Preintervention phase

Ensuring appropriate knowledge levels was identified as being critical to the DPC intervention adoption. For example, knowledge of the clinical history and origins of the PAL were considered an important foundation for HCP professionals implementing the DPC as well as patients’ willingness to accept the intervention. Patients typically reported little or no knowledge regarding penicillin allergies. A consistent message was, therefore, that patients were inclined to accept the label as valid, had not sought out any more information and were unaware of the potential for delabelling.

I was 6 years old. This was 34 years ago and since then everyone’s been scared to check. You were the first one to actually check (Site-C, Patient code: 3057)

Most patients were able to provide some information on the origin of their own allergy label. However, this was relayed with varying degrees of certainty, depending on whether the label was received as a child or adult, and their own or others’ (eg, parents) recollection of the circumstances or associated symptoms. This uncertainty was recognised as a challenge by HCPs in determining the risk of proceeding with delabelling, especially for those who may be less experienced in taking drug allergy histories.

The perceived impact of the PAL was low, reflecting a patient perception that equally effective alternatives are available. Patients who had personal experience of where alternative antibiotics were less effective had an increased awareness of the adverse impact of the label. The information received during recruitment to the study was valued by patients, and this increased their willingness to accept the DPC intervention.

I hadn’t a clue as regards the effectiveness either of penicillin or their substitute. The substitutes that they gave me always seemed to work actually so it didn’t bother me. (Site-B, Patient code: 745)

In three instances where patients declined to undergo the DPC, reasons related to poor health, inconvenience or a combination of both. Reasons for agreeing to the DPC varied (eg, ‘I’ve no idea. I just did it’). Patients cited a range of factors, often in combination, such as perceived benefits to themselves and others, curiosity, altruism and opportunism (ie, taking advantage of an episode of hospital care).

HCPs also acknowledged that they often did not always question the reported PAL. Reasons included lack of confidence in assessing and handling risk, feeling unsupported if anything was to go wrong, or not considering this to be part of routine practice. Their own knowledge of penicillin allergy and professional skills relevant to (de-)labelling were crucial factors. There was some divergence of opinion as to how much knowledge could be codified into a ‘tick box’ or checklist approach. These knowledge and skill deficits were considered most challenging in clinical settings with high volumes or turnover of staff, who were less likely to have sufficient time and resource devoted to allergy history taking, or to building up the experience in delivering the DPC.

…And there’s no foundation doctor who’s ever going to think, ‘oh it’s low risk, [give] Amoxicillin.’… (Site-C, Consultant Anaesthetist)

… confidence is the key, … the correct people who have had the correct training…the worst thing would be that somebody does something and it goes wrong… (Site-C, Consultant Pharmacist)

Specialists felt that they had a professional role in encouraging and promoting the appropriate assessment and review of PALs across all sectors. In view of the challenges of general practice (including limited time appointments and opportunities to monitor), the secondary care setting was thought to be more appropriate than intervention by primary care colleagues. A long-term strategy would require raising awareness and training, generally and in primary care, to shift the culture and approach to assigning PALs in the first place. A broader societal and ethical imperative for all HCPs was also outlined to promote and practice antimicrobial stewardship.

Intervention implementation phase

‘Environmental context’ and ‘beliefs about capabilities and consequences’ significantly influenced the experience and delivery of the DPC. At the macrolevel, participants noted relevant national resources and policies—from shortages of specialists to the national antimicrobial stewardship agenda. Meso-level factors included the use of guidelines, electronic decision support tools and governance frameworks within organisations. At the microlevel, availability of dedicated space (eg, treatment room or clinic), equipment (such as blood pressure monitoring) and medications (for the intervention as well as rescue or supportive medicines in case of adverse events) were considered essential to successful implementation. Choosing the optimal timepoint to carry out the intervention, in terms of both care pathways and the clinician workflows, was seen as important. Clinical areas with the facility for close monitoring and conditions, where there was a need to prescribe antibiotics, such as acute medical units or theatres, were reported as facilitators.

HCPs’ concerns regarding their own capabilities and the consequences of making decisions related to PAL status were influenced by environmental context and resources, which influenced their willingness and ability to implement the intervention. The combination of knowledge, skills, professional role, clinical expertise and level of experience contributed to the level of confidence they held in their own and others’ capabilities. An interesting dynamic was noted between allergy specialists and generalists; while the former considered the intervention to be relatively straightforward and easy to implement, non-allergy specialists and junior staff indicated that their role and confidence levels would inhibit their willingness to implement DPC for fear of adverse outcomes. Similarly, the research pharmacists and nurses noted that in case of any ambiguity at the risk stratification stage, a full consultation would take place with senior colleagues before progressing to the DPC. Although non-specialist delabelling was a key principle of the DPC intervention, senior clinician and if necessary, the option to seek allergy specialist input remained an important element of variability across the three sites. Research pharmacists who delivered the DPC reported that the level of allergy specialist input was relatively minimal and reduced over time with a transferral of responsibility to non-specialists. However, this was more gradual and observed to a lesser degree in one of the sites.

