The Dartmouth Atlas has become an Exhibit A of evidence that the system of providing healthcare in the U.S. is chaotic. It shows enormous variation in regional practice, Medicare spending, and utilization of healthcare resources in the last six months of life for 12 chronic illnesses, without any improvement in quality.
But a new relatively low budget study by researchers at six teaching hospitals in California may have unveiled the Dartmouth study's Achilles' heel. These academic physicians found lower mortality among patients hospitalized at facilities that used more resources—such as days in the hospital and procedures—than those that used fewer resources.
"Our study puts a significant dink in the Dartmouth Atlas armor," says Ted Ganiats, MD, one of the paper's authors and interim chair of the Department of Family Medicine at the University of California Medical Center in San Diego. "The insight the Dartmouth atlas provides may not be accurate."
The study, "Looking Forward, Looking Back: Assessing Variations in Hospital Resource Use and Outcomes for Elderly Patients with Heart Failure," was published last week in the journal, Circulation.
Instead of looking just at care for Medicare beneficiaries in their last six months of life, called the "Looking Back" approach as the Dartmouth researchers have done, the California study also looked at all 3,999 elderly individuals hospitalized between January 1, 2001 and June 30, 2005 for heart failure, the "Looking Forward" Approach.
They also looked at a subset of those 3,999 patients—the 1,639 individuals who died—and compared the care they received during the previous six months.
Ganiats says in addition to the expense of days in the hospital, spending may have involved more aggressive cardiac catheterization and more attempts to determine optimal medication doses or increased echocardiography. It may have also included more coronary artery bypass graft procedures or left ventricular assist devices.
"By only including individuals who have died in the analysis, researchers cannot identify differences on health outcomes such as survival," the researchers wrote.