A Nudge in the Right Direction

Several years ago, our hospital faced a number of Clostridium difficile infection (CDI) outbreaks. Our antimicrobial stewardship program (ASP) decided to look for ways to reduce patient susceptibility to this burdensome disease. Knowing that fluoroquinolone use is a major risk factor for CDI, and that ciprofloxacin – a fluoroquinolone drug – was one of our most commonly used oral antimicrobial agents, we thought we might be able to influence prescribing practice by selectively reporting ciprofloxacin. That decision was further supported by the fact that, at the time, our rates of Gram negative susceptibility to the drug were dropping steadily, making it less useful for infections where it was truly needed. We hoped that selective reporting might be able to reduce the development of resistance by reducing overall use of fluoroquinolones.

The general rule of thumb we established was not to report ciprofloxacin for Enterobacteriaceae that were fully susceptible to all the agents (or only resistant to ampicillin) on the Gram negative panel. The most typical example we saw was pan-susceptible Escherichia coli from urine culture. Unfortunately, because our reporting system isn’t automated, compliance with the policy was not perfect – and many clinicians requested reporting ciprofloxacin for blood cultures, due to its good bioavailability and tissue penetration for deep-seated infections. There were some confounding factors, too, like in any before-and-after study – we were in the process of implementing our ASP during the intervention; new Infectious Diseases Society of America guidelines had just come out urging clinicians not to prescribe first-line fluoroquinolones for uncomplicated cystitis; and clinicians may have been discouraged from ciprofloxacin use because of Gram negative organisms’ low susceptibility to it. Nonetheless, we think the rapid and sustained drop in usage – and, indeed, in E. coli susceptibility – after selective reporting definitely indicates that our intervention was a contributing factor. As a result, we plan to look at other selective reporting practices, now in the context of a more robust and stable antimicrobial stewardship program, to see if they also have an impact on utilization and susceptibility.

But selective reporting comes with a warning label. We found that, after our intervention, the use of amoxicillin-clavulanate increased significantly. There were a number of reasons for this, but the problem was likely exacerbated by selective reporting of ciprofloxacin. Selective susceptibility reporting and other types of restrictive approaches are likely to result in “squeezing the balloon,” a phenomenon where overall antibiotic use does not change, but prescribing of targeted agents shifts to another agent. Although we can use this method to shift prescribing from broad-spectrum or powerful antibiotics to more suitable ones, it’s still a concern because our main goal is an overall reduction in unnecessary antibiotic use. That’s something we can achieve by performing selective reporting in combination with other, more active, approaches like prospective audit and face-to-face feedback.

“Nudging” – the concept of guiding the prescriber to make more rational decisions without removing autonomy – can be applied to many other scenarios as well. Discouraging the initiation of antibiotics in patients with likely colonization or contamination by carefully worded comments or interpretations can be useful, as can adding more detail about each therapeutic option (like cost or breadth of spectrum) on a susceptibility report. Even something as simple as placing a more desirable option first on a susceptibility list may increase the likelihood that a prescriber will choose that option. All of these kinds of communication are important, and so is ensuring that prescribers have easy access to laboratory staff if they have additional questions. We need to keep the lines of communication open if we want our interventions to be effective.

There’s a big “know-do” gap between what we know about appropriate use of antibiotics and what is actually implemented in everyday practice. Education is a key component to help change practice, but it won’t be effective on its own; we also have to recognize the central role that behavior plays in prescribing practices. Fear, often related to uncertainty of diagnosis, or complacency, can lead to unnecessary antimicrobial use. In our opinion, the biggest challenge will be implementing effective antimicrobial stewardship interventions that address these behavioral aspects of prescribing in a variety of settings to a variety of audiences – and “nudging” will likely continue to play an important role.

Our study reflects the need for a truly collaborative, multidisciplinary approach to intervention. There are too many complexities in modern healthcare delivery, and in the emergence of resistance, for laboratories to make assumptions about how clinicians will best interpret and apply our results. As laboratory physicians, we need to actively seek end-user input to formulate individual antibiograms that are clear and useful. Part of that, too, is ensuring that laboratories receive the right specimens at the right times – because as we develop increasingly diverse and sensitive tests, the old adage of “garbage in, garbage out” is truer than ever. Selective antibiotic reporting can target and improve the post-analytical phase of the total testing process, but if we want to make a real difference to stewardship, we need to combine it with education, communication, and interventions at every stage of the process.