Medicare part B spending on medical imaging per beneficiary is down in almost every state since peaking in 2005 and 2006, according to a study published in the American Journal of Roentgenology.

In a Harvey L. Neiman Health Policy Institute (HPI) study, Andrew B. Rosenkrantz, MD, of the NYU Langone Medical Center department of radiology, and colleagues did a study of Medicare part B data from 2004 to 2012.

The authors found that, on average, spending decreased 4.4 percent per year from 2006 to 2012.

Spending was down in 2012 from the 2005/2006 peak in all but two states: Oregon and Maryland. Maryland has seen a 12.5 percent average annual increase since 2005, and Oregon has had an average annual increase of 4.8 percent since 2008.

Rosenkrantz and colleagues noted that both states have federal waivers in place that support their state health plans, and it's those waivers that likely caused the difference.

“It is possible that the unique state systems have shielded Maryland and Oregon from the effects of federal legislation that have affected spending on medical imaging in every other state,” the authors wrote. “This observed divergent trend in a limited number of states highlights the value of granular state-level geographic analyses and thus has potential for more meaningfully informing policy decisions than do analyses with only a national or even a large regional basis.”

In 2012, Florida ($367.25) and New York ($355.67) had the highest Medicare part B spending on medical imaging per beneficiary. Meanwhile, Ohio ($67.08) and Vermont ($72.78) had the lowest.

In addition, Ohio (12.2 percent), Arizona (8.1 percent) and Tennessee (6.9 percent) had the sharpest decreases in spending from 2005/2006 to 2012.

Co-author Danny R. Hughes, PhD, research director and senior research fellow of the HPI, said these numbers should help with the creation of healthcare policy going forward.

“State and regional spending variation is often quoted in addressing national imaging policy,” Hughes said in a statement. “Now knowing that such variation almost universally describes spending decline can help policymakers arrive at evidence-based policies that use resources efficiently and preserve access to care.”