This SWAT protocol is written in accordance with the guidelines for reporting embedded recruitment trials outlined by the MRC START Group [34]. The protocol has been pre-registered on the Northern Ireland Hub for Trials Methodology Research SWAT repository (Queen’s University Belfast) (SWAT ID #159) [35]. The overarching SWAT has received ethical approval (Leeds West REC ref. 22/YH/0121). Each host trial will obtain approval for embedding the SWAT, either from the start of the trial or as a substantial amendment after the host trial has commenced.
Study design
The CONSULT SWAT is a two-arm, parallel group, embedded randomised-controlled trial to investigate the effect of a decision support intervention compared with standard study information on decision-making by consultees and legal representatives of adults lacking capacity to consent. The study uses established SWAT methodology [27] to evaluate the intervention in approximately 5 host trials that recruit adults who lack capacity through personal consultee or personal legal representative involvement (see Fig. 1).
An initial feasibility stage will be conducted in one host trial to test the intervention and SWAT with 20 family members and up to 15 research staff (recruiters and members of the trial co-ordination team). This will explore the feasibility of SWAT procedures including outcome measures, establish acceptability and identify any unintended consequences, help estimate the likely rates of recruitment and retention of participants for the SWAT, and give an understanding about likely SWAT sample sizes. There are generally no formal a priori sample size calculation for SWATs [27]. The SWAT will be undertaken on the basis of the maximum number of recruiters and participants possible per host trial.
An embedded process evaluation and economic evaluation in the SWAT will enable the findings to be contextualised in order to draw robust conclusions about the effectiveness of the intervention and factors likely to affect implementation. Process evaluations enable an understanding of the causal assumptions underpinning an intervention and are a vital part of the use of evaluation to understand how interventions work in practice [36]. In addition to quantitative data such as data completion rates, approximately 20–25 participants from the intervention arm and 15 recruiters will be interviewed. The interviews will be conducted across host trials and will provide more in-depth information about how the intervention works in different trial contexts. The economic evaluation will take the form of a cost-consequence analysis (CCA) where disaggregated costs and a range of outcomes are presented [31]. The CCA will take an NHS and societal perspective by identifying and measuring the resources involved in intervention production and delivery as well as resource use, including time use, by family members and research staff.
Objectives
The objectives for the feasibility stage of the SWAT are to:
-
Establish the feasibility of the novel decision support intervention in a real (rather than hypothetical) decision-making situation
-
Establish the feasibility of conducting a SWAT
-
Establish the feasibility and acceptability of the CONCORD scale in a real (rather than hypothetical) decision-making situation and of measuring resource use
The objectives of the main SWAT are to:
-
Evaluate the effectiveness of the decision support intervention through a randomised SWAT
-
Explore issues affecting future implementation such as reach, contamination, context, adaptation and fidelity through an embedded process evaluation
-
Undertake an economic evaluation through a CCA to explore the resources involved in intervention delivery
Eligibility criteria
Host trials can be based in any setting, including in primary or secondary care, the community, or care homes. Host trials will be eligible if the trial team anticipates that a reasonable proportion of potential participants will lack capacity to consent and will involve personal consultees or personal legal representatives. Trials will be ineligible if they only involve recruitment without prior consent (‘deferred’ consent), use only nominated consultees or professional legal representatives, or if the participation decision needs to be made urgently or within a short timeframe (i.e emergency research). In order to provide a greater understanding about the factors that influence the effectiveness and implementation of the intervention, host trials and sites may be purposively selected to participate in the SWAT where ongoing analysis suggests that conducting the SWAT in specific trial contexts is needed. Informed by the preparatory qualitative research, this may include a range of study types (e.g interventional and non-interventional studies considered ‘burdensome’) with differing risk levels, research settings (e.g care homes, secondary care) and populations/conditions (e.g acute loss of capacity vs long-term impairing condition).
