Incidence of interruptive penicillin allergy alerts in patients with previously documented beta-lactam exposure: Potential for leveraging the electronic health record to identify erroneous allergies.

BACKGROUND Approximately 10% of patients report allergies to penicillin, yet >90% of these allergies are not clinically significant. Patients reporting penicillin allergies are often treated with second-line, non-β-lactam antibiotics that are typically broader spectrum and more toxic. Orders for β-lactam antibiotics for these patients trigger interruptive alerts, even when there is electronic health record (EHR) data indicating prior β-lactam exposure. OBJECTIVE To describe the rate that interruptive penicillin allergy alerts display for patients who have previously had a β-lactam exposure. DESIGN Retrospective EHR review from January 2013 through June 2018. SETTING A nonprofit health system including 1 large tertiary-care medical center, a smaller associated hospital, 2 emergency departments, and ˜250 outpatient clinics. PARTICIPANTS All patients with EHR-documented of penicillin allergies. METHODS We examined interruptive penicillin allergy alerts and identified the number and percentage of alerts that display for patients with a prior administration of a penicillin class or other β-lactam antibiotic. RESULTS Of 115,081 allergy alerts that displayed during the study period, 8% were displayed for patients who had an inpatient administration of a penicillin antibiotic after the allergy was noted, and 49% were displayed for patients with a prior inpatient administration of any β-lactam. CONCLUSIONS Many interruptive penicillin allergy alerts display for patients who would likely tolerate a penicillin, and half of all alerts display for patients who would likely tolerate another β-lactam.