Hospital Compare on East Tennessee Medical News

Hospital Compare

By: CHELSEA FARNAM

Practicing and maintaining standards for quality care is something every hospital aims to achieve. Particularly in the last five years, however, government health agencies have been actively advocating for across-the-board standardization of those quality measures. The benefits and disadvantages of such a move are demonstrated in such efforts as the Department of Health and Human Services’ (DHS) Hospital Compare Web site.

The site, www.hospitalcompare.hhs.gov, was set up in 2004 by the Centers for Medicare and Medicaid Services (CMS), the DHS, and members of the Hospital Quality Alliance (HQA) in order to work toward standardization of quality measures and hospital accountability. By publishing each hospital’s data concerning process of care measures, outcome of care, and patient surveys, patients and their families are able to compare nearby hospitals and evaluate their apparent successes or failures. Hospital treatment of a selection of common illnesses and conditions are examined: heart attack, heart failure, chronic lung disease, pneumonia, diabetes (in adults), and chest pain.

“It’s very useful,” said Tom Tull, assistant vice president for Washington County operations in Mountain States Health Alliance guest services. “It puts more information in the hands of consumers, and they’re asking the right questions.”

As Tull explained, the site enforces accountability through the publication of data, but also through financial reimbursement. Hospitals can choose not to participate, but CMS will withhold a measure of funding to that facility, ensuring that only hospitals pursuing best practice medicine are rewarded.

“Does it force us as providers to use best practices? Absolutely,” said Tull.

But the site has its flaws, as well.

“The data is always 9-12 months old,” said Tull.

Although the hospital records data on a daily basis (using several online surveys and electronic databases), this data is only collected quarterly—generally in March, June, September, and December. According to the site, the mortality and readmission data are updated only once every year.

The time needed to process the data requires that the 12-month time period represented on the site be at least nine months old. For instance, the Johnson City Medical Center’s current data, which was published online in November 2009, represents events that took place from April 2008 to March 2009.

And since most hospitals internally undergo frequent analyzation of their own quality of care, problems evident in the online data may have been identified and eliminated long before the data was published. The site’s lag time brings about inaccuracies and could possibly mislead the site’s users.

“The difference is that we look at (our data) on an ongoing daily basis,” said Karen Beam, market director of quality at Parkridge Medical Center in Chattanooga. “In many cases, by the time the numbers are published, centers are already putting into action measures to improve those things.”

“You’re comparing maybe an apple and an orange,” said Beam. “Then you just end up with fruit salad.”

The site does state its data collection schedule, including the statement that “there may be lag time,” but instead of being highlighted for consumers, this information is located in the link “View Note to Hospitals,” at the bottom of the page.

“The devil is always in the details,” said John Lacey, MD, chief medical officer at the University of Tennessee Medical Center in Knoxville, commenting on how the site’s data on outcome measures could easily be misunderstood by its users.

“For example,” Lacey said, “a patient who comes in is treated for acute myocardial infarction. He is treated and goes home, but dies in a car accident less than 30 days later.”

According to the data collection methods, this patient’s fatality would be included in the hospital’s mortality rate, regardless of the cause of death’s correlation to his previous hospitalization.

“There are efforts to statistically minimize the impact of those obvious things that don’t connect, but are they completely eliminated? I don’t think we can be sure of that,” said Lacey.

Lacey also remarked on the site’s lag concerning new research and practices.

“In early 2005 the core measure on heart failure said the standard of care was that the patient receive an ACE inhibitor, and failure to do that would turn out as a negative on your core measure of heart failure treatment. Prior to that time, angiotensin receptor blockers (ARBs) were considered a safe alternative for patients who could not handle an ACE inhibitor. CMS at that time did not consider giving an ARB as an appropriate alternative, whereas the science of medicine had approved that,” said Lacey.

According to the site, Hospital Compare changed its quality standard to include ARBs in December 2005.

Similar discrepancies have occurred concerning new research on the method of administering antibiotics to coronary bypass patients and patients suffering from pneumonia.

“There’s research going on all the time. We’re always improving the care we provide, and sometimes we’re being asked to adhere to a standard—and penalized when we don’t—when the science of medicine has moved on,” said Lacey.

One common misconception about Hospital Compare is that it is not risk adjusted. On the contrary, the site’s section titled “Hospital Information for Professionals” details the statistical methods for risk adjustment, including age, gender, past medical history, and comorbidities. This risk adjustment attempts to eliminate such inaccuracies as a bias toward hospitals that treat fewer patients or larger medical centers that treat more severe cases.

“Overall, the concept is valid,” said Lacey. “My fervent hope is that patients make these decisions with the help of their physician. I think the physicians are best able to interpret the validity of some of this data.”