1. Trial design and setting
The study will be a multi-centre randomised pilot trial carried out in community settings in Gloucestershire, North Wales and the Vale of Glamorgan where patients are likely to die at home in accordance with their wishes. The three pilot study sites have been chosen as they are representative of the range of sites for a future definitive study.
The trial was funded by the Health Technology Assessment Programme of the National Institute for Health Research (National Health Service). It has received a favourable ethical opinion from the Wales 1 National Research Ethics Committee (REC reference: 17/WA/0208, IRAS project ID: 227970) and the Bangor University Research Ethics Committee. The UK Medicines and Healthcare products Regulatory Agency (MHRA) has advised that this randomised pilot trial is not a Clinical Trial of an Investigational Medicinal Product (CTIMP). The study is registered on the International Standard Randomised Controlled Trials Number (ISRCTN) registry (ISRCTN11211024). Approval was granted from the Research and Development departments of all three sites. SPIRIT 2013 recommendations and CONSORT 2010 statements (including those specific to randomised pilot and feasibility trials) guided protocol development.[31,32] The current version of the protocol can be accessed via the National Institute for Health Research (NIHR) Journals Library.[33]
2. Inclusion/exclusion criteria for participants and recruitment
Inclusion criteria: Dyads of
- An adult (18+) patient in the last weeks of life, who is likely to lose the oral route for medication, and who has expressed a preference to die at home, and
- Their adult (unpaid) lay/family carer, who is willing to have this extended role and have SC-injection training.
There is an assumption that the carer will spend a significant amount of time with the patient. Whilst Australian experience indicates that one lay person generally takes a lead role in this practice, where there is more than one suitable carer we will ask the patient to identify which carer they would like to be included in the study.
Exclusion criteria: Patients who have only paid/formal care or with previously known adverse reactions to the ‘usual’ as-needed medications will be excluded. Patients or lay carers who have not met the risk assessment criteria (see below) will not be approached for consent.
Patient identification: Patient / carer dyads will be identified through the hospice, SPC service, or DN team. When a patient is perceived by the HCP team to be in the last weeks of life and they have expressed a wish to be cared for and die at home, they will be screened for approach.
Screening: To be eligible, dyads must satisfy the risk assessment criteria. A risk assessment screening tool has been refined for CARiAD, based on existing self-medication tools.[34] Risk assessment will take into account several factors, including:
- the carer’s mental state, vision and physical condition
- the dyad’s attitudes to medicines and willingness to engage with the healthcare team
- relational issues including concerns about burden
- history of substance misuse in the family
The risk assessment will be conducted by the healthcare team involved in the patient’s care. If a dyad does not satisfy the risk assessment criteria, they will not be approached.
Approach: The patient will be approached with written material by a member of their healthcare team. The initial patient approach will be done separately from the carer, unless otherwise requested by the patient and if the attending HCP deems this appropriate i.e. there is no perceived risk of patient-carer coercion. As the project involves sites in Wales, the Participant Information Sheets and Consent forms will be translated into Welsh for the Welsh centres and offered bilingually to comply with the Welsh Language Act 1993. Dyads will be given as much time as they need to consider the information sheets and discuss with family, friends or the healthcare team until they decide whether to take part. They will be told that they can refuse participation without giving reasons.
Informed consent: A researcher will seek advance consent separately from both the patient and their lay carer at a time judged to be suitable by the attending HCP. This gives the patient and carer as much time as they need to understand the nature of the research, ask questions and make their feelings clear on trial participation.[35,36]
If the patient is unable to consent, or once they lose capacity after they have previously given consent, the assent of a Personal Consultee will be sought (as required by the Mental Capacity Act 2005) to the patient’s participation in the trial. As the risk assessment will exclude dyads where there are concerns about relational issues between patient and carer, the carer can act as Personal Consultee.
If the carer does not wish to act as the Personal Consultee and there is no additional family member or close friend to take on this role, we will appoint a Nominated Consultee (e.g. a HCP not associated with the research) who will act for all patients in this situation in the trial.
Randomisation: Once the dyad has consented and baseline data has been completed, the dyad will be randomised to one of the trial arms. Secure online randomisation hosted by the North Wales Organisation for Randomised Trials in Health (NWORTH) Clinical Trials Unit will be performed by the researcher who has obtained consent. The system will use a dynamic adaptive method of randomisation stratifying for recruitment centre and diagnosis (cancer/non-cancer).[37] Confirmation of allocation will be sent only to those members of study staff who need to be aware of the result.
