153 - Development of a Pathway for Removal of Inappropriate Penicillin Allergy Labels in Hospitalized Patients

Background: More than 90% of reported penicillin allergies are found inaccurate when formally assessed. These allergy labels lead to decreased utilization of first-line beta-lactam antibiotics, and adverse clinical outcomes. The objective of this study was to develop a multi-disciplinary approach to decrease inaccurate labeling among hospitalized patients with documented penicillin allergy.

Methods: A team of clinicians, pharmacists, and nurses utilized the DMAIC quality strategy to improve accuracy of penicillin allergy labeling. Allergic reactions were stratified to develop a penicillin allergy de-labeling algorithm (Figure 1). Admission to the intensive care unit (ICU) for anaphylaxis was defined as a balancing measure. We reviewed baseline data from patients with a documented penicillin allergy admitted to a single inpatient floor at Mayo Clinic, Rochester between June and October 2019. A cause and effect diagram was used to conduct a root cause analysis. The intervention was then applied to patients who reported penicillin allergy admitted to the same floor from November 2019 to January 2020. Study data were collected and basic descriptive statistics generated.Figure 1: Penicillin allergy delabeling algorithm


Results: 96 patients were included in the control group with mean age of 71 years (range 65-84 years) and 55% females. Breakdown of documented allergic reactions are represented in Figure 2. 58 (60%) received an antibiotic for a median duration of 1.5 days (IQR: 0 – 6). Of these, 7(12%) received penicillin-class antibiotics, and 41 (70.6%) received non-beta-lactam antibiotics. 2 (2%) of these patients were de-labeled without any penicillin skin tests. Detailed metrics of each PDSA cycle are shown in Table 1. During PDSA cycle 2, inaccurate penicillin documentation was removed in 9/19 (47.4%) of hospitalized patients. There were no ICU admissions for anaphylaxis.Figure 2: Graphic representation of proportion of type of documented allergic reactions to penicillin
Table 1: Metrics and outcomes at baseline and during successive PDSA cycles


Conclusion: Various factors contribute to penicillin allergy mislabeling. Our comprehensive algorithm addresses nuances of penicillin allergic reactions and increased accurate penicillin allergy labeling in 47.4% of the cases. Beta-lactam use also increased to 37% through our pilot project while maintaining patient safety. A multi-disciplinary and patient-centered approach aligned with institutional workflows is necessary to improve patient outcomes.