From Wiki Journal Club
Jump to navigation Jump to search
Copaescu AM, et al. "Efficacy of a Clinical Decision Rule to Enable Direct Oral Challenge in Patients With Low-Risk Penicillin Allergy". JAMA Intern Med. 2023. :e232986.

Clinical Question

In adult patients at low risk of true penicillin allergy, is direct oral challenge safe and effective as compared with standard-of-care skin testing followed by oral challenge?

Bottom Line

Those patients at low risk of true penicillin allergy, direct oral challenge was as safe and effective as two-stage skin test.

Major Points

When interrogated, there is a portion of patient reported penicillin allergies are descried as "I was was told I was allergic" with no further explanation. Clinicians commonly do not question these allergies when the decision to start antimicrobials is made, thus forcing them to choose agents that may not be the first choice or first line. Removal of spurious allergy labels is a key part of antimicrobial stewardship. [1] [2] In many areas the gold-standard for testing is a 2-step intradermal skin test followed by drug challenge, but this costly and time consuming. Regardless of skin-test results, a challenge exposure is still required to "prove" and then support de-listing of the allergy.

In this trial, the investigators challenged that the in patients at low risk of allergy, as defined as a PEN-FAST score <3, that skipping the skin test and going straight to an oral challenge was as safe and effective as the gold standard test. They enrolled 382 adult outpatients, 190 to direct oral challenge and 192 as control with the standard of care, from 9 centres in Canada, America, and Australia. In their primary outcome of a positive immune-mediated reaction within 1 hour, they found no difference with 0.5% of subjectives having an unfortunate positive result. There was no statistical difference identified in any of the sub-groups that they analyzed nor in rates of adverse effects observed. There are some criticisms with this trial, one being that the only included PEN-FAST score <3 , and the vase majority were 0-1, so the slightly higher scores are unclear of their risk. Next Patient observation time was not standardized in the clinics and most were observed for 1-2 hours but the time to adverse event was 2-3 times longer than this. The applicability of this approach may also be impacted if it is tried to be employed in a busy setting such as an Internal Medicine clinic or Emergency Department. Finally the majority of patients were Caucasian and recruited from Canada and Australia so the generalizability may be impacted. Overall this is a good study that demonstrates patients at low risk of a true allergy may be challenges with a penicillin may be done safely and effectively without a skin-test.


As of August 2023, no guidelines have been published that reflect the results of this trial.


  • Multicenter, parallel-group, randomized, open-label, non-inferiority trial
  • N=382
    • Direct Oral Challenge (n=190)
    • Two-stage skin test (n=192)
  • Setting: 9 centres (3 Canadian, 3 American, and 3 Australian)
  • Enrollment: 18 June 2021 to 2 Dec 2022
  • Analysis: non-inferior, margin 5%
  • Primary Outcome: positive immune-mediated reaction (within 1 hour)


Inclusion Criteria

  • Self reported penicillin allergy
  • Outpatient
  • ≥ 18 years old
  • low risk (PEN-FAST[3] score <3)

Exclusion Criteria

  • anaphylaxis associated with any drug
  • known history of chronic spontaneous urticaria or mast cell disease
  • history of non–IgE-mediated severe reactions (e.g. severe delayed organ or skin reactions)
  • self-reported allergic reaction was inconsistent with an allergy and compatible with adverse effects such as nausea or headache
  • pregnancy
  • receiving current antihistamine treatment
  • receiving > 200mg hydrocortisone equivalent daily

Baseline Characteristics

Oral Challenge Group displayed

  • Demographics: median age 52, 62% female
  • Atopy: 19% asthma, 37% allergic rhinitis, 13% atopic dermatitis, 4% idiopathic urticaria/angioedema, 1% history of any anaphylaxis, 20% other antibiotic/antifungal allergy, 7% family allergy history
  • Coexisting conditions: median Charlson Comorbidity Index 1, 20% immunocompromised
  • Risk, PEN-FAST core: 42% zero, 52% one, 6% two
  • Allergy label: 78% unspecific penicillin, 2% PenVK, 1% PenG, 18% amoxicillin, 1% ampicillin, 1% amox/clav, 1% flucloxacillin
  • Allergy description: 60% childhood reaction, 13% immediate (<2h) reaction
  • Antibiotics safely tolerated since index reaction: 25% cephalosporins, 14% macrolides, 6% fluoroquinolones, 59% unknown


  • Oral challenge with amoxicillin or penicillin at the lowest therapeutic dose available
  • confirmatory 2-step skin test followed by oral challenge


Comparisons are oral challenge group vs. confirmatory skin test group.

Primary Outcomes

positive oral penicillin challenge consistent with an immune-mediated reaction
0.5% vs. 0.5% (risk ratio 1.02, 90% CI 0.10 to 10.34)

Secondary Outcomes

446 of 643 screened patients eligible (69%, 95% CI 66%-73%)
Per-protocol completion
overall 92% (95% CI 87%-94%)
Allergy Label removal
99.5% vs. 97.9% (risk difference 1.57, 95% CI -0.72 to 3.86) [ 4 patients in control had positive skin test and were excluded)

Subgroup Analysis

Similar across both interventions sub-groups including PEN-FAST score, immediate/delayed reaction, immunocompromised status, age, sex, or dose of oral challenge.

Adverse Events

Any adverse event within 5 days of intervention
0.11% vs. 0.11% (risk ratio 0.97, 95% CI 0.54-1.73)
immune-mediated events
0.05% vs. 0.05% (risk difference-0.45%, 95% CI -4.87-3.96)
time to adverse event, median
4 hours vs. 6 hours
delayed diffuse rash or urticaria, > 1 hour after oral penicillin challenge
3.2% vs. 1.6% (risk difference 1.63, 95% CI -1.46 to 4.72)


  • Majority of patients recruited from two sites: McGill University Health Centre (Canada, 36%) and Austin Health (Australia, 46%)
  • Unclear risk for PEN-FAST score 3, only included scores of 0-2
    • only 6% of participants had PEN-FAST score of 2
  • Patients were observed for 1-2 hours but the majority of adverse events occurred after this window
  • Patients with history of anaphylaxis to any drug were excluded


  • unrestricted grant from the Austin Medical Research Foundation

Further Reading