Participant inclusion criteria:
Participant exclusion criteria:
Recruitment
Participants will be identified in accordance with the host trial processes for recruiting participants who lack capacity to consent. Recruitment processes are likely to differ between host trials and may differ between participating sites. As part of aligning the SWAT with the host trial processes, the procedures for identifying eligible participants for the SWAT will be discussed and agreed in advance for each host trial and participating site. The SWAT will use a proportionate approach to informed consent in accordance with HRA guidance [37]. Brief information about the study will be provided to family members approached to act as a consultee or legal representative through a combined information sheet and questionnaire. The information section will include the purpose of the SWAT and data protection arrangements, and that completion and return of the questionnaire indicates consent to participate in the SWAT. Participants can also provide their contact details if they are willing to be contacted to take part in an interview for the feasibility stage or process evaluation. SWAT participants and research staff delivering the DA intervention will be purposive sampled to ensure that a range of perspectives are included in the interviews. SWAT participants and research staff agreeing to take part in an interview will provide consent prior to commencing the interview.
Intervention and allocation
The intervention consists of the DA (a 12-page A5 colour booklet ‘Making decisions about research for others’) which is provided to family members in addition to standard study information about the host trial. The control is standard study information alone for family members. The DA is intended to be used by the family member at the time they are making a decision about whether the person they represent should participate or not. The development of the DA and its contents are described elsewhere [26]. Briefly, it contains an explanation about why they are being approached, why adults lacking capacity are included in research, and a six-step guide to making a decision. It also includes a values clarification exercise to help them to understand what the advantages and disadvantages might be, and to consider how the person they represent would view them and come to a decision about participating or not. The family member is encouraged to identify any areas they feel that they need more information in order to make a decision and to write down any questions they may have in the spaces provided. Depending on the method of communicating with consultees and legal representatives used in the host trial, the DA may be posted out to the participant to be read and completed remotely, or they may be provided with it (in person or remotely) ahead of a consultation with the recruiter (in person or via telephone or video conference) to discuss the person they represent’s participation in the host trial.
The level of randomisation to either the intervention or control arm will be dependent upon the host trial design e.g whether cluster or individually randomised, number of sites and recruiters (see Fig. 2 for the trial schema). Randomisation will preferably be at an individual level (family member) as the intervention is highly amenable to randomisation at that level. But cluster randomisation (recruiter or site) may be required where the host trial itself is cluster randomised, where individual randomisation might cause disruption to the host trial, or where cluster randomisation is the most feasible option for the host trial. Detailed field notes will be maintained to facilitate decisions about randomisation with host trials and to ensure that the randomisation and decision-making process is reported in full. Randomisation will occur in a 1:1 ratio to either intervention or standard information arm. The allocation sequence/algorithm will be generated centrally by the Centre for Trials Research (CTR) who are co-ordinating the study. As the level of randomisation will vary depending on the nature of the trial and the feasibility of individual randomisation it may not be possible to maintain blinding. However, participants will be blinded to allocation (they will not be informed that the study involves randomisation to a DA or control), and site staff involved in data collection will be blinded where possible.
* CRF – Case Report Form. Return of CRF indicates consent to participate
Outcomes
The primary outcome is the quality of proxy consent decision as measured by the Combined Scale for Proxy Informed Consent Decisions (CONCORD) scale. CONCORD scores will be recorded both for consultees and legal representatives who do provide agreement to participation on the person’s behalf and those who decline participation. The timing of outcome measure completion will be aligned with the host trial processes but it is expected to be completed in a relatively short timeframe following the decision. CONCORD is a novel outcome measurement instrument which has previously undergone feasibility and acceptability testing in a hypothetical decision-making situation [38] and will be concurrently validated during the SWAT.
Secondary outcomes include selected CONCORD subscales of values clarity and preparedness, and where possible the proportion of consultees and legal representatives who provide agreement to participate on the person’s behalf and the proportion who decline participation and subsequent retention in the host trials.
