Posted on Fri, Nov. 12, 2010
last updated: November 13, 2010 07:48:36 PM
WASHINGTON — The Department of Veterans Affairs, which for years has touted the achievements of its health care system, is now highlighting a new study that shows its health outcomes are — about like everybody else's.
The VA touted the study this week, saying in a press release that "VA Health System Shines in Quality-of-Care Study."
"This report is strong evidence of the advancements VA continues to make in improving health care over the past 15 years," VA Secretary Eric Shinseki said in the release. "The systems and quality-improvement measures VA actively uses are second to none, and the results speak for themselves."
The study, just published in the journal Medical Care, synthesized the results of three dozen other studies that compared VA health care to care provided by non-VA providers. It concluded that the VA performed well on many measures of medical care, but that the VA had little impact on the key question of whether the patient lived or died.
Most of the research it depended on to reach its conclusions, however, dates to when Bill Clinton was president. One source for the study is dated 1991, when George H.W. Bush was in the White House.
The issue of how good is VA care is a hot topic among veterans' advocates and Congress. Some recent studies have been used to make the case that VA care is not only pretty good, it's among the best in the country. That's a big turnaround from two decades ago, when the VA was widely derided for poor quality. Since then, the agency has transformed from a hospital-based system to an integrated network of hospitals and clinics that is commended for its emphasis on preventive care.
But the best care anywhere?
What the latest study shows is that the VA performs well on what are known as "process" measures — whether a certain test was ordered, for example. But studies that compare health outcomes — do patients in the VA system do better or live longer? — are equivocal.
Of 12 studies that compared mortality, for example, three showed a better outcome for VA patients, two showed a better outcome for non-VA patients, and seven showed no difference.
That's very different from the process measures, which showed an overwhelming VA advantage.
Researchers aren't sure what causes that disconnect. If veterans are taking their drugs and getting their tests done, the thinking goes, they should be living longer. But for the most part, the data don't show that.
"When it comes to mortality, we found that the VA does no better and does no worse," said study author Amal Trivedi, an investigator at a VA medical center in Providence, R.I.
It would be helpful to study other health outcomes — so-called "intermediate outcomes" that detail health status short of death — but those aren't often measured, Trivedi said.
There were some other caveats to the study. One is what's known as a "publication bias," since most of the studies researchers found were funded by the VA. But researchers weren't sure why that would show a VA advantage on process measure but not one on outcomes.
The lack of fresh data was a central problem with the study.
"We need more recent data," Trivedi said. "There have been a number of efforts in the private sector to improve care." But those efforts wouldn't have shown up in this analysis.
Researchers initially combed through 175 VA quality studies, excluding most for a variety of design or other problems. They were left with 36.
Of those, about 60 percent included data gathered only during the Clinton administration. One included data from 1991.
The freshest data were five years old, collected from 2005.
Joseph Francis, the VA health system's chief quality and performance officer, said the relative age of the research "was kind of surprising to me."
Part of that has to deal with the length of time it takes to complete and publish rigorous research. He also said the VA is working to complete more studies on health outcomes, which will show how VA patients fare compared with non-VA patients.
Other researchers have found it difficult to compare VA care to non-VA care, in part because so many veterans get care from different places.
The Congressional Budget Office, for example, recently reported that 79 percent of veterans in the VA system also had health coverage elsewhere, typically Medicare but also private health plans.
"I think we do have a challenge understanding the totality of care veterans receive — and that includes care rendered outside the walls of our system," Francis said.
In 2008, the VA treated 5.5 million people in its system of about 150 hospitals, 900 outpatient clinics and other facilities.
MORE FROM MCCLATCHY
McClatchy Newspapers 2010