Background: Over 32 million patients in the United States have a penicillin (PCN) allergy label in their medical chart.1 Studies have demonstrated that 15-20% of hospitalized patients are labeled PCN allergic. However, IgE-mediated PCN allergy wanes over time and up to 80% of patients with a listed PCN allergy can tolerate PCN after 10 years. 2 In short, most PCN allergy labels are inaccurate. The previous standard of practice to delabel (remove a PCN allergy) was skin testing. This is labor-intensive, expensive, and conducted in resource-rich areas.4, Studies have shown that PCN delabeling can be conducted in a more efficient, cost-effective, and safe manner via direct oral challenges. 6-10 The 2022 Drug Allergy Practice Parameter guidelines recommend direct oral challenges in patients deemed to have low-risk PCN allergy.11 The purpose of our quality improvement project was to increase the rate of delabeling low-risk PCN allergies in patients admitted to all hospital medicine services from 0.1% to 5% in one year.
Methods: We implemented a protocol in February 2024 directing primary hospitalist clinicians to delabel PCN allergies among low-risk hospitalized patients. This protocol was designed through a clinical pathway and integrated into our electronic health record (EHR). The pathway provides information on how to risk-stratify patients via a validated tool known as PEN-FAST.12 If a patient has no-risk features, they can be directly delabeled. If a patient meets low-risk criteria, the pathway will guide clinicians to consent patients to a direct oral challenge with amoxicillin. We integrated a pre-filled consent form, order set, and EHR smart phrases in the pathway. A medical trainee screened hospitalized medicine patients twice weekly via an EHR-generated report of hospitalized medicine patients with a listed PCN allergy. Afterward, clinicians received secure messages notifying their patient has a PCN allergy and encouraging pathway use to delabel if appropriate. Our primary outcome was the number of patients delabeled. Our secondary outcomes were adverse reactions, inappropriate delabeling, and change in antibiotic therapy.
Results: We have successfully delabeled 82 patients in the eight months post-intervention. 36 were directly delabeled and 46 were delabeled through a direct amoxicillin challenge. Antibiotic therapy changed for 28 patients, enabling a switch to the appropriate penicillin antibiotic for their infections. When evaluating for adverse effects, there have been no instances of anaphylaxis or serious adverse reactions to the PCN challenge, and four patients were inappropriately directly delabeled. Our end-of-year goal is to delabel 105 patients (5%), and we are on track to meet that goal.
Conclusions: Several studies have shown that a PCN allergy label is not benign and significantly impacts patient care. A PCN allergy label has been shown to increase hospital costs due to the use of alternative antibiotics, delay in antibiotic administration, and increased risk of hospital-acquired infections.3 Our current process of notifying clinicians is not an indefinite solution. Therefore, we are exploring ways to ensure sustainability such as an automated EHR notification and collaboration with inpatient pharmacy. Before our intervention, ten patients were delabled in the previous year. Our project illustrates the successful implementation of an inpatient PCN delabeling protocol into the daily hospitalist workflow. We anticipate this will have lasting impacts on antimicrobial stewardship.