Who's Responsible for the Rise in Medicare Costs

Today’s Managing Health Care Costs Indicator is $300 billion

Click image to enlarge
Today, the New England Journal e-published an evaluation of which states and which specialties have “overspent” in Medicare, leading to the SGR (sustainable growth rate) cuts of 27.4% as of January unless Congress acts to overturn this. Congress is having trouble – because eliminating the SGR and freezing physician payments would cost $300 billion over the next ten years.   Ali Alhassani et al have shown huge discrepancies.  This article is a good demonstration of how the threat of across-the-board cuts is unlikely to drive diverse providers to diminish their utilization.  

Excess expenditures in Alaska from 2003-2009, for instance, are 130% of total 2002 expenditures, while Maine has increased its spending at such a low rate that they are responsible for ‘saving’ almost half of 2002 expenditures over the same time period.   I’m guessing that former Senator Ted Stevens’ securing a 35% permanent rate increase for Alaska providers probably plays some role for overspending in the frozen northwest.

The authors also show that large states like Texas, Florida and New York drive much more of the total Medicare costs – but of course these states will also see a larger portion of the total across-the-board cuts. 

The authors also show that radiation oncology has driven very large increases in Medicare expenditure, while thoracic surgery costs have substantially lagged the SGR target.   Of course, there have been huge improvements in radiation therapy over the last decade, while the decrease in smoking and improvements in less-invasive techniques have happily led to much less work for thoracic surgeons.  

The current impasse in Congress makes it more likely that the SGR will not be fixed for the thirteenth or so time since 2003 – which could lead to these across the board cuts. This could well lead to access problems for Medicare beneficiaries – it’s a terrible way to lower health care costs!

There is a better way.   This is reprinted from a 2009 post:

The SGR was not a “glitch,” but it was a poorly designed way of trying to prevent overutilization.  The problem is that the benefit of increased revenue to individual providers overwhelms the risk of a pay cut due to overall higher than expected utilization.  This is a classic “tragedy of the commons” problem – where it pays for each individual provider to do more procedures, knowing that her contribution to the “overgrazing” will be overwhelmed by the practices of the general population. 

SGR is poorly designed because the group of procedures it applies to (the equivalent of the “pasture” in the tragedy of the commons) is too big – and no physician would rationally think about cutting back on utilization to prevent future fee cuts.  There is a better way.  Japan has an SGR-equivalent which is by individual service –not generic across all services.  

Prices are revised individually, adjusted for each procedure and drug, and not by an across-the-board conversion rate. In particular, the prices of procedures that show large increases in volume tend to be decreased. (Ikegami  and Campbell, Health Affairs, 2004)       Harvard Link 

As a practical matter, procedures with large increases in utilization are sometimes those where there is new evidence of efficacy, but they are likely to be procedures with an exceptionally high margin.  This method of adjusting helps diminish the excess margin associated with particular services, so there is less likelihood they will continue to be overused. 

So – we should get rid of the SGR – it’s not effective at changing physician utilization, and would cause politically infeasible across-the-board cuts.  As long as we are using primarily fee for service payments, Medicare should adopt the Japanese approach to targeted fee cuts for certain procedures if the volume increases. 
Click image to enlarge.