Background
Penicillin allergy labels in medical records are common, often incorrect, and associated with increased antibiotic use and worse health outcomes. We aimed to establish whether a penicillin allergy assessment pathway initiated in primary care could safely improve use of penicillins.
Methods
ALABAMA was a multicentre, open-label, randomised pragmatic trial with embedded process and cost-effectiveness evaluations. Participants came from 51 UK general practices and testing took place at four UK hospital sites (Leeds Teaching Hospitals NHS Trust, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Cornwall Hospitals NHS Trust, and Bradford Teaching Hospitals NHS Foundation Trust). Eligible participants were aged 18 years or older, provided informed consent, had a record of penicillin allergy or sensitivity in their electronic medical records, had received an antibiotic prescription in the previous 24 months, and were outpatients at the time of recruitment. Participants were randomly assigned (1:1) by the research team to either a penicillin allergy assessment pathway or usual clinical care, by use of a secure, web-based system. The primary outcome was the proportion of participants who received at least one prescription for a penicillin for conditions for which a penicillin is first-line therapy, up to 12 months after random assignment. The original primary outcome was changed on July 12, 2023, from treatment response failure to penicillin prescribing due to slow recruitment caused by the COVID-19 pandemic. The primary analysis population was defined as all randomly assigned participants for whom outcome data were available. Safety was assessed in the as-treated population (ie, participants analysed by the intervention they received). The study was registered with ISRCTN, ISRCTN20579216, and ClinicalTrials.gov, NCT04108637, and is completed.
Findings
Between Sept 17, 2019, and Oct 9, 2023, 1616 participants expressed interest and 823 were enrolled and randomly allocated (411 to the penicillin allergy assessment pathway and 412 to usual clinical care). 401 penicillin allergy assessment pathway and 410 usual clinical care participants were included in the primary analysis. 584 (72%) of 811 patients were female and 227 (28%) were male, the mean age was 55 years (SD 15·6), 786 (97%) of 811 patients were White, and 13 (2%) were non-White. 72 (18%) of 401 participants in the penicillin allergy assessment pathway group and 14 (3%) of 410 participants in the usual clinical care group were prescribed at least one course of a penicillin for a condition for which it was first-line therapy during follow-up (adjusted relative risk 5·27, 95% CI 3·03 to 9·18; adjusted risk difference 14·21%, 9·92 to 18·49). 83 adverse events occurred in 73 participants in the 28 days after allergy testing; one event was severe and probably related to the intervention. In the as-treated population, 27 (7%) of 365 participants who received the penicillin allergy assessment pathway and 34 (8%) of 446 participants who received usual clinical care had at least one serious adverse event during the 1-year follow-up. There were no deaths related to the intervention.
Interpretation
Our data suggest that the penicillin allergy assessment pathway can increase prescription of narrow-spectrum penicillins with few signals of harm, indicating its potential in antibiotic stewardship.
Funding
National Institute for Health and Care Research.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)