Public Reporting of HAIs: Where Do We Stand Now?

When a patient at a healthcare facility gets an infection, staff members are expected to immediately mobilize resources to stop it. But not all healthcare organizations are on the same page when it comes to addressing the problem of how many such infections occur in the first place. For several years now, a movement has been underway to improve patient safety by tracking healthcare-associated infections (HAIs). Nearly one third of the 50 states have already passed laws requiring hospitals to report rates for various infections. Many other states are in the process of following that lead. Theres still some resistance to public reporting, but the trend is definitely toward more disclosure.

Some organizations feel that reporting shouldnt be mandatory at all, but rather voluntary. Paul Levy, president and CEO of Beth Israel Deaconess Medical Center in Boston, disagrees. The best way to get the best infection reporting is to move ahead, he says. In his widely read blog, Running a Hospital, he wrote in January: I have chosen to post BIDMCs data because I believe the only way for us all to learn how to do this well is to actually do it.


No one contests the fact that HAIs are a raging problem. Figures from the Centers for Disease Control and Prevention (CDC) show that an estimated 2 million patients per year or about one patient per 201 contract HAIs and some 90,000 die from them. With treatments and extra days of hospitalization coming to about $4.5 billion annually, HAIs balloon the overall cost of U.S. healthcare and threaten hospitals bottom line.2

The Centers for Medicare & Medicaid Services (CMS), in fact, is soon expected to stop paying for two HAIs -- catheter-associated urinary tract infections (CAUTIs) and Staphylococcus aureus bloodstream infection/septicemia under new regulations that go into effect in October. CMS is responding to a requirement from the Deficit Reduction Act of 2005 that it cut costs related to at least two high-cost, high-volume conditions that are preventable and hospital-acquired. Driving this cost-cutting is the recognition that most HAIs appear to be preventable with such simple, easy-to-implement protocols as handwashing before examining patients, other hand hygiene practices, better technology such as anti-infective Foley catheters, and more thorough cleaning of facilities and devices.3 Research by the National Quality Forum shows that hospital clinicians compliance with handwashing generally is less than 50 percent.4

Nonetheless, some facilities are succeeding in their efforts. Staff members at Marshalltown Medical & Surgical Center in Marshalltown, Iowa, for example, have improved their infection prevention rate. How do they do it? Its just by compliance with hand hygiene and due diligence by the staff, says Barb Grabenbauer, the hospitals infection control director. They are all very aware of when someone gets a HAI and they all feel bad. Were a small town.

Nationally, Consumers Union and another non-profit organization called the Committee to Reduce Infection Deaths have helped lead the national movement urging states to require hospitals to disclose their infection rates. The strategy is pure economics. With data in hand, the public will search for the safest hospitals they can find. Hospitals in turn will be motivated to be more pro-active about preventing infections because they need to fill beds.

As of this writing, 15 states have passed laws requiring public reporting of infection rates, with Washington the most recent. Three of those states Pennsylvania, Florida, and Missouri have issued public reports of hospital infection rates. Momentum is plainly on the activists side all of this legislation has passed in the last three years, and 13 more states are considering similar action, including healthcare leader Massachusetts.6

Pennsylvanias report, the most meticulous yet produced, reveals the potential value of infection data. It showed that patients with an HAI had a mortality rate of 12.9 percent, about six times the 2.3 percent rate for patients without infections. Hospital stays were extended an average of 16 days by infections, and insurers paid nearly $46,000 more for these patients than for patients who were infection-free.7 For consumers, it provided detailed infection rates for 168 Pennsylvania hospitals, giving folks a chance to select the safest place for their treatment, imperfect though the data may be.

Issues and Answers

Although the trend is clearly toward mandatory reporting, details in laws and proposed legislation differ widely from state to state. One of the major areas of difference concerns the particular infections to be reported. Many people involved with the issue feel that reporting should begin with those infections that are most prevalent, dangerous and costly that is, the ones whose reduction would make the biggest impact. Presumably, other infections can then be added to the list in phases as the reporting systems and infrastructures mature.

