Objective To evaluate the use of a questionnaire-based decision-making algorithm to triage children with reported antibiotic allergies to proceed directly to an oral provocation challenge.

Design Cohort study.

Setting Children aged 2–16 years attending paediatric emergency department over 1 year (1 June 2018 till 31 May 2019) or identified from four primary care centres in Sheffield with a recorded antibiotic allergy and no previous testing.

Participants 313 children with 325 recorded antibiotic allergies.

Exposure Clinical decision-making algorithm used to either exclude, directly delabel or stratify children to oral antibiotic challenge in outpatient department or primary care practice.

Main outcome measures To assess the safety of using the questionnaire-based algorithm for proceeding to a direct oral provocation challenge.

The secondary outcomes were to look for associations and predictive factors in positive challenges and to assess parent/carer acceptability of the service by using Likert Scale.

Results Successful contact was made with 200 children, of which 153 children could be evaluated based on inclusion criteria, engagement and availability of medical records.

15 children were directly delabelled based on history and records. 138 children underwent challenges in outpatient and primary care. 6% of challenges were reactive with a mild, delayed reaction. Overall, a delabelling rate of 91% was achieved. There were no clear predictors for a positive challenge.

Conclusion Our questionnaire-based algorithm for stratifying children with antibiotic allergies to proceed directly to an oral outpatient or primary care challenge was found to be safe, feasible and acceptable.