Geisinger Antibiotic Allergy Pilot Program: Assess and Address

Overview

This is a prospective non-inferiority study to evaluate penicillin allergy history in patients with reported penicillin allergy, who require penicillin or penicillin-derivative antibiotic during inpatient admission using a focused questionnaire. A simplified scoring system will be assigned to patient responses, and the total score will be utilized to identify low-risk patients that have a minimal risk of allergic reactions on exposure to penicillin or its derivative. Patients determined to have low risk based on this questionnaire will be offered a test dose (graded challenge) of amoxicillin in a supervised setting, and if they tolerate it, penicillin allergy label will be removed from patient's chart. We hypothesize that at least 95% of low-risk patients will successfully pass the graded amoxicillin challenge so the penicillin allergy label can be removed from their charts. A proportion as low as 0.85 would be a good clinical outcome and considered non-inferior to the expected proportion of 0.95.

In low-risk patients, as determined by the GAAP questionnaire, the investigator will discuss the option of graded challenge. This will involve initially administering 1/5th (56 mg) of the goal dose of amoxicillin (256 mg) by oral route following the procedure outlined below. The total goal dose to be administered is amoxicillin 256 mg (3.2 ml of 400 mg/5 ml). Patients will be monitored for 30 minutes after administration of the first amoxicillin dose. If no adverse/allergic symptoms (as listed below) are noted, the rest of the amoxicillin dose will be administered, and the patient will be monitored for an additional 90 minutes. The amoxicillin challenge will be performed on a monitored bed with telemetry and appropriate nursing coverage. The presence of bedside Benadryl 50 mg, Epinephrine (0.3 mg intramuscular), systemic steroids, and inhalational albuterol should be ensured. Glucagon will be added to the anaphylaxis order set to be used in patients with anaphylaxis who are on the beta-adrenergic blocker if needed throughout the duration of the study. However, no adverse event associated with the interference with anaphylaxis treatment due to the use of beta-adrenergic blocker has been reported in literature on amoxicillin graded challenge. Also, no prior study done on the graded challenge excludes the patients on beta-adrenergic blocker because the patients deemed appropriate for the graded challenge are not likely to develop anaphylaxis. The benefit in de-labeling penicillin allergy in this group far exceeds the risk of the theoretical but unreported adverse event. We will refer to EPIC order set "Anaphylaxis" if any adverse allergic reactions occur and medications are already available for use on each unit/floor to combat allergic reactions as standard of care. The below steps were taken to design and modify the GAAP questionnaire to optimize the validity for use in this study to identify risk of participating in the graded challenge : 1. Conduct a Literature Review: Our goal is to ensure the questions align with the intended use of the questionnaire, which is to safely stratify the hospitalized patients with history of penicillin or penicillin-derivative allergy to make decision on amoxicillin graded challenge during the same hospitalization. Some of the questions have been used in the reviewed studies in different formats. 2. Develop Items: Each item in the questionnaire is developed to ensure that the users (i.e., the providers who use the questionnaires to interview the patients) and the patients understand them clearly and the provider answers with no or minimal error. We removed any ambiguous and technical terms that may require the assistance of allergy/immunology specialty so that any general provider with basic medical training can understand each item. 3. Conduct Expert Validation: We have had three allergy/immunology specialists establish that individual items are relevant to the construct being measured and that the key items have not been omitted. This systemic review improved the overall quality and representativeness of the scoring system in the questionnaire, which is the unique element of this questionnaire. 4. Conduct Cognitive Interviews: Once the questionnaire was built and reviewed by the experts, we asked several general providers (i.e., house staff, internist, etc.) and patients to assess how prospective participants interpret the items in the questionnaire. 5. Conduct Pilot Testing: A pilot test was done and reported in the protocol that has been submitted for your review previously. The purpose of the pilot test was to 1) identify items that may not be functioning in the intended way, 2) ensure the clarity and easiness of each item for prospective participants to understand, and 3) support the safety in using the questionnaire in deciding on amoxicillin graded challenge. Procedure for the Graded Challenge: 1. Obtain baseline vitals (heart rate, O2 saturation, blood pressure and respiratory rate). 2. Administer 0.7 ml (56 mg) of amoxicillin 400 mg/ 5ml orally followed by 30 minutes of observation for allergy symptoms and other adverse reactions (e.g., itching, hives, swelling, coughing, wheezing, throat tightness, difficulty breathing, abdominal pain, vomiting, lightheadedness, hypotension, low oxygen saturation, tachycardia) 3. If no allergy or adverse symptoms are noted, administer 2.5 ml (200 mg) of amoxicillin followed by 90 minutes of observation. 