Several important players feel that reporting should begin with the so-called big four consisting of surgical site infections, central line infections, catheter-associated urinary tract infections (CAUTIs), and ventilator-associated pneumonia (VAP). One expert who agrees with this starting point is Paula Bussard, senior vice president of policy and regulatory services for the Hospital & Health-system Association of Pennsylvania. Back when the [Pennsylvania reporting requirements] were being phased in, we felt very strongly about the device-related infections central line, ventilator-associated pneumonias, Foleycatheter associated UTIs and then also on an array of surgical site infections, says Bussard. We knew from the infection control professionals that those infections tended to carry with them greater quality concerns and also were more costly to the hospital. The Association for Professionals in Infection Control and Epidemiology (APIC) recommends reporting on surgical site and bloodstream infections. But data from Pennsylvania first-year statistics would seem to argue for also reporting CAUTIS and VAP. CAUTIs were the most common infection cited in the Pennsylvania document and were associated with 423 deaths in just that one year. This is a rate of one death per 15 CAUTIs. VAPs were nearly as deadly 393 fatalities and the mortality rate was frightening: More than one in four VAP cases. For reasons such as this, Consumers Union supports the reporting of both VAP and CAUTIs.2

A final argument for including CAUTIs involves the likelihood that CMS will require institutions to track them anyway, and identify which infections were acquired in the hospital rather than prior to admission. Shouldnt legislation be consistent with this requirement, advocates ask? Pennsylvania, New Hampshire, and Nevada already require the compilation of data on all big-four infections, and pending legislation in Massachusetts, Oregon and Georgia takes a similar stance.

One of the thorniest issues in mandatory reporting involves the appropriate methodology for doing the tracking. If hospitals employ different methodologies for identifying infections, how can infection rates from one institution be compared to another, APIC and others ask. The utility of the information is a related issue. At Penn State Milton S. Hershey Medical Center, in Hershey, Penn., administrators hired a private company that specialized in medical chart abstracting to track infection data for the initial state report. The company performed its task relatively well for purposes of complying with the state mandate, but the methodology gathered information too slowly to be useful internally. That is, the feedback wasnt immediate enough for infection control personnel to measure their own progress in preventing infections.

Hershey Medical Center has since adopted a different methodology an electronic surveillance system that runs algorithms on microbiology data. Kathleen Julian, MD, who led a study of Hersheys first-year reporting, likes what shes seen from the system so far, especially in terms of timeliness. Its certainly more helpful to us for internal purposes, she says. To me thats the bottom line getting data that hospitals can use and respond to and address in a timely manner.

Shes hesitant to recommend it yet as a national or statewide solution for the methodology issue, however.

Julian is among those who are reluctant to endorse reporting until the various controversies can be settled. One of those controversies involves patient mix. Some who represent hospitals with high proportions of AIDS or geriatric patients note that because their patient populations are subject to more infections, its not completely fair to compare their data to hospitals with fewer infection-prone patients. APIC is also in this camp, contending that for certain patient populations, some infections may be inevitable. The counterargument goes: Because most HAIs are preventable with basic measures such as compliance with hygiene and sterile technique, is patient mix really that substantial an issue?

The cost of mandatory reporting troubles some players. Julian estimates that it took one FTE in infection control to track data for the first Pennsylvania report, plus the cost for the private company her hospital hired. Grabenbauer says that the cost for Marshalltown was not a major factor because her staff was tracking the infections they had to report to the state anyway, as part of its regular duties. She adds, however, that tracking CAUTIs would be more labor-intensive because of the large number of Foley catheters used, and the fact that a single patient may have a catheter inserted, removed, and reinserted during a single stay. Disclosure advocates agree that all interested parties need to work together to ensure hospitals have sufficient resources for data compilation.

Any objections to public reporting must be weighed against the cost to hospitals of treating avoidable infections, plus the extended stays for infected patients. For instance, one Pennsylvania hospital notes that it loses $15,000 for every central line infection. Ideally, reporting will lead to a dramatic reduction in infections and the widespread use of appropriate preventive techniques more than paying for the cost of tracking.

On the state-vs.-national standards issue, skeptics including the American Hospital Association (AHA) decry the patchwork of state laws that is now emerging. They argue that because hospital corporations operating in more than one state may have to comply with different requirements in different states, their efforts to institute system-wide improvement measures may be hampered. The AHA proposes a national reporting standard instead.

