567 - Penicillin allergy de-labeling: Adaptation of risk stratification tool for patients and families

Poster Number: 567
Publication Number: 567.315

Background: Penicillin allergy is reported in 10% of the population, however, over 90% of patients are deemed non-allergic upon allergist assessment. A validated mobile penicillin allergy risk assessment tool was previously developed to help healthcare providers identify patients at low risk of penicillin allergy and de-label them.

Objective: The goal of this quality improvement project is to validate a patient-driven version of the assessment tool.

Design/Methods: Patients 6 months to 18 years old and pregnant women referred to the institution’s allergy clinics for penicillin allergy assessment were invited to use the patient tool to complete a self-assessment prior to their appointment, resulting in the assignment of a risk category: allergic, high risk, low risk, or not allergic. At their appointment, the allergist conducted an assessment with the standardized validated tool. The risk categories determined using the patient tool were compared against the allergist assessment. Patients were provided satisfaction surveys.

Results: The allergist assessment and patient tool demonstrated agreement in 57/84 (67.9%, 95% CI [56.7%,77.4%]) assessments, intra-class correlation (ICC) = 0.618, p < 0.001. Of these, 45/57 (78.9%) patients were stratified as low risk or not allergic. In 22/84 (26.2%) assessments, the patient tool determined a higher risk category than the allergist, primarily due to differences in patients’ perceived timing and description of symptoms. Only 5/84 (6.0%) patients were placed in a lower risk category using the patient tool compared to the validated tool. Of these, one patient responded to the self-assessment accounting for a mild reaction, whereas the allergist assessment accounted for a second, more severe reaction. The other 4/5 patients mistakenly answered that they had never taken penicillin so were stratified as not allergic on the patient tool; however, they had previously taken penicillin and were found at risk of possible allergy by the allergist. Of those who responded to the satisfaction survey, 53/62 (85.5%) agreed that the patient tool was easy to use.

Conclusion(s): The patient tool demonstrates good agreement in determining penicillin allergy risk, generally aligning with the validated tool and offering potential as a method of empowering patients to assess their risk and advocate for their care. For a small minority of patients, the patient tool determined a lower risk category than the allergist assessment. To improve safety, agreement, and accessibility, iterative cycles of assessment and changes to the patient tool structure will be applied.