Healthcare executives may experience deja vu this month as the government publicizes something not seen for a whilehospital death rates.
On June 21, the Hospital Quality Alliance will be posting hospital-specific cardiac-mortality rates on the Hospital Compare Web site. The alliance is a public-private partnership of hospitals, government agencies, purchasers and consumer groups that includes the CMS. It created the site to provide quality-of-care information to the public.
In March 1986, the Health Care Financing Administration, as the CMS was then known, first released death rates for hospitals based on data from 1984. At the time, the media referred to the information simply as "the death list," and many hospital executives contended the raw data could mislead patients. This time around, federal officials and healthcare quality experts are hoping the refined mortality data release will be better received than previous efforts.
This month the CMS will post 30-day, post-admission mortality rates for heart attack and heart failure of patients for more than 4,000 hospitals that are Medicare fee-for-service beneficiaries. Nancy Foster, vice president for quality and patient safety at the American Hospital Association, a member of the Hospital Quality Alliance, said unlike most information on the Hospital Compare site, the data are strictly for Medicare payments. The CMS said only Medicareusing its claims database that reflects care to the Medicare fee-for-service populationhas sufficient national data to meaningfully assess outcomes for patients who suffer from acute myocardial infarctions and heart failure.
Prior to the release, the CMS conducted a "dry run," which provided participating hospitals an opportunity to view performance (that was not reported publicly) from November 2006 until January 2007, based on hospital admissions data from 2003. The posting this month will reflect data taken from hospital admissions between July 2005 and June 2006. Hospitals will fall into one of three categories: "no different than the U.S. national rate," "better than the U.S. national rate," and "worse than the U.S. national rate," according to Michael Rapp, director of the quality measurement and assessment group at the CMS.
"We're looking at this as one component of a broader effort of transparency," Rapp said. "The main significance of this is (it's) the first foray into outcomes, as opposed to process measures."
Rapp called the methodology "sophisticated," and said the National Quality Forum endorsed the methodology to measure those rates. He said there are "potential criticisms" about using administrative data vs. clinical data. Some are skeptical that administrative data can measure outcomes effectively.
"Claims data are useful because they are inexpensive, ubiquitous and readily available, so the tendency is to use it," said Fred Edwards, a cardiothoracic surgeon at the University of Florida at Jacksonville. "But just because it's readily available, doesn't mean it's the best data to use. It's not designed for quality."
Angie Ramirez, medical director of performance improvement and clinical safety at Parkland Health & Hospital System, Dallas, agreed that claims data do not, as she put it, "tell the whole story." For example, she would expect a different outcome for a 35-year-old diabetic who is healthy and suffers a heart attack than an 80-year old diabetic with hypertension who is not managing the disease properly with medication. Claims data would show that both patients are diabetics, but not the severity of their conditions. In addition, Ramirez said there are other outside factors that cannot be considered in administrative data. In the case of Parklandwhich she said treats an indigent populationsome patients come in later for treatment, may not have access to primary care and cannot afford to purchase medication after they leave, all of which could affect their mortality rate.
"It's good as a start if it's understood only as a starting point," Ramirez said of the data. "Once you put it out there, people will start making judgments about hospitals."
While Ramirez said this effort has good intentions, she also said patients do not make decisions based on this type of data.
Eileen Sampanes, the clinical excellence and patient-safety officer at Christus Health in Houston, agreed. "People are still selecting quality on convenience or where their doctor sends them," Sampanes said. "It is more relationship-driven than data-driven. Still, you have to have confidence in your caregiver. Having that, people are more likely not to pay attention to that data."
She added that Christus embraces the notion of transparency with its financing and is already posting clinical outcomes data.
This story initially appeared in this week's edition of Modern Healthcare magazine.
What do you think? Write us with your comments at [email protected]. Please include your name, title and hometown.