Whats more, the AHA wants the standard to be voluntary, pointing to the example of the Hospital Quality Alliance to which it belongs. The Alliance is a public-private partnership that includes CMS, the Joint Commission on Accreditation of Healthcare Organizations, and many other healthcare organizations. The participating hospitals, of which there are thousands nationwide, voluntarily submit data on several performance measures. Those that dont participate receive a somewhat lower payment from CMS.2

But only one of the performance measures involves HAIs in this case, surgical infections. If the alliance efforts at infection control are so effective, one might ask, why do HAIs remain such a problem? Lisa McGiffert, director of Consumer Unions anti-infection campaign, has stated that while national standards are a laudable goal, it makes more sense to start with state laws and let the states be laboratories for different approaches. The best approach will eventually be obvious, she says, at which point a national system can be instituted.

In the meantime, disclosure champions such as Beth Israels Paul Levy say that reporting is not only good for the public but for hospitals themselves. First, the public has a right to know about fundamental measures of patient quality and safety in institutions in their communities, he writes. Second, as a management tool, there is nothing more effective for hospital administrators than to be able to remind their staffs that actual clinical results will be made public.

Alan Reder is a freelance writer who frequently reports on healthcare issues.


1. RID 2. Rollins G. Reporting infections: Mandatory vs. voluntary debate heats up as states enact legislation. Hospitals & Health Networks. May 15, 2005.




6., except 13 states: 

7., except 13 states:

Q & A
Hospitals in Pennsylvania Support Public Reporting

By Alan Reder   

In this Q & A interview, Paula Bussard, senior vice president of policy and regulatory services for the Hospital & Healthsystem Association of Pennsylvania, discusses public reporting of hospital- associated infections (HAIs). Pennsylvania has been a national leader in requiring hospitals to report infections rates for the most prevalent serious infections. Its initial report, the most detailed of its kind in the nation, covers HAI rates in Pennsylvania hospitals in 2005.

Q: What is the current situation in Pennsylvania regarding reporting of HAIs and how did it come about?

A: Unlike other states where there has been recent legislation, the Pennsylvania Healthcare Cost Containment Council (PHC4), which is a state agency, just chose to finally implement what previously existed as a law. We had a law back in 1986 that created PHC4. In that law, it said that the council could collect and report information on infections, but they didnt do anything for a long time. Then after their last reauthorization in 2004, they issued to the hospitals a requirement to now report the infections. They get, and have been getting on every patient discharge, the age, the diagnosis, and the discharge status. So in Pennsylvania effective Jan. 1, 2006 it was finally completely phased in for every patient discharge you have to put in a code. Either they didnt have an HAI, or theres a series of codes if they did.

Q: Are the codes specific to the type of infection?

A: No, theyre by body system ww circulatory, surgical site, etc. This is different than most of the other states have done. My familiarity with the other states is that many are following the guidelines of the Centers for Disease Control and Prevention, which suggest that you collect the information by the infection, but you do it by the major, riskier infections that is, the device-related infections which are more costly and more likely to have a negative patient outcome.

Q: Has the Pennsylvania approach been effective in reducing infections?

A: Weve only had one full year of full reporting, so we dont have an accumulated base of data to say yes, theres less. And, of course, with any type of new reporting, you usually see things increase at first before they decrease.

But the reporting phase- in, and the full reporting coincided with the Institute for Healthcare Improvement report, the 100,000 Lives campaign and now the 5 Million Lives campaign which most of the Pennsylvania hospitals are participating in.

So I think you have the combination of improved implementation of best practices coupled with public disclosure. Because the council did produce a hospital- specific report last fall, it has received greater attention in Pennsylvania hospitals and I think we will see the numbers come down in subsequent reporting.

Q: Is there any sign yet that consumers are considering hospitals reported rates of infection when they select a hospital to patronize?

A: No. We are not seeing any significant change in market share, or purchasing pattern.

Q: Is there widespread awareness in the public that this kind of information is available about each kind of hospital?

A: No. In fact, when you do public opinion polls and ask people the information they want and where would they look to get it, theres very little awareness of the availability of any information, whether its from the CMS or our PHC4 or our department of health.

