YES – Eliminating HAIs is feasible and may help better focus prevention efforts.

For many years, research suggested that reducing hospital-acquired infections (HAIs) by one-third was the best we could do. But recent landmark efforts, such as the Pittsburgh Regional Healthcare Initiative (Morb. Mortal. Wkly. Rep. 2005;54:1013-16) and the Michigan Keystone Project (N. Engl. J. Med. 2006;355:2725-32), have shattered our notions of how many HAIs might be preventable, achieving reductions of nearly 70% in central line–associated bloodstream infections (CLABSIs). Similarly, a national campaign in England has achieved a 68% reduction in Clostridium difficile infections (Health Prot. Rep. 2012;6:38).

These new data suggest that possibly all HAIs are preventable, and there are now many published reports in which institutions have reached zero. For example, the Hawaii experience has shown that a prevention initiative achieved a median rate of zero catheter-related bloodstream infections (Am. J. Med. Qual. 2012;27:124-9).

The advances in HAI prevention are coming in the context of an ever increasingly sick patient population, with patients who are more complicated than ever before. In fact, many of the greatest gains in CLABSI prevention have been among the very sickest patients in our hospitals – in the intensive care unit – as at the Johns Hopkins Hospital in Baltimore (Crit. Care Med. 2004;32:2014-20). So the rationale that we cannot get to zero because our patients are too sick simply is not relevant anymore.

We won’t know how many HAIs are preventable until we hit the bottom. But if our goal is not zero, we will likely end up accepting some infections that might be preventable. Therefore, we must set the goal at zero infections.

What if we don’t get there? That would suggest that some of these infections might not be related to health care delivery but rather to patient risk factors. And these are the infections that we can target. New research could end up preventing some of them. Modifications in definitions might be necessary to account for some of these infections.

One might then argue that we are just defining our way to zero. I would counter that plenty of places have, in fact, hit zero CLABSIs without any changes in our current definitions. But I will also argue that rational changes in our definitions will be needed to help us differentiate truly preventable HAIs and allow us to better focus our prevention efforts.

An excellent example is the new definition of the mucosal barrier injury–related bloodstream infections, which was discussed during the recent IDWeek. Making such an evidence-based change to a definition to focus prevention efforts is neither gaming nor cheating—it’s called good science and good policy.

The zero goal will also improve medical care through the wider adoption of best practices. Published reports of HAI prevention efforts show the successes have been obtained through the implementation of best practices, not through gaming, not through cheating. Zero HAIs as a goal is already driving real improvements in patient safety and quality.

Pushing for zero HAIs will also further research. How much can we argue for pushing the prevention research envelope if we decide that some of these infections are simply okay? How are we going to argue for funding for more prevention research if we tell them the infections aren’t really preventable? We will be much more successful if it’s clear that we need it to get to zero HAIs.

Getting to zero has been put forth as a goal that we should aim for, not as a standard that people should be punished for in the event it is not attained. In fact, value-based purchasing rules are based on relative infection rates – not on zero infection rates.

Moreover, it’s a long-range goal. This is not something that anyone is advocating for overnight. Evidence of that can be seen in the National Action Plan to Prevent HAIs, which calls for 30%-50% reductions, not zero, as the target for the first 5 years.

It must be acknowledged that sometimes an HAI is someone’s fault. Many U.S. institutions have yet to implement best practices for preventing CLABSIs, for example. How long can we hold these hospitals blameless for failing to do this?

In sum, getting to zero HAIs is not the right medicine, but the only medicine. It is our successes, published in our own medical literature, that have prompted the push to get to zero. We shouldn’t run away from the success that we have had; we should embrace it and build on it. Patients will no longer accept a goal of preventing some HAIs, and neither should we.

Dr. Srinivasan is associate director for health care–associated infection prevention programs, Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention. He disclosed no relevant conflicts of interest.

NO – Eliminating HAIs is unrealistic and has unintended negative consequences.

