April 9, 2012

Study: Public disclosure of hospital infection rates vary by state

Patients ‘walking in blind’ with little access to quality and outcomes data

Only 21 states require public reporting of hospital data on surgical site infections, and even when disclosure is mandated, the information is often not easily accessed by patients who could use it to make decisions about their medical care, according to new Johns Hopkins research.

The findings suggest that a haphazard, state-by-state system for reporting these critical measures of health care quality isn’t working, and that only national guidelines governing disclosure can paint a clear picture of how well hospitals are doing at preventing patient harm, the researchers say.

Reporting accurate data on measures such as rates of surgical site infections can be an inexpensive way to actually reduce them, the authors note in their study published online in the Journal for Healthcare Quality.

When patients have access to this information and use it to take their business to hospitals with lower infection rates for select operations, the researchers say, hospitals with higher infection rates will have financial and reputational incentives to quickly find ways to do better.

“A lot of information is not available to the public and, if it were, hospitals would be motivated to improve,” said study leader Martin Makary, an associate professor of surgery at the Johns Hopkins University School of Medicine. “Right now, a hospital can have high complication rates, high readmission rates and high infection rates, but because patients can’t look up this information, they’re essentially walking in blind.”

One example of the impact of such transparency occurred in New York State two decades ago, Makary says. Rates of mortality from coronary artery bypass surgery varied widely among hospitals before the state began requiring public reporting of death rates from the procedure; four years into mandatory reporting requirements, average hospital death rates from the operation fell by 41 percent. Makary says that he thinks one reason for the precipitous drop is that “poorly performing hospitals had an incentive to look better to consumers making health care decisions.”

Researchers estimate that surgical site infections occur in up to 25 percent of patients after major surgical procedures, and cause more than 8,000 deaths a year. The occurrence of a surgical site infection is increasingly recognized to be largely preventable, and, as a result, rates are being used as a surrogate measure of broader health care quality.

The Centers for Medicare and Medicaid Services recently announced that hospitals must soon report surgical site infection rates for select procedures. Failing to meet benchmarks will result in financial penalties. But Makary says that the new requirement covers only a small number of procedures, and wider reporting of complications will initially be voluntary. Medicare, he says, needs to quickly expand the program and speed up the transition to uniform public reporting for all hospitals.

In the new study, Makary and his colleagues found that as of September 2010, 29 states had no laws regarding the monitoring and reporting of surgical site infections. Of the 21 that did have such laws, only eight made the data publicly available in an easy-to-access format.

Even then, he says, the data shared are limited, covering between two and seven procedures. Seven of the eight states reported surgical site infection rates following coronary artery bypass graft procedures, six did so for knee or hip replacement surgeries, and two reported rates after colon surgery, which nationally has the highest rates of surgical site infections. Only one state, Ohio, reported rates after gallbladder surgery, among the most common surgical procedures in the United States. The average time lag between collection and publication of data was six months, with a range of two to 11 months.

Makary also says that states don’t always specify how data are to be collected, resulting in lack of uniform reporting that can make comparisons impossible. The lack of national standards, he says, may also disadvantage hospitals that are better at collecting information because their rates may appear higher than those at hospitals that don’t look as rigorously for infection cases.

“It is important to use a common method or, at a minimum, ensure common parameters, inputs and definitions are used,” he said. “Without that, it is difficult for consumers, payers or regulators to compare infections within or across states. Unless we are comparing apples to apples, public disclosure has the potential to mislead patients instead of help them.”

Other Johns Hopkins researchers involved in the study are Monica S. Aswani, Andrew M. Ibrahim, Elizabeth C. Wick and Peter J. Pronovost.


Related website

Martin Makary