We spend about 25-30% of Medicare spending on patients who are in their last year of life, according to Gerald F. Riley and James D. Lubitz in their 2010 study, "Long-term trends in Medicare payments in the last year of life" (Health Services Research, April 2010, 45(2):565-76). They also find that this number hasn't changed much over the last 30 years--that is, health care spending during the last year of life is rising at about the same pace as other Medicare spending-- and that the percentage isn't much affected by adjusting for changes in age or gender of the elderly. On their estimates, Medicare spending on those who die in a given year is much higher than on those who survive the year: in 2006, Medicare spending in 2006 on those who died in that year was $38,975, while Medicare spending in 2006 on those who survived the year was $5,993.
Total Medicare spending in 2012 was about $560 billion. Thus, 25% of that amount would be $140 billion spent during the last year of life. It's often unclear at the time whether someone is actually in their last year of life, but say for the sake of argument that such cases could be identified, and spending in this area could be reduced by half. If attainable, cuts of this size would be $70 billion in annual savings, which is certainly a substantial sum. But to keep it in perspective, total U.S. health care spending is in the neighborhood of $2.6 trillion. Thus, the potential gains from even fairly aggressive limits on end-of-life health care spending through Medicare is a little under 3% of total U.S. healthcare spending.
To what extent is the U.S. health care system changing its practices in end-of-life care? In the February 6, 2013 issue of JAMA, a team of writers led by Joan Temo address this question in an article called: "Change in End-of-Life Care for Medicare Beneficiaries" (vol. 309, #5, pp. 470-477). They find an intriguingly mixed set of patterns.
One common measure of end-of-life care is to see what share of patients died in a hospice or at home, compared to dying in the acute-care ward of a a hospital. Their results show that from 2000 to 2009, the share of patients who died in the acute care section of a hospital declined from 32.6% to 24.6%; the share of patients who died at home rose from 30-7% to 33.5%; and the share of patients who died in a hospice rose dramatically from 21.6% to 42.2%. On the surface, these kinds of numbers certainly suggest a pattern of less aggressive end-of-life care.
But when Temo et al. dug just a bit deeper, they found that many of the hospice stays were extremely short--just a few days. Looking at the use of intensive care units in the last month of life, they found that it has risen from 24.3% in 2000 to 29.2% in 2009. In addition, the number of health care "transitions" from one care setting to another has risen both in the last 90 days of life and the last three days of life. For example, 10.3% of patients had a "transition" in the last three days of life in 2000, while 14.2% of patients had a transition in the last three days of life in 2009.
As Temo et al. put it: "Although a hospice stay of 1 day may be viewed as beneficial by a dying patient and family, an important yet unanswered research question is whether this pattern of care is consistent with patient preferences and improved quality of life. ... Our findings of an increase in the number of short hospice stays following a hospitalization, often involving an ICU stay, suggest that increasing hospice use may not lead to a reduction in resource utilization. Short hospice lengths of stay raise concerns that hospice is an "add-on" to a growing pattern of more utilization of intensive care services at the end of life."
Few questions in health care policy are harder than what should be spent on end-of-life care. It's fairly common for the elderly, when healthy, to say that they don't want extreme end-of-life measures. But when those same people become very ill, both they and their families often start thinking that extreme care makes a lot of sense. In addition, while perhaps the diagnostic and statistical techniques for figuring out a few months or a year in advance who is likely to die will improve over over time, right now they are not very accurate. Thus, the common sense policies in this area tend to revolve around earlier counseling for the elderly, so that patients (and their families) can have a more clear sense of what they want in terms of end of life care, and improving hospice and end-of-life home care--after all, basic palliative services like intravenous fluids and antibiotics don't need to happen in a hospital setting.