There are a variety of reasons why getting to zero HAIs is not the right medicine. First and probably most important is that it is dishonest. Patients today are sicker and more immunosuppressed, and the devices we use are ever more invasive. When we have patients in our hospitals who, for example, have total artificial hearts in place for more than a year, can we realistically imagine that there would not be any HAIs? I don’t think so.

Extensive medical literature documenting infection prevention initiatives attests to this. For example, a recent analysis suggests that even if all U.S. hospitals implemented all of the measures known to prevent HAIs, at best 55%-70% of common HAIs would reasonably be preventable (Infect. Control Hosp. Epidemiol. 2011;32:101-14). None were found to be 100% preventable.

If one looks closely at the Hawaii experience showing a median rate of zero HAIs, you actually see a 61% reduction in the rate of CLABSIs over 1.5 years, but it never reaches zero (Am. J. Med. Qual. 2012;27:124-9). That translates to up to 10 infections per quarter. So we can use words in loose ways and talk about zero median infections, but there were not zero infections in that study. If the final rate of 0.6/1,000 catheter-days were applied to my hospital, it would translate to 11 CLABSIs annually – hardly a number that would allow me to say I had eliminated these infections.

Efforts to achieve zero HAIs may be a manifestation of postmodernism, a philosophical paradigm in which there is no absolute truth, and one that puts evidence-based medicine on par with practices such as homeopathy and the notion that the measles, mumps, and rubella vaccine causes autism.

Aiming for zero HAIs drives a punitive culture. If we accept the concept of getting to zero, it then means that zero is actually attainable, and if that is true, then all HAIs are preventable. And if that is true, logic tells us then that the occurrence of an HAI must be someone’s fault.

The zero goal places enormous pressure on infection preventionists and their programs, and it creates adversarial relationships between infection prevention services and clinicians as they argue about whether an event is an infection. Hospital administrators ask why programs are not reaching zero. And the infection prevention people are caught in the middle of all this. They are in a terrible position: In some cases, they are deciding – in a way – whether people will get a pay raise or will be fired from their jobs.

Trying to eliminate HAIs fosters problems with surveillance, such as outright cheating, making subtle changes to case definitions that reduce infection rates, and underfunding infection prevention programs to reduce their sensitivity for case ascertainment.

It also leads to inappropriate medical practices. For example, many hospitals now check urine cultures on admission in patients with urinary catheters, or obtain blood cultures on asymptomatic patients simply because they have a central line. We know what those kinds of practices lead to nonindicated treatment and overuse of antibiotics.

Aiming for zero separates infection prevention from quality and safety because now, the "be all, end all" becomes an infection-free hospital stay, when maybe that is not the main goal from a patient perspective.

The zero goal also fosters expedient solutions over the hard work of behavior change. A report predicts that, in 2016, the market for infection-control devices and products will be $18 billion – triple that of the market for antibiotics to treat those infections. So there’s a lot of industry out there waiting to get into the market. That in turn contributes to a conflict of interest. Some leading infectious disease associations now have strong sponsorship from these industries that is likely not serving us well.

Aiming for zero also punishes hospitals that care for poor and sicker patients. As an example, in public reporting of hospitals’ infection rates, academic medical centers may appear to have comparatively worse performance.

Finally, if we already know how to get to zero, why would we ever invest any more in research to reduce infections? It really weakens the rationale for funding in the whole field of HAI prevention.

In sum, getting to zero HAIs is not a realistic or beneficial goal and may actually produce many unintended negative consequences. Clinicians and patients alike would be better served by a focus on achieving realistic reductions.

Dr. Edmond is the Richard P. Wenzel Professor of Internal Medicine and chair of the infectious diseases division, Virginia Commonwealth University Hospital, and an epidemiologist at the VCU Health System, both in Richmond. He disclosed no relevant conflicts of interest.

Point-Counterpoint: Hospital-acquired infections: Is getting to zero the right medicine?