Overuse also varied by region. Across six categories—cancer screening; diagnostic and preventive testing; preoperative testing; imaging; cardiovascular testing and procedures; and other surgery—researchers found all categories of unnecessary medical care climbed or fell by region, except cardiovascular testing and procedures, where the study found no regional correlation.
The results underscore the potential for savings from efforts to eliminate medical care that has been identified by previous research as unnecessary, and wasteful, such as colorectal screening among those age 85 and older with no history of colon cancer.
Such savings have been the target of quality campaigns, such as the American Board of Internal Medicine Foundation's Choosing Wisely push, which highlights potentially unnecessary services identified by multiple medical groups. But that effort also has drawn criticism for failure to include more costly services that may diminish physician revenue.
The latest findings account for a fraction of potential overuse, and the study's 26 measures are an initial step toward more direct metrics that can be widely used by policymakers, health insurers and others, the lead author said.
“We won't fix the healthcare system by getting rid of just these 26 services,” said Aaron Schwartz, a student in the Harvard University School of Medicine, who co-authored the paper with Harvard University health policy professors Dr. Bruce Landon, Michael Chernew and Dr. J. Michael McWilliams and Adam Elshaug of the University of Sidney.
The study also highlights the challenges in identifying useful measures.
Researchers screened hundreds of potential measures but excluded those for which the evidence is unclear about whether services lack benefit, Schwartz said. The analysis reviewed—and in some cases, revised—measures of low-value care from the Choosing Wisely campaign, the U.S. Preventive Services Task Force, the National Institute for Health and Care Excellence, medical literature and Canada's Agency for Drugs and Technologies in Health.
Researchers also limited the list to metrics that could be tracked using claims data, which is widely available, though it lacks the detail of medical records, he said. Greater detail would help to more accurately identify when services are inappropriate, but medical records are not nearly as accessible as claims.
The researchers sought to develop a list that might eventually be useful to policymakers in efforts to measure appropriate use and spending under new delivery models, such as bundled payments, Schwartz said. Overuse measures may be less helpful and more risky in efforts by insurers to exclude coverage or increase out-of-pocket spending for unnecessary care, he said, because the measures may fail to accurately identify inappropriate from needed care. “We've developed a tool and a tool is not the right tool for every job,” he said.
Variation from one region to the next is “a promising finding that suggests we're not just measuring 26 idiosyncratic things,” he said. Underlying the patterns may be regional variations in patient preference, physician practice or insurance policy, he speculated, though researchers did not examine that question.
Follow Melanie Evans on Twitter: @MHmevans