made since the 1970s, and continue to be made in spite of the fact that there is no economic "mid-death crisis" that can be distinguished from the general trend of health care cost increases. As one analysis summarized in 2001, "In the late 1980s and early 1990s analysts questioned whether ‘the high cost of dying’ was a major driver of the growth of health care spending. The ‘no’ answer reached then still appears to be true."

Second, analysts who emphasize the large portion of Medicare costs incurred by people who die ignore the fact that a lot of the most expensive beneficiaries do not die. There are many Medicare beneficiaries who have much the same disease profile as the beneficiaries who die, but live. As the same report put it, "Much of what has been labeled the ‘high cost of dying’ is just the cost of caring for severe illness and functional impairment."

In fact, there has been a gradual decrease in the share of Medicare spending that goes to the most expensive beneficiaries in a given year, and this appears to be associated with somewhat less concentration of costs in the year of death, costs for the very ill being spread out over more years (which suggests treatments are preserving life for a while), and a proliferation of all sorts of services for people who are not quite as sick.

So, yes there is a lot of spending on patients at the end of life. But that makes no special contribution to health care spending, which has grown just as much (actually more) for care that is not at the end of life. A lot of the highest-cost patients do not die, and they look a lot like the ones who do. That brings us to the next flaw in the argument.

(3) Actually, doctors are not very good at figuring out who is going to die, when. The "mid-death crisis" crowd seems to think physicians are omniscient, even if not omnipotent. Although Mr. Duncan’s good reporting provides a couple of stories in which prognoses of quick death were wrong, the reader might think these are just anecdotes that systematic data would contradict. In fact, as one of the most eminent advocates of palliative care has written, "on the day before death the median prognosis for patients with heart failure is still a 50% chance to live 6 more months."

A synthesis of the evidence sponsored by the Canadian government in 2004 concluded that "end-of-life prognostication is unreliable, despite a considerable amount of research attention since 1972." So physicians have good reason to hesitate to dismiss all hope, and patients good reason not to be eager to listen.

(4) & (5) Is much of the care that patients receive "dehumanizing," denying them a "good death" that they want, or would want if they were conscious? Perhaps, but these arguments, though worth making in pursuit of a more decent medical care system, can be overstated. Robert Burt has addressed this question in a striking book, Death Is That Man Taking Names: Intersections of American Medicine, Law and Culture. As he argues, some of the advocacy to change care at the end of life even seeks to make dying "not something that happens to you but something you do," and a good death a form of "self-determination." But, face it, death "dissolves personal identity." It is not a form of self-actualization.