Monday, September 15, 2014

Brookings v. Dartmouth on Health Costs

The Brookings Institution released a study last week that could turn the debate over health spending on its head. While many health analysts—including several key advisers to the administration during the debate over Obamacare—believe that variations in physician practice patterns could represent the key to unlocking a more efficient health system, the Brookings paper questions the degree to which such variations even exist.

At its core, the debate boils down to a difference in two econometric models, both of which attempt to explain geographic variations in spending— for instance, why Medicare spends so much more per patient in Miami than in Minneapolis. Researchers affiliated with the Dartmouth Atlas of Health Care previously found what they consider large, unexplained variations in health spending. Their research—which examines data from individual Medicare beneficiaries, controlled for health status—led them to conclude that differences in physician behavior may account for much of the unexplained spending variations.

The Brookings study, however, uses a different model, one that examines spending data from the state level, and controls those state data using average health attributes in that state, rather than using data from individual Medicare beneficiaries. This state-based model explains much more of the previously unexplained geographic variation in spending, arguing that states with similar demographics have similar spending levels. As a result, the Brookings paper concludes—contra­ Dartmouth—that “geographic variation in health spending does not provide a useful way to examine the inefficiencies of our health system.”

It’s unclear who has the more accurate model, and why. While Brookings’ state-level model incorporates data from both Medicare and non-Medicare beneficiaries, the Dartmouth research focuses just on Medicare patients—and may therefore be skewed by traits particular to the Medicare program, or Medicare beneficiaries, that do not apply to the population as a whole.

The debate over spending variations has profound policy implications. Former Obama administration official Peter Orszag, who has cited Dartmouth research in his writings, believes that variations in physician practice patterns—doctors performing too many tests, for instance—lie at the root of the unexplained variations in spending.  Mr. Orszag and others used this theory to inform many policy choices related to Obamacare, which included a variety of carrots and sticks that attempted to change physician behavior and reduce spending variations.

The Brookings study undermines the basis of the Dartmouth thesis, and one of the reasons why Obamacare’s adherents believe the law will ultimately reduce health costs. Despite its arcane details, the debate between Dartmouth and Brookings will have profound real-world consequences for our health system in the coming years.

This post was originally published at the Wall Street Journal Think Tank blog.