Comparative Effectiveness, Safety, and Real-World Outcomes of a Nurse-Led, Protocol-Driven Penicillin Allergy Evaluation From the Hong Kong Drug Allergy Delabelling Initiative (HK-DADI)

Background: There is a high prevalence of unconfirmed penicillin allergy, which is associated with a multitude of adverse clinical outcomes. With the overwhelming burden of currently incorrect labels and the lack of allergy specialist services, new delabeling strategies are urgently needed.

Objective: To assess the effectiveness, safety, and real-world outcomes of a nurse-led, protocol-driven evaluation of penicillin allergy, the Hong Kong Drug Allergy Delabelling Initiative (HK-DADI).

Methods: Adult patients with suspected penicillin allergy were recruited into HK-DADI. Allergy and postdelabeling outcomes were retrospectively compared between patients evaluated via HK-DADI or traditional allergist evaluation.

Results: A total of 312 completed penicillin allergy evaluation: 84 (27%) and 228 (73%) via HK-DADI and traditional pathways, respectively. Overall, 280 penicillin allergies were delabeled (90%). The delabeling rate between HK-DADI and traditional pathways was similar (90% vs 89%; P = .796). Among patients of the HK-DADI pathway, the delabeling rate was significantly higher among low-risk (LR) compared with non-LR patients (97% vs 77%; P = .010). Skin tests did not add diagnostic value among LR patients. No patients developed severe or systemic reactions during the evaluation. Upon 6- to 12-month follow-up (median, 10 months), 123 patients experienced infective episodes (44%) and 63 used penicillins again after delabeling (23%). This proportion was significantly higher in patients who were delabeled via HK-DADI compared with the traditional pathway (32% vs 19%; P = .026).

Conclusions: The Hong Kong Drug Allergy Delabelling Initiative, a nurse-led, protocol-driven evaluation, was safe and effective in penicillin allergy delabeling. It led to an even higher rate of future penicillin use after delabeling and mitigated the need for unnecessary skin testing among LR patients.