I've never yet seen a [penicillin de-labelling] study that’s genuinely completely led and delivered by non-allergists. There’s always someone right there in the middle of it who’s got a lot of expertise, who answers a lot of questions. (Site-B, Consultant Anaesthetist)

Few patients expressed doubts about undergoing the DPC and, in many cases, they were reassured by the DPC being conducted in a hospital setting as well as from their ‘total confidence’ in the staff involved. Some patients recounted anxiety immediately before or during the DPC, expressed as being ‘scared’ or recalled symptoms associated with their initial label and concerns at these being repeated.

I think it depends on how severe it is, and if it’s very mild which I presume mine was, then you should take the test, but if you’ve got problems like you know, life threatening ones, like your tongue is swelling up or something like that, then I don’t think you should risk it. (Site-A, Patient code: 1077)

I would have had concerns if they’d have sent me away outside of hospital because obviously if I did start having a rash or have a severe reaction to it then obviously, I’d be, I wouldn’t be in healthcare hands. (Site-C, Patient code: 3038)

However, all those who underwent DPC reported positive experiences, indicating that the process was well managed, with clear communication and advice. Facilitators of participation included reassurance of being observed, ease and simplicity of the DPC, use of oral formulation and travel reimbursement. No negative experiences or feedback comments were reported. Patients advocated raising awareness and increasing access to the risk stratification and, for those stratified as low risk, undergoing DPC to remove the PAL, recognising that …it is personal choice, but who’d want to keep a label that you know was perhaps incorrect at the time? (Site-A, Patient code: 1061)

Postintervention phase

In the post-DPC phase, ‘behavioural regulation’ relating to future communication about allergy status and reinforcement through updates to the medical record were reported as important. For example, all patients expressed a preference and expectation that the change in their PAL status would be communicated to their GP by the team conducting the DPC. This was partly linked to the logistics of arranging an appointment with their GP and partly due to uncertainty about how to respond if asked about their allergy status in future. Some indicated that they would still mention the historical label to make the HCPs aware in case of any reactions or need for close supervision if penicillin antibiotics were prescribed.

I think it depends on the circumstances… if it was something you know… like sepsis… I would explain that I for years was under the impression I was allergic to penicillin but I’m not anymore, you know? (Site-A, Patient code: 1044)

HCPs at all sites recognised the importance and value of the intervention in improving antimicrobial stewardship and patient outcomes. They also acknowledged the extensive changes to clinical behaviours, systems and practice that were necessary for implementation. Key determinants to simulate and reinforce changes to increase adoption, spread and sustainability included published evidence, incentives and alignment with national initiatives such as specialist society recommendations. Organisational policy and governance frameworks were considered essential to support individual clinician decisions and practices.

…if something goes wrong here, another clinician has said they’ve got a Penicillin allergy, I’ve completely ignored it and given them Penicillin anyway. So having that framework and guideline so that you’re backed by the organisation is absolutely critical… (Site-A, Consultant Respiratory Medicine)

Financial implications, in terms of incentives or penalties, leveraging existing mechanisms for clinical quality improvement, such as local and national clinical quality indicators,18 resources and cost of services were also mentioned as ways to influence individual behaviour and organisational change. Other initiatives, such as incorporating allergy history into national training and membership programmes, were identified as ways of embedding this into existing professional development pathways. All participants noted that recognition of the wider, public health implications of spurious PAL at a global level would also reinforce the importance of the intervention.

Emergent cross-cutting themes

Digital systems and socioprofessional influences were central to all stages of implementation, from screening patients with a PAL to communication of the outcome. The impact of digital systems on documentation, and general quality of documentation of allergy was noted as an enabler as well as a barrier to accurate allergy labelling, within and across care sectors. A recurring theme from the focus groups was the knowledge gaps caused by poor medical record keeping; once recorded, the information influenced decisions made by others involved in the medicines management process. In the case of electronic health records, there was an added layer of complexity from any resulting clinical decision support rules that were triggered.

Social influences were also identified between professionals working in the sites. In some cases, these had a long history and high levels of trust, which was seen as predisposing colleagues to support the intervention. In the absence of this, the need to actively influence other colleagues was seen as crucial, and overcoming resistance through peer-to-peer social influence was acknowledged to be challenging.

…it’s not so easy to … be like ‘no actually I want this patient to have a penicillin for their cellulitis’. … You need everyone to be on that page with you … Going against the clinical team sometimes is difficult and I’ve felt that resistance. But most of the time you’re okay and, you know, people see what you’re saying. (Site-A, Research Pharmacist)