Protocol, charters and analysis plans
We published a protocol describing the CARE pilot trial collaboration, the methods of the pilot RCT and the QRI.1 We registered the CARE pilot trial with International Standard Randomised Controlled Trial Number (ISRCTN) prospectively (number ISRCTN41647111) before recruitment started. We created charters for the Trial Steering Committee (TSC) (see Report Supplementary Material 3) and Data Monitoring Committee (see Report Supplementary Material 4) and the Patient, carer and public involvement and engagement Advisory Group (PAG) terms of reference (see Report Supplementary Material 5). We wrote date-stamped documents describing the statistical analysis plan (see Report Supplementary Material 6) and health economic analysis plan (see Report Supplementary Material 7).
The CARE pilot trial was a prospective, randomised, open-label, assessor-blinded, parallel-group trial at neuroscience centres in the UK and Ireland. Comprehensive information about trial design is available in the published protocol and trial report.1,40 We aimed to recruit ≈60 people of any age, gender and ethnicity who had mental capacity, were resident in the UK or Ireland and had a symptomatic brain cavernoma. Computerised, web-based randomisation assigned participants (1 : 1) to treatment of their symptomatic brain cavernoma with medical management and intervention (using neurosurgery or stereotactic radiosurgery) or medical management alone, stratified by the neurosurgeon’s and participant’s consensus about intended type of intervention before randomisation. We chose a single, two-arm, 1 : 1 parallel-group trial, apparently ‘lumping’ the two types of intervention in one arm, but with randomisation stratified by preferred type of intervention, in order to provide information about the frequency of preferences and uncertainties about the type of intervention and separate assessments of the effects of types of intervention in two as-randomised subgroups (thereby addressing some of the criticisms of ARUBA). Assignment was open to investigators, participants and carers but not clinical outcome event adjudicators.
The embedded QRI within this trial35 was a mixed-methods study that involved collection of screening log data (numbers and proportions screened, eligible, approached and randomised for each site and in total), 79 audio-recorded recruitment consultations between patients and healthcare professionals, 19 interviews with healthcare professionals, 11 interviews with patients who declined or withdrew from participation, discussions at 5 investigator meetings and observation of Trial Management Group (TMG) meetings.41 The QRI researchers identified and addressed barriers and facilitators with corresponding actions to optimise informed consent and recruitment in collaboration with the chief investigators and CARE trial team before recruitment began (e.g. co-design of patient-facing information, recruiter training at the time of site initiation visits) and throughout the recruitment period (e.g. tips and guidance documents for recruiters, CARE chief investigator narrated videos, described further below, investigator meetings and individual recruiter feedback).
Websites
We created a website (www.ed.ac.uk/care-study) that hosted information for patients and investigators and linked to the web-based trial database. The patient advocacy organisation, Cavernoma Alliance UK (CAUK), also created a website using plain English for patients, parents/guardians and carers describing the purpose and design of the study, as well as the support that CAUK provided for patients who were contemplating participation (https://cavernoma.org.uk/care-study/).
Investigator self-training materials
We developed self-training materials for investigators as narrated videos with PowerPoint® (Microsoft Corporation, Redmond, WA, USA) presentations (https://media.ed.ac.uk/channel/CARE±pilot±trial/205181893). Investigators had to watch these videos and then score 100% in an online quiz (https://forms.gle/HQ1nhyGy5ahCshAE7) in order to conduct study-related procedures. As a minimum, the principal investigator (PI) and site co-ordinator had to provide a certificate of completion of the quiz before a site initiation visit could take place. The training videos covered the following topics:
- CARE Study Training with the Chief Investigator.
- CARE Study Protocol Training for the QRI (Part 1 – Introduction). This training module provides a brief introduction to the Information Study.
- CARE Study Protocol Training for the QRI (Part 2 – Putting it into practice). This training module provides an overview of what site teams need to do for the Information Study.
