Skip to main content

Who's Being Transparent With HAC Data?

 |  By cclark@healthleadersmedia.com  
   April 14, 2011

For months, the American Hospital Association argued against the public release of hospital-acquired condition data linked to specific hospitals.  Last week the Centers for Medicare & Medicaid Services made it public anyway.

CMS posted on Hospital Compare a link to a 26,888-line spreadsheet documenting rates of preventable harm, called hospital-acquired conditions or HACs, for each of 3,361 hospitals, culled from some 19 million Medicaredischarges.

The data makes public rates of falls and trauma, retained surgical objects, incompatible transfusions, poor glycemic control, vascular catheter-associated infections, catheter associated urinary tract infections, complications from pulmonary embolism, and stage III and IV pressure ulcers for each hospital.

Now payers, beneficiaries, and providers who can open a spreadsheet can see, for example, that at Florida's South Bay Hospital .751 per 1,000 discharged patients acquired pressure ulcers during their hospital stays.

Or that Presbyterian/St. Luke's Medical Center in Denver had nearly four vascular catheter-associated infections per 1,000 discharges while at Fresno Surgical Hospital in California there were 8.655 incidents of falls or other hospital-acquired traumatic incidents, such as shocks or burns.

To compare, hospitals such as Eastern Long Island Hospital, Peconic Bay Medical Center in New York, and Community Memorial Health Center in South Hill, VA reported none.

"Any potential preventable complication of care is unacceptable," Don Berwick MD, CMS administrator, said in a statement April 6.

Nancy Foster, vice president for quality and patient safety policy for the AHA, says her organization did try to keep CMS from publishing these hospital-specific statistics because "measures used to portray safety and quality data should be good, reliable, valid measures of what they're trying to portray. There are much better measures than those included in the HAC data."

"We urged CMS to take a step back," she said, but the agency declined. "They just told us 'these are the data we have and we're going to put them out.' "

For starters, Foster says, the numbers are not adjusted for severity of patient illness, age, co-morbidities, or other factors that might even the playing field for hospitals with sicker patients.

Second, this Medicare claims data was assembled for billing or administrative purposes, not for measuring quality. Much more appropriate would be to use the Centers for Disease Control and Prevention's National Healthcare Safety Network data, which is much better for tracking patient records for avoidable errors.

Third, she points to what she calls "the lunacy of little numbers," that hospitals with few eligible discharges might have one or two mishaps that disproportionately make their overall HAC rates look terrible.

Hospitals are already getting hit with financial penalties when they harm patients. For example, prior to Oct. 1, 2008, a hospital caring for a patient admitted with a stroke who subsequently developed a preventable pressure ulcer would pay an average DRG of $8,030.28 for care related to both the stroke and the pressure ulcer. But since Oct. 1, 2008, CMS no longer pays for care for hospital-acquired pressure ulcers, bringing the hospital's payment down to $5,347.98.

Realistically, that's not a lot of money to force change and the actual penalties imposed are rare. That's because the patients who suffer these hospital-caused mishaps already have multiple diagnoses that put them into a higher DRG category anyway.

 

And truth be told, it's not as great an incentive for hospitals to improve as is the threat that the numbers are now public.

Two sets of anticipated federal regulations will further penalize hospitals with high rates of medical errors that cause patient harm, and they'll be publicly viewable too. One is the rules for value-based purchasing criteria; the other is a set of rules that specifically penalizes hospitals with high rates of HACs.

Jim Conway, senior vice president of the Institute for Healthcare Improvement, says that despite the CMS' spreadsheets flaws, public release of these hospital specific details is essential for the sake of transparency as long as people understand the limitations "and don't use it as a club."

In Massachusetts, he recalls, when members of the Commonwealth Quality and Cost Council were debating whether to make quality data public "one member said 'It may not be perfect, but it will allow us to ask questions we didn't know to ask before. This has to be about transparency.' "

With the CMS' data, Conway emphasizes, the numbers are as revealing for what they do not show about rates of HACS as for what they do.

"There are a number of organizations that reported none. I am more concerned about the 'none' than I am about (those that) reported a lot, because I want to know if the culture at those organizations that reported none encourages reporting," he says.

"When you're on this journey, you have to become expert at looking for trouble, the more you find," he adds. "I'm worried a 'zero' means ... the organization didn't know there were any. We know there's a lot more harm in the healthcare system than we now know."

Conway adds that these statistics are important not just for patients to use for comparison, but for payers and health plans to benchmark against their competition, for providers to reduce their misuse of services, for employers who are choosing providers and for policy leaders to see ways to reduce variation.

Gina Pugliese, vice president of the Premier Healthcare Alliance Safety Institute, says that hospitals are gradually coming to understand that public reporting of these numbers is inevitable. That has prompted organizations to "really raise the stakes in investing in, developing and supporting programs to minimize these HACs. I do think ... it's helping hospitals prioritize."

Linda Greene, director of infection prevention at Rochester General Hospital in New York and a member of the board of the Association for Professionals in Infection Control, also says more hospitals are being more proactive, getting cultures on more patients upon admission so they can avoid getting blame for infections the patients brought with them.

"People are understandably worried about [the] preciseness of this documentation of whether something was coded as present on admission or not," she says.

The spreadsheet, which will eventually be transferred to an easier-to -decipher forma, allows users to compare individual hospitals with national averages for each category:

Falls and hospital trauma—10,564 or .564 per 1,000 eligible discharges

• Vascular catheter-associated infections—6,868 or .367 per 1,000 eligible discharges

Catheter-associated urinary tract infections—5,928 or .316 per 1,000 eligible discharges

Stage III or IV pressure ulcers—2,521 or .135 per 1,000 eligible discharges

• Poor glycemic control—944 or .05 per 1,000 eligible discharges
• Retained surgical objects—484 or .09 per 1,000 eligible discharges
• Complications from an air embolism—53 or .003 per 1,000 eligible discharges
• Incompatible blood transfusions—23 or .001 per 1,000 eligible discharges.

Eventually, the numbers, which reflect discharges between Oct. 1, 2008 and July 1, 2010, will be available with other quality measures posted on Hospital Compare. Until then, the spreadsheet can be downloaded here.

 

Tagged Under:


Get the latest on healthcare leadership in your inbox.