Review of Pharmacist Driven Penicillin Allergy Assessments and Skin Testing: A Multi-Center Case-Series

Abstract

Objective: To describe the impact of pharmacy driven penicillin allergy assessments on de-labeling penicillin allergies and antibiotic streamlining opportunities for hospitalized patients. Design: Multi-center, retrospective case-series study. Setting: A health system of 4 non-teaching hospitals. Participants: Patients aged 18 years and older with a physician order for a pharmacist penicillin allergy assessment. Exclusion criteria consisted of patients with anaphylaxis or a type II penicillin allergy, anaphylaxis of any cause within 4 weeks, refusal of penicillin allergy skin test (PAST), antihistamine use within 24 hours, penicillin intolerance, immunosuppression or immunosuppressive medications, or skin conditions that could interfere with PAST. Interventions: The primary endpoint evaluated the number of de-labeled penicillin allergies after pharmacists provided penicillin allergy assessments. Secondary endpoints evaluated the percent of patients with antibiotics deescalated to beta-lactam antibiotics and classification of notable interventions made by pharmacists. Measurements and Main Results: There were 35 patients who met inclusion criteria. Twenty-four patients underwent both penicillin allergy skin testing and oral (PO) amoxicillin challenge. Five patients had allergies de-labeled only after a pharmacist interview. Four patients received only the PO amoxicillin challenge and 2 patients received only PAST. Penicillin allergies were de-labeled from the electronic health record (EHR) in 31 (89%) patients despite all testing negative for a penicillin allergy from PAST or a PO amoxicillin challenge. Four patients had the allergy re-added to the chart on subsequent admissions. No patients experienced a reaction from PAST, PO amoxicillin challenge, or subsequent beta-lactam antibiotics. Twenty-eight (80%) patients had their antibiotic therapy changed as a result of the allergy assessment. Seventeen patients were de-escalated onto beta-lactam antibiotics and aztreonam was stopped in 6 patients. Conclusion: Results from this study suggests that pharmacists expanding their scope of practice with PAST is a safe and effective allergy de-labeling tool. Pharmacist-driven penicillin allergy assessments could provide antibiotic cost savings and avoid aztreonam use. The study supports the need to emphasize education for patients and caretakers regarding allergy testing results to avoid relabeling in future hospital visits.

Introduction

Ten percent of hospitalized patients report a penicillin allergy, however recent studies indicate that 98% of these patients are not acutely hypersensitive. ¹ Additionally, penicillin allergies have been shown to disappear in over 80% of patients after 10 years. Beta-lactams are the treatment of choice for many infections due to their narrow spectrum, low rate of side effects, and affordability. Erroneous penicillin allergies are associated with many deleterious outcomes, such as increased healthcare costs, multi-drug resistant infections, and the use of non-preferred beta lactams.^2-4 Less preferred regimens cost 1.82 to 2.58 fold higher than first-line beta-lactam antibiotics for patients with a penicillin allergy. Additionally, there is a higher prevalence of Clostridioides difficile, methicillin-resistant Staphylococcus aureus, and vancomycin-resistant Enterococcus. ⁴

Penicillin allergy skin testing (PAST) has been recommended by multiple organizations to optimize antibiotics for patients with beta-lactam allergies, including the Infectious Diseases Society of America, American Academy of Allergy Asthma and Immunology (AAAI), and American College of Obstetricians and Gynecologists (ACOG).^5-7 Previous studies have shown that PAST and by mouth (PO) amoxicillin challenge provides a safe and accurate way to clarify penicillin allergies with a 97% to 99% negative predictive value. ⁸

Penicillin allergies are classified as mediated by either immunoglobulin-E (IgE), immunoglobulin-G (IgG), or immunoglobulin-M (IgM). Type I allergies are IgE-mediated with reactions including itching, rash, anaphylaxis, and angioedema. IgE-mediated reactions may be lost over time due to the loss of IgE antibodies. Type II, IgG reactions manifest as Stevens-Johnson syndrome and toxic epidermal necrolysis. Type III, IgM-mediated reactions consist of serum sickness, glomerulonephritis, and vasculitis. PAST is only recommended for Type I penicillin allergies. ⁹