Q: Do you think the Pennsylvania approach of tracking infections by body system instead of different types of infection is a good model for other states to follow or do you think that the more infection-specific approach is better?

A: The professionals who work in our hospitals want data that is both useful to the public and useful to them as providers. If you have a hospital-wide infection rate, what does that tell a pregnant woman about the quality of care in the obstetrics unit?

Or if you have a family member whos going to be having major oncology procedures, again a hospital- wide rate doesnt quite tell you as much as that hospitals central line-associated blood stream infection rate. Chances are, if youre going to have major chemotherapy, youre going to have a central line. So the information is there here in Pennsylvania. The providers are paying attention to it because it obviously is out in the public domain, and they are certainly looking at themselves in some gross kinds of comparison. It could be better, so were continuing to work with our state agencies to make the information better. By better, again it has to be useful to the public and to the professionals.

Q: In other words, because Pennsylvania doesnt collect information on specific infections like central line related infections, the information is not as useful as it could be?

A: Exactly. The neat thing is that the states in this country are serving as the laboratories on reporting. In several years, when more states have been reporting data and Pennsylvania has several years of data reporting under its belt, then well be able to step back and all determine whether the approach taken by other states is better, or is Pennsylvanias a good model to follow. Its hard to make a definitive judgment after one year, but we certainly can see where there are some gaps.

Q: Assuming that there is some merit to the approach of reporting specific types of infections, do you have a position on which infections should be reported?

A: Back when it was being phased in, we felt very strongly about the device related infections central-line, ventilator-associated pneumonias, Foley catheter-associated UTIs and then also an array of surgical site infections. We knew from the infection control professionals that those infections tended to carry with them greater quality concerns and also were more costly to the hospital.

Q: Do you think a mandatory reporting law should be inclusive of all the crucial infections at the start, or should a state start with reporting certain infections and phase in others over time?

A: We would suggest phasing in, because again are you reporting just to be compliant with a law or are you reporting because that data informs the public and practitioners? If all of a sudden you go from no reporting to 100 percent reporting, the focus of your professionals is going to be on compliance with the reporting, not also the use of the data. By phasing in and focusing first on those infections that are more costly and impact quality of care, youre going to see improvements from the use of that data.

Q: You suggest that states start tracking with the so-called big four catheter-associated UTIs, ventilator-associated pneumonia, central line infections, and surgical site infections. How should decisions be made about other infections to phase in?

A: It really depends on the priorities. This is why you bring in the professionals, because if youre having problems with admissions from nursing homes coming in with staph infections, then maybe thats a priority where you start. The state agency should ask the professionals about the problems theyre seeing.

Q: Some people complain that because different institutions may collect data in different ways, it is harder for the public to compare infections rates from various hospitals in a meaningful way. Have you found a widely applicable methodology for infection data reporting in Pennsylvania and if not, is there a methodology you can recommend as a better approach?

A: I have to again go by ICD-9 Codes which are put on uniform bills that a lot of state agencies in other states and the federal government collect and use for quality purposes. They rely on clinical abstracts looking at charges, abstracting to make that coding. It is the same experience with infections. Thats why you have to have standard definitions. You have to have periodic training programs.

There may be some value in what are called electronic surveillance systems in aiding the hospitals in flagging. But all those systems are new.

They all have their strengths and weaknesses. So youre going to have improved data over time, but there isnt any way to get to the goal of comparing apples to apples at this time. Those who recommend these electronic surveillance systems note that they use lab results. Well yes they do, but that doesnt necessary help you to identify ventilator-associated pneumonia. So you really have to focus on periodic education and retraining with your people who abstract the information from your medical records.

Q: On what would the training focus?

A: Youre training them to look for the results, flag the case for the infection control professionals to say yes indeed. it was hospital- acquired or no, it was community acquired. Its no different in many ways from the clinical abstractions that occur now in hospital medical records departments that result in a diagnostic code. Thats a trained and educated person looking at a medical record based on these kinds of flags, and coding it into this diagnosis. The same kind of process is occurring with infections.

Q: How would you summarize the position of your organization on public reporting?

A: The hospitals in Pennsylvania support public reporting. Thats our commitment to report to the public what occurs. Were supportive of transparency.