CDC issues template for national mandatory infection-rate reporting
Chasing a genie that’s out of the bottle
Trying to head off a patchwork of mandatory infection-rate disclosure laws that vary from state to state, the Centers for Disease Control and Prevention (CDC) has created a guidance document to serve as a template for new state laws or a national reporting system.
The draft paper’s key recommendations were released recently in Atlanta at a meeting on the controversial but surging movement to require individual hospitals to reveal infection rates to the public. Four states (Pennsylvania, Illinois, Florida, and Missouri) have passed infection-rate disclosure laws, and 20 other states throughout the country have bills introduced for their 2005 legislative sessions. Crafted by the CDC’s Healthcare Infection Control Practices Advisory Committee (HICPAC), the document recommends three process measures and two outcome measures that could be reported under disclosure laws. Infection control professionals should be consulted in the development of such legislation, HICPAC emphasized.
"There is a significant amount of public distrust; and in some ways, we are viewed as foxes in the hen house — but in order to have the best process, ICPs and health care epidemiologists have to be involved," said Patrick Brennan, MD, HICPAC chairman and an epidemiologist at the University of Pennsylvania in Philadelphia. "But we have to accept collaboration. There are many other stakeholders."
The guidance drew immediate endorsements from the two leading infection control groups in the country: the Association for Professionals in Infection Control and Epidemiology (APIC) and the Society for Healthcare Epidemiology of America (SHEA). APIC, a co-sponsor of the meeting, has long emphasized that mandatory rate disclosure systems only will work if there is use of standardized definitions, surveillance methods, and reporting procedures of risk-adjusted data, said Georgia Dash, RN, MS, CIC, APIC liaison to the HICPAC committee.
"Once you review the full HICPAC document, you will see that all of these issues have been addressed," she told those attending. "It is a true collaboration between public health professionals, hospital epidemiologists, infection control professionals, and consumers. It has been written in a way so as to produce a credible and valid public reporting system that really ought to be utilized — not just in the development of individual state reporting programs — but as an actual, national standard."
Michael Tapper, MD, the SHEA liaison to the HICPAC committee read from a SHEA position paper in noting: "However flawed it may be, reporting of quality data to external agencies has become a way of life. SHEA members should do what they do best, promote practices that are epidemiologically sound and evidence-based. SHEA is strongly encouraging its local members to be proactive and offer their services to enhance the process."
Inclusion of SSIs sparks debate
The HICPAC process measures selected for rate disclosure laws are central line insertion practices in intensive care units; appropriate timing of surgical antimicrobial prophylaxis; and influenza vaccination coverage rates for health care personnel and patients. Some changes in current practice may be necessary to track the processes, but there is no denying they are proven patient safety measures.
"In hospitals that are not all already capturing this information [e.g., aseptic practices during line insertion] it will be necessary to have checklists and observation of line insertion practices," said Brennan. "But these are unambiguous indicators, and I think most of us would agree that these practices should occur in every central line insertion."
The outcomes indicators are surgical-site infection (SSI) rates and central line-associated laboratory-confirmed bloodstream infections. The inclusion of SSIs drew a reaction from some in the audience who questioned the validity of such data. Audience members noted that lack of adequate post-discharge surveillance in many settings would invalidate the surgical infection data.
"As best I can tell, we are not doing anything to keep the patients in the hospital longer," said Glen Mayhall, MD, hospital epidemiologist at the University of Texas Medical Branch in Galveston. "They are moving out faster and faster. I have not seen anything published that indicates that there is any sort of valid post-discharge surveillance. I think the [SSI-infected] people remaining in the hospital — 15% or so — are a very skewed population and very sick people. How can we be using surgical-site infections as one of the major things we’re looking at? There is no accuracy in what you’re reporting."
In addition, by definition total-joint replacement SSIs should be counted within a year of the procedure, creating a situation where previous data reports may need to be revised to be accurate, another audience member noted.
Brennan conceded the system was flawed, but SSIs could be included if the type of surgery was chosen carefully. For example, cardiac bypass surgical patients are frequently readmitted if infected so data would be more accessible.
"[SSI] reporting is certainly not something that could be required across the board," he said, adding in regard to the joint-replacement issue that "most of the patients we encounter are not long-term outliers, but it is going to be a fly in the ointment from time to time."
