Categorical Risk Perception Drives Variability in Antibiotic Prescribing in the Emergency Department: A Mixed Methods Observational Study

Research Area: Antibiotic Resistance

The Question

This study, published in The Journal of General Internal Medicine, finds that clinicians who perceive the risks of antibiotics as negligible are more likely to prescribe them than those who consider the potential side-effects a concern and are more mindful of an effect on antibiotic resistance.

The study involved a survey of 69 emergency department clinicians about their perceptions of potential harms of antibiotic use and how they frame their decisions to prescribe antibiotics. The researchers then correlated clinicians’ survey responses with their actual antibiotic prescribing practices for patients with acute respiratory infections over 17 months, beginning in August 2014.

What we found

To interpret the prescribing behavior of clinicians, researchers used “fuzzy-trace theory,” an innovative approach to evaluating decision-making under risk, developed by study co-author Dr. Valerie Reyna. According to fuzzy-trace theory, providers prefer to rely on simple “gists” to make medical decisions, even though they are able to make decisions involving complex tradeoffs between benefits and harms. When deciding whether or not to prescribe an antibiotic for a particular patient, two gists dominated: (1) “why not take a risk,” in which providers perceived the risks from antibiotics as negligible compared to the patient’s current illness, and (2) “antibiotics may be harmful,” in which the provider agreed that the side effects of antibiotic therapy could be significant.

The clinicians who treated patients in this study followed a well-documented history of overprescribing in the United States: overall, antibiotics were prescribed for three-quarters of patients with diagnoses sometimes requiring them, according to guidelines; and antibiotics were prescribed for half the patients with diagnoses for which antibiotics are definitely not indicated.

But clinicians were true to their gists: those in the “why not take a risk” group were about fifty percent more likely to prescribe antibiotics than those in the “antibiotics may be harmful” group. This was especially true when patients were diagnosed with conditions that, according to prescribing guidelines, do not actually require antibiotics. Patients with pneumonia—where antibiotics are almost always indicated—were prescribed the drugs in almost all cases, regardless of gist.

Why it matters

The findings could inform the design of antibiotic stewardship strategies and educational tools that aim to reduce inappropriate prescribing and preserve the effectiveness of antibiotics. Until now, most stewardship interventions have assumed that clinicians are poorly informed about treatment guidelines and have focused on education to improve guideline adherence.  However, these educational interventions have only been modestly effective. This study may provide an explanation for why interventions of this type are not more effective.

Related Tools