- CARE Study Protocol Training for the QRI (Part 3 – Processes and SIV prep). This training module covers the QRI processes and tasks the site teams should complete before the site initiation visit.
- CARE Study Training with the Trial Manager. This training module focuses on the study assessments and time points covered in the protocol.
- Sponsor standard operating procedure (SOP) training for CARE. This training module covers the sponsor’s SOPs that are applicable to this study.
Sites
We provided a public list of participating sites, including the contact details of the PI and co-ordinator at each site (www.ed.ac.uk/usher/edinburgh-clinical-trials/our-studies/all-current-studies/care/care-study/sites) in order to facilitate the patient advocacy organisations in the UK and Ireland signposting the trial to potentially eligible patients (see Patient and public involvement), so that they could determine whether their local neuroscience hospital was participating in the trial. We engaged all investigators and project partners with weekly e-mail newsflashes (containing a recruitment update, any good news and signposts to the latest learning from the QRI or other important issues of trial conduct) and a monthly newsletter with lengthier updates and recruitment league tables (see www.ed.ac.uk/usher/edinburgh-clinical-trials/our-studies/all-current-studies/care/care-study/newsletters for the newsletter archive).
Participant information leaflets
We developed short and supplementary information leaflets in collaboration with the PAG and QRI researchers, including separate versions customised for adults, or for parents and guardians of children, three leaflets for children (ages 0–5, 6–10 and 11–15 years) and separate information leaflets for Ireland (www.ed.ac.uk/usher/edinburgh-clinical-trials/our-studies/all-current-studies/care/care-study/get-involved).
Frequently asked questions
After recruitment began, we had foreseen that some procedural and logistical issues would arise. We also encountered barriers and facilitators to screening and recruitment as part of the QRI. In order to address these issues and frequently asked questions, we created short, narrated PowerPoint presentations as videos – involving discussions between the chief investigator (Rustam Al-Shahi Salman), co-chief investigator (Neil Kitchen), the QRI researcher (Julia Wade) and key TMG members, known as ‘CARE Chats’ – which were disseminated via the weekly newsflashes and posted on the trial website for investigators (www.ed.ac.uk/usher/edinburgh-clinical-trials/our-studies/all-current-studies/care/care-study/faq). These pre-recorded videoed discussions have not been used in previous QRIs. The topics included:
- screening logs
- putting the QRI into practice
- approaching patients diagnosed long ago and treated without intervention
- logistics of recruitment
- recruiter tips and guidance
- tips for conversations about the CARE study
- patients’ frequently asked questions
- tips for recruitment conversations
- QRI audio-recordings
- tips for describing randomisation
- tips about timing of randomisation
- logistics of using stereotactic radiosurgery.
Outcomes
In the protocol, we pre-specified a primary feasibility outcome involving 13 measures of the feasibility of a definitive RCT; a primary clinical outcome of ICH or new persistent/progressive FND due to brain cavernoma or intervention; several secondary clinical outcomes [death, modified Rankin Scale score, National Institute of Health Stroke Scale (NIHSS) score, EuroQol-5 Dimensions, five-level version in adults and EuroQol-5 Dimensions, three-level version in young people (EQ-5D-Y), Karnofsky Performance Status scale in adults and Lanksy Play-Performance Scale in children, and Liverpool Seizure Severity Scale]; and collection of data to estimate health service use and healthcare and socioeconomic costs.1
In addition, the contract with NIHR pre-specified the following success criteria:
- At least 30 sites in the UK and Ireland collaborate.
- Project delivered according to the major milestones identified in the timetable.
- Recruitment to within 10% of target in the CARE pilot RCT.
- Brain cavernoma radiographic diagnosis confirmed by expert neuroradiologist review in > 95% of patients recruited.
- Retention of > 95% of participants at 6 months in the CARE pilot RCT.
- < 10% treatment group switches or lost to follow-up in the CARE pilot RCT.
- QRI is associated with an improvement in recruitment.
- Definitive RCT appears feasible and affordable.