Feasibility of a Centralized, Pharmacy-Led Penicillin Allergy Delabeling Program Free

A convenience sample of patients with a reported history of penicillin allergy admitted to our academic children’s hospital was identified by a central pharmacist. Patients with a penicillin allergy label were identified during the intake process by nurses, who then documented a penicillin allergy into the patient’s electronic medical record (EMR). We developed an internal EMR penicillin allergy dashboard visible to the pharmacy team that displayed all patients within the hospital with a penicillin allergy label. All patients on a nonpsychiatric or behavioral health team (including pediatric hospital medicine and surgery specialties) were evaluated for enrollment and eligible patients were randomly selected for risk stratification. Patients excluded from risk stratification included those who were pregnant, had a previous positive penicillin skin test result, were admitted with concern for allergic reaction or nonaccidental trauma, presented with unstable wheezing or a rash on admission, and those with antihistamine use in the previous 24 hours.

Once a patient was enrolled and selected for intervention, the pharmacist contacted the primary team and obtained permission to perform risk stratification and an oral challenge if the patient stratified as low-risk. The decision to perform risk stratification and an oral challenge was at the sole discretion of the primary team after consent was given by a parent or legal guardian. Patients were risk-stratified by using a previously developed and validated penicillin allergy-specific questionnaire (regardless of the need for antibiotics) (Supplemental Fig 3).^5,6  Those screened as low-risk underwent an amoxicillin oral challenge after informed consent was obtained from a parent or guardian. The amoxicillin oral challenge was performed by using an institutional standard-of-care protocol consisting of a 250 mg oral amoxicillin dose followed by 60 minutes of monitoring for reactions while on a cardiorespiratory monitor and pulse oximetry. Vital signs were obtained at 0, 30, and 60 minutes after the amoxicillin administration. An adult (≥18 years of age) other than the patient was required to be present in the room during the monitoring period. This could be a family member or hospital staff.

After the completion of each patient’s risk stratification and delabeling process, the Acceptability of Intervention Measure, Intervention Appropriateness Measure, and Feasibility of Intervention Measure surveys⁷  were distributed to health care personnel (pharmacists, nurses, nurse practitioners, house staff, and attending physicians) involved in the patient’s care that day. Measures were scored by using a 5-point Likert scale ranging from strongly disagree (1) to strongly agree (5) (Supplemental Fig 4).⁷  The study was approved by our institutional review board (201337, ClinicalTrials.gov Identifier: NCT04441021).

Of the 23 patients who screened as low risk, 20 underwent a penicillin allergy evaluation and an oral challenge. Of these, the penicillin allergy label was removed in 19 (95%) patients (Fig 1). The median age was 7 years (range 11 months–18 years). One 13-month-old patient had pruritus reported by her mother 1 hour into the oral challenge observation period. There were no vital sign changes associated with the parent-reported reaction, and medication intervention with antihistamines or epinephrine was not required. The patient’s original penicillin allergy label was due to a family history of reported penicillin allergy, but the patient had never received penicillin before the amoxicillin oral challenge, and suspicion for true amoxicillin IgE-mediated reaction was low. She was referred to the allergy clinic for skin testing after discharge but was lost to follow-up.

A total of 48 of 52 (92.3%) health care personnel who were approached completed the survey. Participants rated the risk stratification and delabeling favorably overall, with high ratings on all 3 implementation measures: acceptability (mean 4.55, ± standard deviation [STD] 0.65), appropriateness (mean 4.58, STD ± 0.6), and feasibility (mean 4.51, STD ± 0.73). Measures of acceptability, appropriateness, and feasibility remained high when stratified by health care worker type (Fig 2A) and provider type (Fig 2B). Comments from the survey are represented in Table 1 and revolved around timing, communication, clinical need, and required effort.

Given the high prevalence of inaccurate penicillin allergies and the significant benefits of removing such labels, improved approaches are needed to identify and safely delabel inappropriate allergy labels in children. In this study, we piloted a centralized, pharmacy-led penicillin allergy risk-stratification and delabeling program that was demonstrated to be safe and effective and perceived as acceptable, appropriate, and feasible by associated health care workers. Hospitalized patients provide an ideal point-of-care for allergy evaluation because hospitals are a location of frequent patient interaction with the health care system, provide close patient supervision, and often include sufficient downtime in which to perform allergy evaluations, regardless of the need for antibiotics.^8,9 

There are several aspects of this pilot intervention that were key to the success of this new delabeling process. Importantly, the pharmacy, hospital medicine, nursing, and allergy/immunology teams were intimately involved in the development of the study protocol. This multidisciplinary, unified approach across subspecialties was crucial in obtaining early buy-in for the project. Second, this intervention was pharmacist-led. A pharmacist identified patients with a penicillin allergy using an EMR dashboard, obtained consent from families, ordered medications, and provided documentation. All of these were performed in coordination with the primary team and often through routine interactions with the primary provider. Next, the use of a single dose, rather than a graded oral challenge requires less monitoring and is less of a burden to nurses and pharmacy staff. There was also face-to-face communication with nurses and families that included education on the benefits of delabeling and the likelihood of reaction related to the oral challenge. Specifically, discussion regarding the availability of medications and protocols for managing allergic reactions, as well as the likelihood of allergic reaction with or without the need for epinephrine or respiratory support, and the duration of monitoring was judged as important. As a result, as-needed orders for epinephrine, antihistamines, steroids, and β-agonists were added to the protocol and included for every patient before amoxicillin administration. Enhanced team communication also allowed real-time nursing staff feedback to optimally schedule oral challenges within daily workflows. Finally, before implementation, we engaged hospital leadership, including nursing, pharmacy, unit medical directors, clinical informatics, and hospital medicine to develop internal support for the process to improve patient care of children with drug allergies.

Findings from this project suggest that a centralized, pharmacy-led delabeling program can be successfully implemented within the workflow of a children's hospital. Additional study is needed to evaluate the scalability of this intervention at a hospital-wide level. Our intervention could be improved by leveraging the EMR to improve and automate patient identification, notification of allergy labels, delabeling orders, and documentation support through the use of clinical decision support tools to improve the efficiency of the delabeling process within hospital workflow. Limitations of this pilot demonstration include the number of patients evaluated as well as the academic hospital setting, which may limit the generalizability of our findings.

Appropriate delabeling of penicillin allergy in children will lower health care costs and improve health outcomes by lowering rates of treatment failure, multidrug-resistant infections, infection-related mortality, antibiotic-related drug–drug interactions, and adverse drug events.^1,10  These promising results provide support for future prospective studies of implementation strategies to most effectively and safely delabel patients at low risk for true allergy.