NEJM Evid. 2026 Mar;5(3):EVIDoa2500038. doi: 10.1056/EVIDoa2500038. Epub 2026 Feb 24.
ABSTRACT
BACKGROUND: We aimed to evaluate the feasibility of conducting a multicenter, double-blind, placebo-controlled trial of prolonged (i.e., multidose) compared with single-dose amoxicillin challenge in patients with a history of delayed penicillin allergy and to assess the frequency of immune-mediated adverse events.
METHODS: We conducted a parallel, double-blind, placebo-controlled multicenter randomized feasibility trial at four Australian hospitals. Adults who had a history of penicillin allergy of delayed-onset (2 or more hours after exposure) or unknown timing and had tolerated a single-dose oral challenge to amoxicillin were eligible to enroll. Participants were randomly assigned 1:1 using a permuted block design to either a 5-day amoxicillin challenge (intervention) or a placebo (control). The primary outcome measures were adherence to the intervention (N, % taking at least 80% of capsules), need for unblinding (N, %), and recruitment feasibility (N, % of eligible patients enrolled). The secondary outcome measures included positive oral challenges, defined as immune-mediated adverse events up to and including day 7.
RESULTS: We screened 302 adults and randomly assigned 120 participants to intervention (n=60) or control (n=60). The mean (standard deviation) age was 49.4 years (17.6). Allergy histories included 42% (50/120) delayed reactions and 58% (70/120) of unknown timing. Adherence to the intervention was 83% (100/120; 95% confidence interval, 75 to 90) (88% intervention and 78% control). No patients were unblinded and recruitment feasibility was 71% (120/169). The number of positive oral challenges was 7 of 60 (12%) in the intervention group and 3 of 60 (5%) in the control group (risk difference, 6.7 percentage points; 95% confidence interval, -3.2 to 16.5). One serious adverse event (1/60, 1.7%, severe cutaneous adverse reaction) was observed in the control group.
CONCLUSIONS: Conducting a double-blind, placebo-controlled randomized trial of single-dose versus multidose oral amoxicillin challenge was feasible. A fully powered efficacy trial is needed to assess between-group differences in immune-mediated adverse events. (Funded by Austin Health and others; trial registration: ANZCTR — ACTRN12623001242617, registered on November 30, 2023.).
PMID:41733406 | DOI:10.1056/EVIDoa2500038