Blinding: CARiAD is an open trial where blinded outcome assessment is not feasible, and therefore it is important that outcomes are as robust as possible in light of the lack of blinding. Outcome assessors will be experienced research nurses. Only the trial statistician providing data analysis will be blinded to treatment allocation.
Withdrawal criteria: Participants remain free to withdraw at any time from the trial without giving reasons and without prejudicing their further treatment. This will be made clear to all potential participants at the time they consent to participation and throughout their time in the trial. Non-completion of the follow-up questionnaires will not constitute formal withdrawal from the trial, and unless the participant requests withdrawal of their data completely, it may be used to impute values for the analysis. The risk assessment will be reviewed at intervals based on HCP judgement and if the criteria are not met the dyad will be withdrawn from the trial.
3. Interventions
Training of carers in the intervention arm will be supported by a manualised training package based on the Australian package ‘Caring Safely at Home’.[5] Lay carers will receive training on: common symptoms that may occur in the last days of life, and how to assess if their loved one needs medication for a particular symptom; how to prepare (draw up) medication and dispose of sharps (glass ampoules and drawing up needles); how to administer SC medication by needle-less technique (utilising a ‘butterfly’ SC catheter); how to assess the effect of the medication; and support available, including primary care team as well as dedicated 24/7 SPC support. If a symptom occurs for which medication is deemed necessary (either as expressed by the patient if able, or as assessed by the carer), the carer can use the training outlined above to administer the appropriate medication.
Medication regimens: Guidelines for anticipatory prescribing for last days of life care are in place across the UK.[38,39] They cover common symptoms in the dying phase: pain, nausea and/or vomiting, restlessness/agitation and noisy breathing/rattle. CARiAD recruitment sites will be advised to follow usual prescribing practice. For patients in the intervention arm only, prescribers will be provided with specific additional advice, including instructions not to prescribe dose ranges/steps, and that dose changes can only be made after a face-to-face assessment (and not remotely, i.e. over the telephone).
Care Pathways: The usual care arm has an unchanged care pathway for dealing with breakthrough symptoms at home for a dying patient, with usual palliative care and DNs administering as-needed SC medication.
‘Usual routes’ for support in each recruitment area are different. For some areas there is direct access to a 24/7 SPC advice line for carers in addition to support from the patient’s primary care team within or out-of-hours. In other areas support for the carer will be via their primary care team, while the GPs and DNs can call on advice from SPC clinicians.
In the Intervention arm, carers will be trained to administer as-needed SC medication, although they will not be obliged to do this. If the carer needs the support of a HCP, either because they would feel more confident having a HCP present when they administer medication, or they wish the HCP to assess and give medication, they can obtain it via the usual routes in their area. If the carer has reached the limit of the number of injections which can be given in 24 hours (maximum three injections for each indication per 24 hour period unless the prescribing clinician advised a maximum of fewer than three), they will be asked to contact a HCP as review is indicated. Usual routes for support might include DN team, GP, GP/DN out-of-hours, Hospice at Home team, or a hospice advice line. The use of such support will be captured in carer diaries (see Study procedures).
HCP training requirements
In order for nurses to train carers, they will themselves receive training on: the standardised manualised education package (adapted from the Australian work); the legal framework (see Additional File 1); guidelines for medication handling and administration in a community setting; and on trial-specific materials and processes.
4. Study procedures
For an overview of study assessments, see Table 1.
The main outcomes of interest will be those appropriate to a pilot trial, including feasibility, acceptability, recruitment rates, attrition and selection of the most appropriate outcome measures. Outcomes will be measured for patients, their lay carers and HCPs. System barriers will also be noted. These measurements will be made at baseline, on a daily basis for symptom control and lay carer confidence, at 6-8 weeks post bereavement, and at 2-4 months for a sub-sample (carer interviews).
Recruitment measurements are: the number of eligible patients who fulfil the inclusion criteria and are willing to be randomised expressed as a percentage of the numbers screened, the number who withdraw after baseline assessment and randomisation, the number who complete the various outcome measurements at baseline and at later time points, and reasons for any non-completion.
