To evaluate the cost-effectiveness of implementing a penicillin allergy assessment pathway (PAAP) versus usual care within the NHS.

A decision tree analysis over a 5-year time-period, informed by a randomised controlled trial (RCT) of PAAP and systematic review. Value of information analysis was also conducted to estimate the value of conducting a new trial.

Model inputs were informed by the ALABAMA RCT participants included in the primary analysis, 811 adults with penicillin allergy labels and recent antibiotic prescriptions, and data from published literature.

Participants in the ALABAMA trial included in the primary analysis: PAAP (n=401) and usual care (n=410).

Costs are presented in GBP (£) at 2022–2023 prices, quality-adjusted life years (QALYs), incremental cost-effectiveness ratio, incremental net monetary benefit (INMB), the probability of cost-effectiveness at the £20,000 and £30,000 per QALY threshold, and the cost effectiveness of a new follow-on trial.

PAAP had incremental costs of £–83 (probability of cost saving 47.5%) and incremental QALYs of 0.036 (probability of positive benefits 47.5%). The INMBs (probability of cost-effectiveness) were £806 (48%) and £1167 (48%) under the decision thresholds of £20,000 and £30,000 per QALY, respectively. PAAP was more cost-effective among females, people aged >65 years, and more frequent antibiotic users. A new follow-on trial involving 1267 participants was estimated to cost £2.4 million and, by reducing uncertainty in the evidence, would avoid £19.6 million in costs of incorrect management decisions for eligible patients over the next 10 years.

The PAAP was considered cost-effective, but significant uncertainty remained. Future trials with adequate power and longer follow-up are needed to determine the most cost-effective models for penicillin allergy testing.

ISRCTN20579216.

To develop a behavioural intervention package to support non-allergist healthcare workers (HCWs) to remove incorrect penA records from medical and surgical adult inpatients. This paper describes the development of the penicillin allergy de-labelling (PADL) intervention and the implementation intervention that will support non-allergist-delivered PADL.

We combined evidence-based, theory-based and person-based approaches. Qualitative research with healthcare professionals and patients explored barriers and enablers to implementation of the proposed PADL pathway. Key intervention design objectives and the key features of the implementation intervention required to achieve each objective were then developed and captured as guiding principles. We produced a logic model, integrating the theoretical domains framework to identify the behavioural influences on PADL and the behaviour change wheel to show how the implementation intervention is hypothesised to address the target behaviours. The implementation intervention package was then reviewed by stakeholders and topic experts for further refinement and optimisation. Finally, we outline how the implementation intervention will be evaluated.

Single-centre District General Hospital in the SW England servicing a rural community of 575 000 people without local allergy services.

HCWs reported PADL needed to be structured, standardised, evidence based and supported by hospital approved guidelines with easy to access patient information leaflets, supported by a sustained programme of education and training with named PADL leaders and visible PADL champions. Patients wanted a good explanation of the benefits and risks of testing and the benefits of having their ‘penA’ record removed. The identified HCW target behaviours were: taking a penA allergy focused history and to risk assess the patient’s penA history; to then either de-label the patient on history alone (direct de-label; DDL) or prescribe a direct oral challenge (DOC) dose; to perform baseline and post-test observations and counsel the patient on the risks of penA records and on the risks and the benefits of PADL. We identified barriers to target behaviours that we considered both important and modifiable, which included: lack of confidence in taking a penA focused history, PADL not viewed as a priority, low confidence with differentiating low-risk and high-risk penA histories, concerns about the safety of DOC, a requirement for senior support for nurses to deliver the observations and senior support for the other HCWs to deliver PADL, access to an expert for advice when required, a lack of PADL champions to promote PADL, and PADL not being supported by the organisation. The identified patient target behaviours were acceptance of the opportunity to be de-labelled via either DDL or DOC and willingness to take penicillin when prescribed. We developed intervention components to target the HCW and patient target behaviours which included: Education, expert advice made available from Infection specialists, a named PADL champion, hospital endorsed PADL guideline with necessary tools to enable PADL and patient information leaflets. The implementation intervention was further optimised through workshops with PADL researchers and stakeholders. The Consolidated Framework for Implementation Research outcome addendum was used to define both implementation intervention and PADL intervention outcomes.

We have developed a theory-based and stakeholder-developed implementation intervention to support inpatient PADL delivered by a multiprofession workforce. The intervention will be tested in a single hospital and scalability explored.