4. Interpretation 1. If a patient does not develop significant adverse symptoms after the challenge, he/she does not have a serious life threatening immunoglobulin E (IgE)-type allergy to penicillin and penicillin-derivatives and can be administered with penicillin-based antibiotics during the admission. Such patients should have their penicillin allergy label removed from the chart. This will be completed by the GAAP Team Member who administered the questionnaire by logging into EPIC and providing appropriate documentation in the allergy section of the chart. The provider will also click on the drug under the allergies and remove it from the list. 2. This test does not rule out delayed onset drug reactions. Patients can develop delayed-onset symptoms, but typically the symptoms are not serious. In case a patient develops a delayed onset rash or other adverse symptoms, discontinue the penicillin/beta-lactam-based antibiotic and re-label the patient as penicillin allergic. Such patients can be referred for outpatient penicillin allergy evaluation in Allergy and Immunology. Patients will be advised to contact the investigators for such delayed reactions with the informed consent. 3. If the patient develops convincing signs and/or symptoms consistent with an immediate reaction during or shortly after the graded challenge (within a few hours of receiving the full dose), no further penicillin or penicillin-derivatives should be given, and symptoms should be treated appropriately with Benadryl 25-50 mg for mild symptoms, albuterol via nebulizer for respiratory symptoms, and epinephrine for serious anaphylaxis. These patients should be labeled with IgE-mediated penicillin allergy. If they require penicillin and/or penicillin-derivatives in the future, it should be administered via a formal desensitization protocol. 4. Some patients can develop nonspecific symptoms during the procedure, which can mimic symptoms of true allergy. Most commonly, these include perioral tingling, pruritus without urticaria, throat and lip discomfort, headache, tachycardia, and nausea. These may be anxiety-related, and spending time with the patient explaining the safety of challenge procedures in advance may help to reduce the incidence of these nonspecific reactions. This is a prospective non-inferiority study. Patient characteristics will be summarized with means and standard deviations or medians and interquartile ranges for continuous variables and with frequency counts and proportions for categorical variables. The number and proportion of patients who complete the GAAP questionnaire and are classified as high and low risk will be described as well as the number and proportion of low-risk patients who pass the graded amoxicillin challenge. A one-sided proportion hypothesis test with a non-inferiority margin of 0.10 will be used to compare the observed proportion to a hypothesized proportion of 0.95. We hypothesize that 95% of low-risk patients can be de-labeled appropriately based on clinical reasoning and strong evidence from previous research. First, it was noted in a previous study that fewer than 1% of the population is truly allergic to penicillin with true IgE-mediated reactions. However, approximately 10% of all U.S. patients report having an allergic reaction to penicillin previously.15,16 It is also noteworthy to consider that approximately 80% of patients with IgE-mediated penicillin allergy lose their sensitivity after 10 years. Preliminary data collected to validate the GAAP questionnaire showed about a 0.93 success rate. This means that about 93% of the identified low-risk patients successfully passed the amoxicillin graded challenge. The lower bound of the 95% confidence interval for the estimated observed proportion must be greater than 0.85 to claim non-inferiority. Compared to current practice successfully de-labeling a proportion as small as 0.85 would reflect better quality treatment, better patient care, and cost savings. This would be a significant improvement over current practice where fewer than 10% of patients with documented penicillin allergies get de-labeled. We chose 0.85 as the non-inferiority margin because we believe a hypothesized fail rate of 5% and a fail rate as high as 15% are acceptable and not harmful to the patients for the following reasons: Studies have shown that performing a direct penicillin or amoxicillin challenge in low-risk patients is safe and effective method of de-labeling penicillin allergy. 8,9 In other words, life-threatening and/or clinically significant reactions are not expected in this patient group even if allergic reaction(s) occur.Understanding the minimal risk potentially associated with this protocol, the providers would be willing to sacrifice a failure rate as high as 15% in exchange for the potential benefits offered by this new process. Currently, there is no formal protocol or process to identify and de-label penicillin allergy in hospitalized low-risk patients in any Geisinger institutions. Only IRB-approved study staff will have access to data collected for this research. Electronic data will be stored on Geisinger's secure network. Any hard copy data will be secured in a locked area. A data analyst will create and maintain a database for project data collection and tracking in a secure Research Electronic Data Capture system (REDCap). This electronic data will be stored on Geisinger's secure network. The database will contain Protected Health Information (PHI) elements, such as name, date of birth, medical record number, and date of consent. Only IRB-approved study team members authorized by the PI will have access to this database and data collected for this research. These records will be kept indefinitely as communicated to patients and described in the informed consent form. No data will be transmitted externally. The following data, including relevant dates, will be collected: * Names * Medical record number * Date of birth/date of death * Allergy list in Epic, including types of reactions and any relevant allergy testing or referrals * Information relevant to current hospitalization, including location, medications administered, problem list entries, vital signs * Baseline demographic variables of patients (age, sex, ethnicity, comorbidities) * GAAP questionnaire results and score * Data related to graded challenge, including dosage of amoxicillin, adverse reaction(s) and subsequent outcomes