Dartmouth Atlas Project moves beyond Medicare data

During the last decade, the Dartmouth Atlas Project has helped spark a national discussion about geographic variations in both the volume and cost of health care across the country. But most of the work on the Dartmouth Atlas has been based on Medicare claims data.

In January, the Dartmouth Institute for Health Policy & Clinical Practice announced that it will undertake two new projects that it hopes will shed light on whether the same variations exist for patients with private health insurance and patients younger than 65.

This isn't the first time the Dartmouth Atlas Project has looked at commercial claims data, but previous studies have been "very small looks" compared with the ones in the works, said David Goodman, MD. He is director of the Dartmouth Institute's Center for Health Policy Research and professor of pediatrics and health policy at Dartmouth Medical School.

Pediatric patients will be the focus of the first project, made possible in part by state programs in Maine, New Hampshire and Vermont that require all insurers to report claims data, Dr. Goodman said.

Dartmouth researchers will look at claims from private and Medicaid plans to identify whether pediatric care shows the same variation found in the Medicare population, he said.

"Unlike Medicare data, the data for children's care is often incomplete, either including only Medicaid or one commercial plan, and restricted to a very specific place," he said. The claims reporting mandate in the three states solves that problem, allowing researchers to see data from all payers, he said.

The work will be funded through a two-year, $500,000 grant by the Boston-based Charles Hood Foundation, which funds scientific and medical research in support of children's health.

The second Dartmouth Atlas study will use data from Blue Health Intelligence, an affiliate of the Blue Cross and Blue Shield Assn., and focus on orthopedic procedures -- specifically, joint arthroplasty and knee arthroscopy. (See correction)

BHI Chief Informatics Executive Andrea Marks said Dartmouth researchers approached BHI about looking at commercial insurance claims data to see whether it would match what they had found in the Medicare population. The Blues and Dartmouth Atlas researchers knew that the number of orthopedic joint procedures has been growing rapidly and has varied in volume geographically, so it made sense to look at those claims, Marks said. That doesn't mean some of those procedures are unnecessary, she added.

"Understanding that variation is one of the key foundations before you can even start to address the appropriateness of care," she said.

She and Dr. Goodman said gaining that understanding is particularly urgent, because those orthopedic procedures are expected to become increasingly common, not only among the elderly but also among the middle-aged.

"I would describe them as both high value and high cost," Dr. Goodman said. "They make a real difference in patients' lives. They're also quite expensive."

He said researchers expect to have results to report for both projects in 18 months to two years, but there is no exact timeline for publication.

The Atlas Project is best known for uncovering variations in Medicare spending that Dartmouth researchers said were not explained by regional differences in cost and income or by severity of illness. Their research was reported in two Annals of Internal Medicine studies in 2003.

Surgeon and journalist Atul Gawande, MD, MPH, used Dartmouth Atlas data in his 2009 New Yorker article examining the reasons spending varied so dramatically between the towns of McAllen and El Paso, Texas. Dr. Gawande concluded that per capita Medicare spending for patients 65 and older in McAllen was 86% higher than in El Paso. By comparison, a Health Affairs study published Dec. 7, 2010, said spending per member, per year, was 7% lower in McAllen than El Paso among patients younger than 65 who were insured by BlueCross BlueShield of Texas.

President Obama cited Dr. Gawande's work as evidence of the need for health system reform.