Patient measurements are: Baseline information (including demographic information, medical history, capacity assessment, preferred place of care in the last days of life, current drug management) and a daily Carer Diary during the study related to the presence and treatment of breakthrough symptoms (for use in both study arms). Data points in the diary will include: initial time breakthrough symptom triggered perceived need for an additional SC dose, whether noted by patient or lay carer, medication and dose, and time given, reason for medication (pain, nausea, restlessness, noisy breathing), symptom score before and 30 minutes after medication administration, when symptom control/reduction of symptom to acceptable level was achieved. Hospital or hospice admissions during last illness and actual place of death will also be recorded.
Carer measurements are: Demographic information at baseline, Quality of Life in Life Threatening Illness – Family Carer Version (QOLLTI-F) (at baseline, after the first as-needed SC medication, then every 48 hours until the patient’s death), whether HCP support was sought, Carer Experience Scale at baseline and post-bereavement, Family Memorial Symptom Assessment Score – General Distress Index (MSAS-GDI) at 6-8 weeks post bereavement visit, and qualitative interviews for a sub-sample at 2-4 months post-bereavement. Specific to the intervention arm, confidence (in administering injection) and competence at intervals after training will be recorded.
Health care professional measurements are: Baseline measurements of attending team structure, primary prescriber, carer trainer, and evaluation of the training package.
Safety
The CARiAD project contains a number of safety outcome measures at different stages of the clinical journey taken by patients, carers and HCPs. Safety outcome measures include the risk assessment, competency checklist and Significant Event reporting. Significant Event reporting will include the following: the appropriateness of administration (is administration accompanied by evidence of need?), proportionality (has the correct dose been administered?), side effects both anticipated and not anticipated, drug accountability (do stocks tally?), and carer events (e.g. distress; needle stick injury; accidental or purposeful self-administration).
All Serious Adverse Events (SAEs) will be captured via SAE reporting forms. As this is a study in patients who are terminally ill, death is an expected outcome. It will be recorded and reported to the sponsor, but will not be considered a SAE if, in the opinion of the Principal Investigator (PI), it was a natural conclusion to a patient’s terminal illness. Due to the nature of the study, events of death will not require immediate reporting to the Data Monitoring and Ethics Committee (DMEC).
5. Exploratory endpoints/outcomes for a future definitive trial
The most likely candidates for primary outcome measures for a future definitive trial are: MSAS-GDI (a measure of overall symptom burden/distress in the last seven days of life),[10,40-42] and QOLLTI-F (a measure of quality of life of carers looking after someone with a life-threatening illness, incorporating elements of control and self-efficacy).[43]
In addition, we will measure carer confidence using a five point Likert scale (where the carer is asked after administration of every as-needed SC injection to rate their level of confidence in administering this injection, 1=not at all confident, 5=very confident), and probe carer experience during qualitative interviews.
Criteria for assessing feasibility as primary outcome measure: All outcome measures will be assessed on the same criteria (applicability, acceptability and level of completeness) for consistency. Once the feasibility of the outcomes is established, the design of the definitive trial will be confirmed. The potential suitability of the following secondary outcomes will be considered: ‘Time to symptom relief’, and Carer Experience Scale.[44-46]
6. Embedded qualitative study
The aim of the embedded qualitative component is to inform the design and assess the feasibility of a phase 3 trial of carer-administered medication. The study will collect interview data from HCPs and carers to:
- Assess clinical willingness to randomise patients for a future full RCT.
- Understand the experience of randomisation between intervention and control, to identify relevant patient-centred outcomes for a phase 3 trial, and consider time points for assessment.
The qualitative study aims to include interviews with non-consenters to the trial, as well as in-depth qualitative exploration of carer and HCP acceptability of carer-administered SC medication e.g. strong opioids, anti-emetics, sedatives. The study will use a phenomenological and pragmatic approach to understand the meaning that carer-administration of injectable strong opioids and other as-needed medication has for bereaved carers and HCPs and the practicalities involved.
Sample: Face-to-face qualitative interviews across the three recruitment sites will be conducted with:
- 6-10 carers who have experience of supporting a patient in the intervention arm.
- 6-10 carers who have experience of supporting a patient receiving usual care.
- 6-10 carers who declined to be randomised to the trial. For carers in all three groups, sampling criteria will include gender and rurality.
- Up to 30 HCPs – to include prescribers (e.g. GPs and ANPs), administering HCPs (e.g. DNs) and SPC clinicians. Sampling criteria will include years since qualification, experience of supporting home deaths and practice characteristics.
Consent: Carers declining to take part in the trial will be approached upon declining and invited to participate in an interview about the reasons why they chose not to participate. They will be given a separate information sheet for this.
