Why are There Large Differences in Medicare Spending Across States?

Research by the Dartmouth Atlas team finds large differences in Medicare spending on patients with similar health conditions across the country. States such as Florida, for example, spends much more per beneficiary than Minnesota. Dartmouth researchers tend to attribute these differences to differences in the practice patterns of physicians. Some doctors practice medicine conservatively; other’s do not.

A new study, however, seems to rebut this idea — finding that variations in medical conditions are the main cause of regional variations in health care costs, once one controls for the health status of patients (The Dartmouth Atlas is based on spending averages).

Another finding of the paper is that the elderly move around a lot.  In fact, one in five Medicare beneficiaries receive medical treatment in more than one census division. These beneficiaries could either be snowbirds or the medical care they require is not available near where they live.

(HT to Jason Shafrin)

Comments (6)

Trackback URL | Comments RSS Feed

  1. Erin O says:

    Interesting…it seems the more we find out, the more questions we have.

  2. Devon Herrick says:

    I suspect the variation has more to do with patient characteristics than provider characteristics. I read an article that explained how in retirement communities in Florida, when meeting friends the conversation often leads to discussions about which specialists you see. Retirees have plenty of time so seeing the doctor (where you might run into acquaintances) is somewhat of a social activity in Florida retirement communities. By contrast, in Minnesota this may not be the norm.

  3. Linda Gorman says:

    There are a lot more questions about the Dartmouth Atlas work than this. These include questions about whether looking at those who died in a two year period gives a clear picture, or spending, whether Medicare spending is a proxy for total spending, and whether its service area roll up makes sense.

  4. Seamus Muldoon MD says:


    A new study, however, seems to rebut this idea — finding that variations in medical practice are the main cause of regional variations…

    @Study abstract:

    Principal Findings. Among high-cost beneficiaries, health was the predominant predictor of costs…

    John, did you mis-type, or am I reading this wrong?? (added emphasis is mine)

  5. Seamus Muldoon MD says:

    @Linda Gorman-

    I agree. I have always maintained that the Dartmouth Atlas two year window with death as the endpoint is not necessarily representative of other aspects of health spending either within or outside of Medicare. Here in Grand Junction, with a preponderance of FP’s and fewer specialists, the local Medical Society poobahs, (mostly FP’s) tout the low spending from the Dartmouth Atlas rankings as a selling point for an Accountable Care Organization model. In fact though, many patients travel to Denver for their higher cost specialty care or procedures, artificially lowering the GJ score. Also, when the only measured outcome is death, I do not think the Dartmouth Atlas Data says anything noteworthy about ‘quality’, although it is often cited as if it is linked to quality.

    Jim Schroeder, MD (aka Seamus Muldoon MD)

  6. John Goodman says:

    Thanks for catching the error, Seamus. I was traveling yesterday and was unable to stay on top of things.

    Error has been corrected.