TREatment of ATopic eczema (TREAT) Registry Taskforce: protocol for an international Delphi exercise to identify a core set of domains and domain items for national atopic eczema registries

Atopic eczema (AE) (synonymously ‘atopic dermatitis’) is among the most common chronic inflammatory disorders, with a lifetime prevalence of 15–30% in children and 2–10% in adults [1] and is known to give a high burden of disease to patients, their families and society [2]. Whilst most patients with AE can be treated effectively with emollients and topical anti-inflammatory agents, a significant number will require photo- or systemic immunomodulatory therapies to induce disease remission and maintain long-term control [3]. Ciclosporin is currently the only systemic treatment approved by the European Medicines Agency (EMA) as a treatment for severe AE [4]. Alternative, but off-label, systemic immunomodulatory therapies include methotrexate, azathioprine, mycophenolic acid (mycophenolate mofetil or mycophenolate sodium), systemic glucocorticosteroids, and intravenous immunoglobulin. Novel agents are currently tested in clinical trials (e.g. dupilumab, lebrikizumab, apremilast, ustekinumab, and CRTH2 antagonist (chemoattractant receptor-homologous molecule expressed on Th2 cells)).

The current evidence to guide clinical management for moderate-to-severe AE originates from a small body of randomised controlled trials (RCTs) [5] and observational studies [6‐8]. There are no long-term, comparative and real-life data on the effectiveness and safety of these treatments in children or adults from large-scale, multicentre cohort studies. Several scientific guidelines and a systematic review highlight these gaps [5, 9, 10] and lament the resulting lack of clear management guidance to inform clinical practice.

Nevertheless, as shown in a recent survey among over 700 dermatologists and paediatricians from eight European countries, these immunomodulatory treatments are frequently prescribed as off-label medicines in children [11]. It is likely that these treatments are used more frequently in adults, but no research has yet been performed.

The establishment of national AE patient registries with a research focus that use the same methodology would not only partially fill the gaps in the current evidence, but could also operate as a research platform to inform the design of future RCTs and basic research and serve as an example for other initiatives. Such a platform will allow the collection of prospective daily practice data on effectiveness, safety, cost-effectiveness, subgroup characteristics, quality of care and personalised medicine. Our research group, the TREatment of ATopic eczema (TREAT) Registry Taskforce, is currently developing an interoperable patient registry platform based on best-practice guidelines for this purpose [12].

Experience has shown that disease registries that were set up without an a priori agreed core set are likely to yield disparate datasets, hampering direct comparisons of results and data pooling [13]. To overcome this problem, an internationally agreed core set for AE patient registries is required. The European Commission funded PAtient REgistries iNiTiative Joint Action (PARENT JA) has developed methodological guidelines for the development of patient registries in cross-border settings [12] to reduce heterogeneity, enhance direct comparability of individual country data and improve data pooling between countries.

In alignment with the PARENT guidance, the objective of this eDelphi study is to reach international consensus between different stakeholders on a core set of domains and domain items (what to measure), for existing and future atopic eczema patient registries with a research focus, that collect data of children and adults on photo- and systemic immunomodulatory therapies. This protocol outlines the methodology.

This study will be conducted following the recommended checklist proposed by Sinha et al. [14], of items that should be reported in studies using the Delphi technique [15]. Details of our study have been included in the Core Outcome Measures for Effectiveness Trials (COMET) database and are available at www.​comet-initiative.​org/​studies/​details/​825?​result=​true. The Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist can be found in Additional File 1.

By the end of this Delphi study we hope to have reached consensus on the core set of domains and domain items to capture in AE patient registries with a research focus, that collect data of children and adults on photo- and systemic immunomodulatory therapies. This internationally agreed core set has the potential to unify data collection within existing and future AE patient registries to allow direct comparisons and data sharing and pooling. Analysis of these registry data will facilitate insight into the effectiveness, safety and cost-effectiveness of photo- and systemic immunomodulatory therapies across dermatology centres and country boundaries.

Applying the Delphi method has several advantages. First, it allows us to include a large group of international experts and patients who are geographically dispersed to participate. In this way, broad consensus can be obtained. Second, the anonymity of the survey may avoid the dominance of certain persons in contrast to face-to-face meetings. However, this could diminish the positive effects of interaction to be found in face-to-face meetings, which helps to identify reasons for disagreement [17]. Therefore, a final consensus meeting will be held in case of remaining disagreements. Third, feedback is given to participants, allowing them to consider the answers of not only their own but also of other stakeholder groups.

Despite the obvious advantages of the eDelphi process, there are several sources of possible biases. Specifically, there is a risk of disproportionate representation among individual groups, including a dominance of clinicians and researchers, and relative underrepresentation of other groups such as patients and families [14]. By contacting patient organisations and using personal contacts, we hope to enrol sufficient numbers within this important stakeholder group. Dominance of individual participants or representatives of one specific stakeholder group at the face-to-face consensus meeting may be another risk. The deployment of an independent facilitator will minimise this risk by ensuring that all voices will be adequately heard. Furthermore, we have chosen to use a prespecified list of domains and domain items in the survey next to an obvious list suggested by our research group. This may bias responses of participants and overstate the importance of domains and domain items recognised by our research group [25]. However, this was done in an effort to optimise the efficiency and feasibility of the survey. Additionally, participants will have the opportunity to add domains and domain items and to comment on the obvious list in the first round. Another consideration is that patients may struggle to interpret the medical terminology used in the eDelphi, which in turn may limit the value of their response. We will, therefore, involve patient organisations and patients in the development and pilot phase, include help texts in the survey and use plain English where possible. A further potential problem is a declining response rate from round to round [17]. Reminder emails will encourage completion of questionnaires, and we will also only invite people who responded to a pre-Delphi invitation, as they are likely to be more motivated to complete the survey. In addition, the extent of the core set may be a hurdle for patients as they may struggle to tell the difference between what is important to themselves and what is important to be included in a patient registry that is focussed on research. This may encourage them to include more domains and/or domain items than other stakeholder groups or to dismiss domains and/or domain items that they see as unimportant, e.g. surveillance data. This may impact the feasibility of the final core set. Therefore, we will highlight the importance of agreeing on a core set in each round of the eDelphi survey and at the consensus meeting. Another point of debate is the challenge to accommodate potential cross-cultural differences in a core set. We will, therefore, extend our invitation to take part in the eDelphi exercise to colleagues in (East) Asia, Africa, South America, and Australia and New Zealand, in addition to European and North American participant representation (see Tables 1 and 2 with dermatology societies and eczema patient support groups to be invited). Lastly, a core set will only have impact if it is consistently implemented in patient registries with a research focus. To achieve implementation in these registries, we must actively engage with all stakeholders, especially international societies and regulatory bodies after the Delphi exercise is completed to ensure that this core set is accepted and used.