This 2017 Health Affairs article has a fairly common cocktail party factoid as part of its frame that it is writing against:
end-of-life medical spending is often viewed as a major component of aggregate medical expenditure, accurate measures of this type of medical spending are scarce. We used detailed health care data for the period 2009–11 from Denmark, England, France, Germany, Japan, the Netherlands, Taiwan, the United States, and the Canadian province of Quebec to measure the composition and magnitude of medical spending in the three years before death. In all nine countries, medical spending at the end of life was high relative to spending at other ages. Spending during the last twelve months of life made up a modest share of aggregate spending, ranging from 8.5 percent in the United States to 11.2 percent in Taiwan…
One of my good friends and colleagues is the lead author on a palliative care/end of life evaluation methods paper that we’re submitting this week. We’re writing about a palliative care program that enrolled participants who were very ill and we had challenges in performing a reasonable evaluation. One of the many things that we talked about over the past couple of years with this program is the problematic framing of palliative care as a cost saver in the last X days of life. There can be palliative treatments that are good, that are valuable and that don’t save money. We have no problem paying for attempts to use curative treatments that are good, effective and not cost-saving. We don’t apply that same threshold to palliative payment policy.
Let’s go on a quick little thought experiment about why the factoid that we spend a lot in the last X days of life is problematic. We are picking on the dependent variable and only seeing deaths instead of extended lives.
Let us pretend that there are 10 people. They currently have a disease with a 99.9% fatality rate within the next six months. We can predict that these folks will have $20,000 in medical claims between now and death on average. A year from now we will say that the last six months of life that group had an average medical expense of $20,000 and think about how to compress spending variation.
Now let’s spin out a counter-factual. There is some awesome innovation ( a drug, a surgery, a miracle worker etc ) that costs $200,000 per patient and has well defined outcomes. 9 of the patients who receive the treatment will be cured and live another 20.5 years at the same level of health and capacity as the patients had at age 25. 1 patient will die a week after receiving the treatment. We don’t know who will fall into each group before the treatment.
From the perspective of end of life costs, this new innovation spikes costs tremendously. The dead people are now costing $200,000 in the last portion of life instead of $20,000. From the point of view of population health, this is amazing news. An average of 18 years of life per person had been bought at $11,100 per year of life. Under any cost effectiveness threshold, that is dirt cheap and high value. However since we were picking on the dependent variable of death we miss this huge gain in value.
So when you hear the cocktail party factoid of “we spend too much near death” be aware of the assumptions that aren’t being stated.