Another Reason to Dismiss 'Death Panels'

End-of-life care isn't costing as much as we thought.

Maybe not in his last year.

Photographer: Thomas Samson/AFP/Getty Images

I am often asked whether we could constrain health-care costs by trimming only end-of-life care. After all, the argument goes, so much money is spent on treating people in their last year of life.

The problem with this perspective -- leaving aside the complicated ethical issues involved -- has always been the difficulty of predicting which year will be a person's last. If you don’t know, what exactly do you change in practice?

But now research has revealed something else: The share of health-care spending that goes to end-of-life care is not as large as we've thought. Using Medicare survey data, a team of researchers led by Mariacristina De Nardi of the Federal Reserve Bank of Chicago examined health-care spending on Medicare beneficiaries. They found, as previous research had, that spending is very high in the months and years prior to death. Average costs in the final calendar year of life amount to more than $40,000 -- twice the average annual spending on the elderly (people over 65) and roughly six times the average for the nation as a whole. About half of the expenditures in that final year are spent on hospital care and about a quarter on nursing home care.

The research also suggests, though, that spending in the final year of life amounts to only 17 percent of total spending on the elderly -- about a quarter less than earlier estimates. Furthermore, since almost three-quarters of deaths occur in that age group, the share for the population as a whole is much lower -- less than 7 percent of total U.S. medical expenses.

The researchers also calculated how concentrated health-care costs are. If you rank people over 65 by their medical expenses, the top 10 percent account for more than half the total. The top 30 percent represent 80 percent of spending. And these patterns are persistent: About half of those in the top quintile of health spending in one year are still there two years later. So focusing on high-cost beneficiaries makes a lot of sense; it’s just that most of them are not in their final year of life.

None of this is to say that we should ignore difficult questions about expenditures on end-of-life care. I have long favored, for example, requiring patients admitted to a hospital to specify their preferences through an advance directive, which is the ultimate form of consumer-driven health care.

The new data, by the way, should put to rest any remaining false charges around so-called death panels -- which were never part of the Affordable Care Act. Even if they had been real, they wouldn’t have been all that appealing to an accountant.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.