Yet another report on geographic variation in Medicare spending is given to us by the Kaiser Family Foundation. (KFF Report) The report uses 2013 data and examines trends in spending and spending growth rates by county from 2007 to 2013. Across the nation average per capita Medicare beneficiary spending in 2013 was $9415. In the 20 counties with the highest spending, it was $13,149 and in the 20 with the lowest, $6726. The high ones tended to be in the Northeast, Mid-Atlantic and South-i.e. Florida; the lowest in the West. Differences in hospital spending were by far the greatest contributor to the higher spending in the top counties. If you adjust the numbers for price differences and beneficiary health risk, the amount of difference between the top and bottom 20 counties declines from 96% to 22%. On an adjusted basis, 19 of the highest spending counties are in the south, with 14 in Texas and Louisiana alone. These high-spending counties have fewer physicians per capita and more post-acute care providers.
The average annual unadjusted growth rate in spending has been 2.2% since 2007, but the range reflects an actual decrease of .9% in the 20 lowest counties and an increase of 4.6% in the 20 highest. 15 of the counties with the lowest growth rate are in the south and declines in hospital, home health and durable medical equipment spending seem to be driving the spending slowdown. None of the counties with the lowest growth rates were in the lowest per capita spending group and none of the counties with the highest growth rates were in the top 20 counties in spending, indicating a regression to the mean at work. There were no real provider supply or demographic or health status differences that seem to be related to differences in growth rates. Counties that had high or low spending in 2007 tended, however, to keep that rank in 2013, even though their spending wasn’t growing unusually rapidly or slowly. Overall, the amount of variation in spending growth, adjusted or unadjusted, is declining.
After lots and lots of research, it is starting to appear to geographical variation research doesn’t actually tell us much that is useful and we would be better off focusing on differences in practice style at the individual physician level.