Regional variations in healthcare spending have been described for decades. The Dartmouth Atlas Project has spent years examining Medicare spending patterns across the country. Findings indicate that some areas spend much more than others, sometimes twice as much. Improved health outcomes do not appear to be associated with Medicare spending, after controlling for disease-severity and other factors, including age.

These studies got widespread political attention in a recent article by Atul Gawande in the New Yorker – President Obama called a meeting to discuss it in the Oval Office. The evidence Gawande presents point to the most emotionally compelling cause of variation: supply-driven demand, sometimes pushing deep into fraudulent territory.

This evidence has led to the disturbing observation that supply sometimes may be driving demand. Fixing the source of these variations could halt the subsidization of high-cost regions by lower-cost. The question remains: how much of the demonstrated regional variation can be painlessly reduced by politically feasible options?

Concrete solutions are hard to come by, but in December 2008 the Congressional Budget Office described 115 options to reduce health care spending. Some of these options were intended to address these variations, and their immediate economic impact was quantified.

One option called for the incremental reduction of physician fees in Hospital Service Areas that are in the 90^th percentile. This would have the effect of decreasing physician fees by as much as 30 percent in some areas, while leaving 90 percent of areas untouched. The total savings over the next 10 years would be $5.3 billion. A second, related option calls for reducing payments to hospitals in areas with an exceptionally high volume of elective admissions. This would save a modest $2.5 billion over 10 years.

The most dramatic option proposed would call for global reduction in payments in areas that spend greater than 10 percent more than the national average. These reductions would be scaled so that the higher spending areas receive proportionately larger reductions. For example, a region that spent twice the national average would have payments reduced by 45 percent across the board, thereby bringing total spending to the national average. These calculations assume that the volume of patients would constant, which history shows it would not.

One could anticipate a combination of effects. Experience shows that increased volume would partially make up for lost revenue. Decreases would simultaneously occur due to reduced provider participation in Medicare, and perhaps even small-scale migration of providers. The repercussions on provider availability would be enormous in the highest spending markets, leaving lower-spending areas free to continue on their current, relatively frugal tracks.

That said, if these changes were to result in real savings, the disruptions could possibly be bearable. As the CBO shows, however, the savings would be about $50.9 billion over the next 10 years, an average of only $5 billion per year. To put this into perspective, total Medicare expenditures in 2007 alone were about $420 billion. There would in all likelihood be an adverse spillover effects on the rest of our national health spending, as hospitals in heavily cut areas begin to replace revenue from other sources, namely private insurance.

As the Dartmouth Atlas researchers repeatedly state, these heavy-handed solutions are not what they envisioned. Restructuring health care delivery by emulating successful models of accountability is touted as the ideal solution. On a national scale, this would take drastic intervention that today lacks sufficient political backing. However, smaller models, such as the Mayo Clinic or Cleveland Clinic, expanding gradually as they succeed, could thrive within the frameworks currently being proposed in Congress.

Regional variations are not the disease to be cured; they are the symptoms of a health care system with some counterproductive incentives. A long-term restructuring of health care to reward quality and not quantity could organically, though perhaps not completely painlessly, reduce spending. The difficulty lies in implementing this on a large scale.

Anthony Marfeo is a fourth year medical student at Yale. Anthony.​Marfeo@​yale.​edu.