William Scheckler, MD, hospital epidemiologist at St. Mary’s Hospital in Madison WI, reminded the audience that SSIs must be included because the whole rate-disclosure issue is consumer-driven to a large degree. "From a consumer point of view, surgical-site infections are the biggie," he said. "That’s why we have the Missouri law, that’s why we have some other laws. That is why there are a lot of the [infection] stories out there."
Those stories are being told to Consumers Union, the Washington, DC-based publisher of Consumer Reports that has been pushing the adoption of state laws. Lisa McGiffert, senior policy analyst at Consumers Union and director of the infection-rate disclosure campaign, said the HICPAC guidance document was "a significant step in trying to find solutions. It is one of the most important steps that has happened so far."
While the HICPAC document was released with no recommendation in favor or against state disclosure laws, the issue was discussed with an air of inevitability throughout the two-day meeting. In a pointed exchange with ICPs questioning the validity for such reporting systems, McGiffert reminded them that "the genie’s out of the bottle."
Saying she came before a skeptical audience in "good faith," McGiffert told attendees Consumers Union believes disclosure of individual hospital infection rates ultimately will reduce infections.
"We are doing [this] because it is a problem, and I urge you to accept that premise," she said. "It is a problem. [Infected] patients endure prolonged stays in the hospital and sometimes suffer debilitating and long-term health care consequences. Many of them never survive that hospital infection. We have heard from hundreds of people. They are writing to us and telling us their stories. They are real people who have suffered from this."
While consumer groups argue that public disclosure will make hospitals accountable and increase patient safety, many question whether the data gathered will yield meaningful comparisons or create disincentives to aggressively pursue and report all infections. However, ICPs are not necessarily against the idea in principle and many concede it could improve patient safety and increase infection control program resources.
"The controversies have existed over the methods [of reporting]," Brennan said. "That has been construed unfortunately as stonewalling on the part of our [infection control] community. I believe we have a lot to learn about the communication of complex information, such as health care epidemiology. Nevertheless, we believe we can work past the methods and develop good systems."
New takes on preventability, benchmarking
There certainly is no argument that health care-associated infections — the term the CDC now is using instead of nosocomial — are a major public health problem in the United States. In hospitals alone, an estimated 2 million infections, 90,000 deaths, and $4.5 billion dollars in excess health care costs occur annually.
Many infections are, of course, not preventable, but HICPAC edited out language about the issue in order not to appear as "apologists," Brennan said. "We tried to have a finely tuned ear to what is going on in the public. To the extent we appear to be apologizing for poor practice, we thought it would diminish the value of the document."
Likewise, the traditional concept of benchmarking — comparing a hospital’s rate to an aggregate rate for a specific infection — is no longer considered sufficient.
"While it is very good to do that, many hospitals use that [benchmark] as their gold standard instead of really trying to continue to have improvement in terms of prevention of infections," said Denise Cardo, MD, director of the CDC division of healthcare quality promotion. "Being better than a bad hospital doesn’t mean you’re good. It is not good enough just to be better than the average. You really need to look at your [program] and improve more and more over time."
While some infections are inevitable, there are certain areas where better adherence to existing guidelines could increase patient safety. For example, one of the process measures included in the HICPAC guideline is proper timing of antibiotic prophylaxis prior to surgery. Administering the appropriate antimicrobial agent within one to two hours before the incision has been shown to reduce SSIs. Yet CDC surveys and surveillance data reveal that the appropriate timing of antibiotic prophylaxis prior to surgery is performed only about half the time, Cardo said.
"Getting antibiotic prophylaxis at the right time has the same chance as flipping a coin," she said. "It is a concern. We are only going to know how many [infections] we can prevent when we really improve adherence to [proven] practices."
While endorsing the HICPAC effort and the meeting as a "first step in a move to national standards," Cardo also warned that resources to comply with such laws will be an issue. Around 20% of small hospitals don’t have ICPs, so resources to collect and report the infection data will inevitably arise.
"Frankly, in Pennsylvania we have seen that this cannot be done without additional resources unless an organization is willing to divert its attention from prevention activities to the cataloguing of information," Brennan said.
The HICPAC document does not specifically address ratios like the number of ICPs to licensed beds, but it does address resources. "A quality reporting system requires adequate resources," he continued. "There is need for information systems, manuals, staff. The key infrastructure must be available."
Indeed, a labor-intensive focus on gathering data without the resources to enact interventions could actually undermine the effectiveness of infection control programs.
"First of all, we should cause no harm," Cardo added. "We should cause no harm in terms of spending a lot of time collecting data and not preventing infections."