Data collection
Following receiving the intervention (or standard information alone), participants will be asked to complete a questionnaire which includes questions relating to demographic data, use and views about the information received, decision outcome (agreeing to participation on the person’s behalf or declining) and the CONCORD scale. The questionnaire will be available in hard copy and online format depending on the host trial processes and will include an option to indicate if the participant is willing to be contacted about taking part in an interview. Where feasible, linked pseudonymised data will be collected on the recruitment and retention of the host trial participant. A web-based system (Qualtrics) will be used to enter data online, either directly or from completed paper questionnaires. All data will be stored securely and in compliance with the Data Protection Act 2018.
Interviews will also be conducted remotely with SWAT participants and research staff in both the feasibility stage and process evaluation. Interviews will be semi-structured using topic guides that have been informed by previous qualitative research [19]. Interviews will be audio-recorded and transcribed verbatim and pseudonymised prior to analysis. Additional resource use data will be collected and assessed in each host trial using a method that corresponds with the host trial processes for approaching and informing potential consultees and legal representatives. For example, where recruitment is conducted face-to-face or during a consultation, the time required for the consultation and discussion will be recorded where possible.
Analysis
Statistical analysis will be performed using statistical software (Stata). Descriptive statistics will report participant demographic data across the two arms of the SWAT. A two-stage meta-analysis strategy will be used to analyse each individual SWAT taking into account whether it was individually randomised, or cluster randomised to generate trial-level summary statistics, with the results from each individual SWAT then combined across trials. Primary analysis will be on an intention to treat basis based on those who return questionnaires. Sensitivity analyses will be undertaken to assess the robustness of the results to non-response to questionnaires, and modelling undertaken if there is a differential response rate between those who receive the decision aid and those who do not.
Where additional data has been collected on recruitment and retention of host trial participants, subgroup analyses will investigate differences between decision outcomes, i.e whether or not the consultee/legal representative’s decision was that the person should participate in the host trial. Further subgroup analyses and sensitivity analyses may investigate differences between host trial results based on factors such as underlying recruitment rates of the host trials, and timing to outcome measure completion.
Qualitative analyses will be supported by data analysis software (NVivo). At the feasibility stage, interview data will be analysed thematically to provide an understanding of participants’ views about the acceptability of the SWAT, intervention, and outcomes. During the process evaluation, qualitative data from the interviews will be analysed using framework analysis and integrated with the quantitative data to contribute to the assessment of fidelity and other implementation factors. Attention will also be paid to any adaptations that were made in different contexts in order to implement the intervention (which may also undermine intervention fidelity), any contamination between arms, and any changes over time. A key focus of the analysis will be to further develop and test the logic model underlying the decision support intervention to examine the likely mechanisms of action, and the active components.
The contextual factors that may be associated with variation in outcomes will be explored in order to analyse how implementation may vary from one context (host trial, population, setting etc) to another, and whether the intervention has different effects in different contexts even if its implementation does not vary [30]. Contextual moderators will be analysed through detailed description of the nature of the host trial, trial population and consultee/legal representative population, setting, and trial processes that are recorded by the research team through discussion with the host trial team. Qualitative interview data will also be analysed to explore the use and implementation of the decision aid in a range of different contexts.
As part of the concurrent validity testing of the CONCORD scale, factor analysis will be conducted to identify items that most clearly represent the content domain of the underlying construct and any redundancy through testing the internal consistency. Supported by qualitative data from the feasibility stage and process evaluation, this will enable further exploration of existing hypotheses about relationships between constructs and items, and ongoing revision of the underlying theory and construct validation.
The cost-consequence analysis will analyse quantitative resource use data such as the time required for training on the intervention, costs associated with delivering the intervention (e.g printing, postage), and time required to deliver the intervention, in addition to the usual costs associated with delivering standard trial information. The costs will then be tabulated from the NHS and societal perspectives and presented alongside the CONCORD measure and relevant contextual information using a descriptive table [31].