The phrase “health care-associated infections” (HAIs) gives quite strongly the impression that illnesses falling under this category are acquired in the hospital. But a recent study found that three of the most talked-about HAIs are not so definitively associated with hospitalization.
An analysis of diagnostic codes from four million discharges at 167 teaching hospitals in 2009 found that 70% of sepsis/septicemia cases, 86% of methicillin-resistant Staphylococcus aureus (MRSA) infections and 67% of Clostridium difficile infections were coded as “present on admission.”
In an article published in the November/December 2011 American Journal of Medical Quality, the researchers concluded that public reporting of hospital infection rates should include information about whether infections were present on admission.
Lead author Donald E. Casey, Jr, MD, MPH, MBA, FACP, chief medical officer of Atlantic Health System in Morristown, N.J., and associate professor of medicine at Mount Sinai in New York, recently talked to ACP Hospitalist about additional lessons that can be drawn from the study's findings.
Q: Before you conducted the study, did you suspect that so many of these infections were present on admission?
A: We had some idea, based upon our own data. That caused me to wonder whether other hospitals were experiencing the same result, and lo and behold, it looks like that pattern is pretty consistent.
Q: Why did you focus specifically on sepsis, MRSA and C. diff?
A: Those are major areas of concern for us on the inpatient side and we actually have several initiatives in place in concert with our hospitalists and many other people to reduce or eliminate these infections. Once we uncovered…that this is pretty much a national trend, it allowed us to bring back to the teams the question of whether we can do something now to prevent some of these infections that are occurring prior to hospitalization.
Q: How should hospitalists feel about your results? Is it good news that many of these infections aren't beginning under their care?
A: There's a natural reaction to think that way. Yet on the other hand…there are traditional assumptions about this issue that need to be challenged and studied further. There were assumptions that they were all coming from nursing homes, and it turns out that many of these patients were not coming from nursing homes. In fact, quite a few were coming from home.
We also don't know from the data when patients first became infected. “Present on admission” doesn't necessarily mean they didn't get it in the hospital. In other words, is it possible that a patient was colonized with MRSA in a hospital two years ago and was then a carrier in a nursing home or at home until ending up back in the hospital with a clinical infection? We're clearly not paying enough attention to eradicating and preventing these infections when they occur outside the walls of the hospital.
Q: What do you think needs to be done for prevention?
A: We've been working with some of our long-term care facilities, for example, over the past couple of years to establish the same infection prevention standards and goals together. They don't have, unfortunately, the measurement systems and surveillance tools that we have in place, and they don't have the detailed expertise that we have, so it's a little difficult. But we have tried to embed guidelines around when not to use urinary catheters, for example, or when to avoid the use of antibiotics that might cause the emergence of resistant organisms. We've had some success with that. The nursing homes are very agreeable to trying to harmonize our efforts.
But we have another set of infected patients who don't come from the nursing homes, which poses a much more difficult dilemma. Some of them may be coming by way of a physician's office. Some infections are acquired out in the community without any direct relationship to health care. There are recent data showing that the incidence of community-based MRSA has been rising over the past five years.
We're used to thinking about problems like this in our silos of care. What we've got to do is not just react by treating this population of patients, but really focus on the antecedent and subsequent effects of the care of these patients. That's going to require a much different mindset than the reactive one we currently have. Perhaps accountable care organizations will recognize this problem as a big opportunity to improve quality and reduce unnecessary hospitalizations. It's so early on, it's hard for us to figure out when that's actually going to happen.
Q: Some of the existing efforts to prevent these infections focus on reporting requirements and payment penalties. Should these results change that?
A: I'm unclear about that. I'm also unclear whether punishing healthcare organizations by reducing their payments is the right way to solve this problem. Well before there was [nonpayment for infections], we took every infection that happened seriously, and we took every effort to prevent infections seriously. We haven't taken them any less seriously because of [the penalties]. I understand the need for an emphasis on prevention and the importance of transparency, but it's still unclear if these policies will change our practice. Wouldn't it make better sense to invest more directly in expertise and systems designed to improve identification and treatment of patients with these infections before they got sick and required hospitalization?
Q: Your study also found that patients whose infections weren't present on admission had longer length of stay and higher mortality. What does that indicate?
A: We know from daily practice that very sick patients with multiple medical problems who end up in ICUs for days with lots of comorbid conditions are at higher risk for infection, by simple virtue of the fact that they've become debilitated and their immune system capacity diminishes. We've got to double up our efforts on the inpatient side to be sure we prevent these life-threatening infections.
The “good news/bad news” is we've been seeing a decline in the numbers of infections that are not present on admission, but we're also seeing in our own facilities a disturbing rise in the number that are present on admission. We didn't do a formal trend analysis by year, but a global trend appears to be happening nationally in this regard.
Q: Are there other lessons for hospitalists in your findings?
A: It's reinforcing the old-fashioned messages, such as wash your hands religiously; use appropriate methods to prevent infections 100% of the time, like the care bundles for catheter-based urinary tract infections, central line-associated bloodstream infections and ventilator-associated pneumonias; avoid whenever you can any injudicious or inappropriate usage of antibiotics; avoid using proton-pump inhibitors, which may increase the risk of Clostridium difficile. If patients have the unfortunate outcome of becoming infected, recognize that early on and treat it before it gets too advanced. All those messages are messages we've given to hospitals, hospitalists and other physicians for a long, long time. This is just helpful reinforcement.