The Center for Disease Control and Prevention has recommended that all patients reporting a penicillin allergy should be evaluated and, if appropriate, tested for the allergy instead of being placed on an alternative broad-spectrum agent. ⁸ Traditionally, penicillin allergy skin testing was done exclusively by allergy specialists, however, pharmacists have begun to provide this testing in the United States both in collaboration with allergists and as a stand-alone service.^10-16 The state of Florida required a collaborative practice agreement and was approved on a case by case basis. ¹⁰ Our institution was granted approval for pharmacists to perform PAST after petitioning the Florida Board of Pharmacy (FL BOP) in March 2019. Removing penicillin allergies from the electronic health record (EHR) has been shown to expand antibiotic options, expedite discharge, and increase cost savings. ¹⁷

Materials and Methods

This multi-center, retrospective case series was conducted at a health system of 4 non-teaching hospitals from April 2019 through March 2021. Patients were stratified into 4 groups—allergy clarified after pharmacist interview, PAST only, amoxicillin PO challenge only, and PAST with amoxicillin PO challenge. Patient inclusion criteria consisted of the following: 18 years or older, documented type I penicillin allergy, physician order for a pharmacist to assess a penicillin allergy (see Figure 1), and a beta-lactam as the preferred antibiotic. Patients were excluded due to declining PAST, any non-type I hypersensitivity or intolerance to a beta-lactam, any anaphylaxis in the previous 4 weeks, or history of a skin condition that could interfere with reading results. The study was approved by the institution’s Ethics Committee. Data collected included age, gender, infection type, antibiotics used, therapy adjustments made post allergy clarification, allergic reaction to PAST or PO amoxicillin challenge, and readmission within 30 days due to infectious causes.

Figure 1.

Inclusion and exclusion criteria.

After FL BOP approval, the penicillin allergy assessment was created. Pharmacists received training on allergy assessment and skin testing. Prior to skin testing, the outpatient medication history was reviewed and the patient was interviewed. If the patient could not confirm they tolerated a beta-lactam previously, education was provided regarding PAST. If the patient was agreeable, the pharmacist advanced to performing penicillin skin testing and an amoxicillin PO challenge.

PAST consisted of a pharmacist performing a skin scratch test with histamine, benzylpenicilloyl polylysine (Pre-PEN), penicillin G 10,000 units/mL, and normal saline (see Figure 2). If the scratch test measured negative for a reaction after 15 minutes, the pharmacist would proceed with intradermal injections of the above previously listed agents, excluding histamine. The intradermal test was evaluated after 20 minutes and if negative, the pharmacist proceeded with the amoxicillin 250 mg PO challenge. Afterward, the patient’s nurse monitored the patient for at least an hour for any signs of allergic reaction.

Figure 2.

Penicillin allergy consult workflow.

A direct to PO amoxicillin challenge was recommended to the ordering physician without PAST if the patient refused skin testing but was agreeable to the test dose of amoxicillin, or was a low risk of a true penicillin allergy. Low risk was defined as a patient with a penicillin allergy history of a mild cutaneous reaction (eg, rash), or an unknown allergy that occurred greater than 10 years ago.

After having negative PAST and/or an amoxicillin PO challenge, patients were provided an allergy result card to keep and counseled on their result. The pharmacist would then remove the penicillin allergy from the patient’s EHR, complete a progress note, and contact the patient’s outpatient pharmacy with updated allergy information.

The primacy outcome evaluated the incidence of de-labeled penicillin allergies after pharmacist-driven penicillin allergy assessments. De-labeling was defined as removal of penicillin allergy from the permanent health record. The EHR was reviewed for allergy de-labeling versus inappropriate re-labeling of penicillin allergies on subsequent admissions. Secondary endpoints included classification of notable opportunities and interventions, number of antibiotic de-escalation or therapy adjustments, discontinuation of aztreonam, discharge after penicillin allergy assessment, cost savings, and infection-related readmission rates. Due to the lack of a comparator group, descriptive statistics were used to analyze the data.

Results

A total of 53 patient charts were reviewed with physician orders for pharmacists to evaluate a patient’s penicillin allergy, in which 35 patients met inclusion criteria. Despite all 35 (100%) patients testing negative for penicillin allergies, only 31 (89%) of patients met the primary outcome and had their allergies de-labeled from the EHR. Four patients had allergies re-labeled on subsequent admissions. There were no allergic reactions post PAST or subsequent administration of beta-lactam antibiotics. Eighteen (51%) patients included were female, and 17 (49%) were male. A breakdown of allergic reactions reported by patients was listed in Table 1. The secondary outcome of therapy adjustment occurred in 28 (80%) of patients of which 17 patients de-escalated to beta-lactams (See Table 2). No readmissions due to infectious causes within 30 days were observed.

