In yesterday’s New York Times, David Leonhardt has an interesting column exploring potential reasons why the United States spends so much more per capita on health care than other nations, without getting better results. This follows an earlier column in which Leonhardt argued that the last half-century’s increase in U.S. medical spending has been overwhelmingly worth it.
In his earlier column, Leonhardt noted that per capita health costs had increased $5,500 in today’s dollars over this time period, but that life expectancy had gone up by a decade as well. “Would you prefer spending an extra $5,500 on health care every year -– or losing 10 years off your lifespan?” he asked. (Presumably everyone other than Jack Benny would choose the former option.)
But in today’s column, Leonhardt notes that life expectancy is greater in countries such as Greece and Canada — which spend far less than we do on health care -– than it is in the U.S. (Some argue that life expectancy is not a good indicator of health care quality, but that’s a topic for another post.) Leonhardt considers several possible reasons for this; he concludes that the most important consideration is that when it comes to medical care, “Americans seem to be less willing to take no for an answer,” even when the care in question is wasteful.
Leonhardt uses a quote from Jonathan Skinner, a health care researcher at Dartmouth College, to sum up the problem: “Basically, anything that doesn’t kill patients is paid for by Medicare and insurance companies.” In February, Skinner was the lead author on a Health Affairs paper [2 week free access] that looked at the rapid decline in mortality from heart attacks from 1986 to 2002. As David Cutler and Mark McClellan had in earlier work, Skinner and his Dartmouth colleagues Douglas Staiger and Elliot Fisher found that on an aggregate basis, the mortality gains more than justified the cost increases over the period.
But the really interesting stuff was going on below the aggregate level. Notably, regions with the largest spending increases were not the areas with the largest spending gains. It turned out that in the Dartmouth team’s view, the factors really driving the mortality gains were low-cost interventions such as aspirin and beta-blockers, not at all the same factors driving the spending increases. Some regions adopted these low-cost, cost-effective interventions more quickly than others, so in any given year, high-spending regions tended to have worse outcomes. Eventually, the cost-effective interventions were adopted widely (although not widely enough) and brought widespread mortality gains, resulting in the favorable aggregate cost-benefit trade-off found by Cutler and McClellan.
This sort of uneven diffusion of cost-effective interventions is a key factor driving costs and outcomes for many conditions, the Dartmouth researchers say, although they are quick to acknowledge that sometimes expensive, high-tech interventions will, in fact, turn out to be cost-effective. They say that the key to getting good outcomes is encouraging the use of the right treatments, not overall spending levels.
As Leonhardt says, there’s some cause for optimism on this front: Under McClellan and his predecessors, for instance, Medicare has pioneered in finding ways to identify and reward high-quality care. But we still have a long way to go, and Leonhardt warns that as long as we’re not restraining health care costs by insisting on cost-effective care, we’ll end up restraining costs in some other way -– like having 47 million uninsured Americans.