Table 1.

Allergy Demographics.

Sex	Itching	Hives	Rash	Unknown
Male (n = 17)	3	2	7	5
Female (n = 18)	2	5	8	3

Table 2.

Therapy Changes & De-escalations.

Initial Medication	Therapy Change and De-escalation
Aztreonam (n = 6)	Cefepime
	Amoxicillin/clavulanate
Carbapenems (n = 3)	Amoxicillin/clavulanate
Cefepime (n = 8)	Ampicillin and ampicillin/sulbactam
	Amoxicillin and amoxicillin/clavulanate
Ceftriaxone ( = 3)	Cephalexin
	Ampicillin
	Penicillin G benzathine
Vancomycin (n = 2)	Cefazolin
	Ampicillin
Fluoroquinolones (n = 2)	Cefepime
	Ampicillin/sulbactam

One patient with chlamydia, suspected neurosyphilis, and a history of intravenous (IV) drug abuse, refused a lumbar puncture. This resulted in the need for 14 days of IV antibiotics in the absence of allergy clarification. After testing negative from PAST and amoxicillin PO challenge, the patient was transitioned from IV ceftriaxone to a one time dose of intramuscular (IM) penicillin G benzathine. Additionally, the patient was discharged 8 days early after the one time IM injection with scheduled follow up to their outpatient primary care provider. Eight patients were transitioned from IV antibiotics to PO amoxicillin ± clavulanate and discharged within 48 hours of pharmacist intervention. One of these patients had osteomyelitis and refused IV antibiotics. Aztreonam was discontinued in 6 patients, with an avoidance of 89 doses. An average of 4 days of aztreonam were avoided for each of these patients resulting in a total cost avoidance of $3,831.80.

Discussion

This study supports previous findings of the scarcity of true penicillin allergies with 100% of patients testing negative for penicillin allergy via PAST.^{11-13,15,18-20} Additionally, of the 28 patients with resultant therapy adjustment, 17 (61%) were transitioned to preferred beta-lactam antibiotics. Despite patient education after testing negative for a penicillin allergy, 4 patients were relabeled with a penicillin allergy in subsequent admissions. This may have been due to a patient’s misunderstanding of their allergy test results, a caregiver providing allergy information for a subsequent admission, or outdated paperwork provided during history taking during an admission. An opportunity for improvement included multiple touchpoints to communicate with patients before discharge that they are not allergic to penicillin antibiotics. This may include physician, pharmacist, and nursing touchpoints to ensure retention of education. Transitions of care evaluation with pharmacists post-discharge may also prevent allergy relabeling. It’s important to incorporate the patient’s caregiver or family in the education, as well as informing the patient’s outpatient providers.

A 2019 study by Englert et al conducted exclusively by pharmacists reviewed 22 patients in the inpatient setting. All patients tested negative for penicillin allergy via PAST. No IgE reactions were reported post PAST or subsequent beta-lactam antibiotic administration, which was similar to our study results. Fifteen of the 22 patients were transitioned to preferred beta-lactam antibiotics. This resulted in decreased utilization of both fluoroquinolone antibiotics and vancomycin.

Limitations of this study were its retrospective design and completion within a single health system. The PAST criteria were conservative and excluded penicillin allergies with any anaphylactic history. This limited the number of patients eligible for the assessment and contributed to a smaller sample size. There was no control arm as this was a descriptive case-series. The vast majority of consults came from 3 infectious diseases providers, so continued education of the service to other infectious diseases providers and other specialties may increase referrals, program expansion, and further antibiotic stewardship efforts.

Aztreonam was 2.6 times more expensive than piperacillin/tazobactam and 25 times more expensive than other oral beta-lactams. Pharmacist-driven penicillin allergy assessments may serve as a cost savings opportunity and provide options for better anti-pseudomonal coverage for hospitalized patients. Future directions for PAST in this health system includes broadening the service for patients with an anaphylaxis history for penicillin allergies more than 10 years ago.

Conclusion

Implementation of pharmacist-driven allergy assessment and PAST can effectively de-label penicillin allergies, optimize antibiotic streamlining, and provide cost savings. Measures should be evaluated to avoid inappropriate re-labeling of penicillin allergies on subsequent hospital visits.