Data gathering: Interview topic coverage was informed by PPI input, the systematic review, and the expert consensus workshop. Attitudes to and experiences of having administered medication including emotional, ethical and practical reflections will be explored, as will issues relating to trial recruitment and feasibility (supply and storage of medication, success of training and perceived competence of carer once trained, choice and recording of the primary outcome). Carers will be interviewed approximately 2-4 months post-bereavement (as suggested by usual clinical follow-up and current literature).[1,47-49]
Interviews will be face-to-face at carers’ homes or alternative preference, or possibly by telephone; lasting 30-60 minutes. The interviews with carers who declined to be randomised to the trial will be shorter, lasting 15-20 minutes. HCP interviews will be by telephone and last around 30 minutes. All interviews will be audio recorded, transcribed verbatim, and the carer interviews will be managed using NVivo. Participants will be asked to consent to publication of anonymised quotes.
Analysis: The analytic frameworks are selected to understand the meaning that carer-administration of injectable strong opioids and other as-needed medication has for bereaved carers and HCPs. Carer interviews will be analysed using Interpretative Phenomenological Analysis [IPA] to allow a deeper, inductive analysis of the data in the context of carers’ and patients’ daily lives and values.[50] This methodology focusses on the subjective experience of participants, as interpreted by the researcher. HCP interviews will be analysed using Framework Analysis with a deductive approach.[51] Framework analysis is commonly used in healthcare and is more appropriate for examining the specific aims and objectives of the HCPs. The data will be summarised thematically and displayed on a matrix linking to the original data.
7. Identification of attributes for a future Discrete Choice Experiment (DCE)
We have identified the need to determine carers' preferences for HCP versus own administration of medication to patients, using a DCE. The preferences of carers towards administering SC medications will have a bearing on their willingness to adopt this practice, and the effectiveness of carer-administered medication.
While the DCE (aiming to ascertain carers’ preferences for their administration of SC medications) will be conducted as part of a future main study, the preparatory work required to identify relevant attributes and levels will be done as part of the embedded qualitative study component of the randomised pilot study. This will be done with each of the three carer groups in the second part of the interviews and will take up to 20 minutes. Attributes may feasibly include cost, time, perceived competency, confidence, and potential risks. The process of attribute development will be informed by best practice.[52]
The use of interviews for the determination of DCE attributes enables a greater opportunity for in‐depth exploration of particular issues and concepts than would otherwise be possible in focus groups (which are more common in DCE development). Individual interviews are also better suited to discussions concerning sensitive topics. Within the first five interviews in each group, carers will be presented with a range of attributes, identified by the research team as being likely to affect carers’ choice for own versus HCP administration of SC medications. Interviewees will have an opportunity to add other factors of their own choosing to the list, and be asked to identify and rank attributes in order of importance. Thereafter, we will use the interviews to pilot the presentation of the highest ranked attributes. The ordinal ranking across each group will be determined, and those ranked highest will be taken forward for DCE development. We have successfully implemented this method in previous DCEs,[53] and it is consistent with the reductiveness approach of attribute development.[52]
We will also pilot the Carer Experience Scale as a means to estimate carer utility.[46] The index values derived from this scale offer a preference-based approach to incorporate the effects on carers in economic evaluation, focusing on care (rather than health)-related quality of life.
8. Statistical considerations
Sample size: A fully justified sample size is not required; size has been justified by estimating what a future definitive RCT will need. Assuming an important difference of 0.4 (SD=1) on the Family MSAS-GDI a sample of about 216 is required to achieve 90% power to detect a difference of this size with a significance level of 0.05 using a two sided test. Equivalently a sample of about 550 would be required to detect a difference of 0.5 points (SD=2) using the QOLLTI-F.
Using the larger of these estimates for the feasibility trial, we will assume about 9% of the main trial size, to give an 80% confidence interval to exclude a clinically important difference, requires ~ 25 in each group. [54] Sim and Lewis recommend a sample of about 50-55 to ensure robust estimates of the variance. [55] Using estimates of dropouts we predict we need to approach 200 potential participants to achieve 100 randomised participants, with 50 completers. (‘Completer’ is defined as a dyad who completed all the study measures from baseline to follow-up at 6-8 weeks post-bereavement.) We will therefore need to approach 5.5 dyads per month from each of the 3 sites, and randomise 2.7 dyads per months from each of the 3 sites to meet our recruitment target. Assuming we will recruit equally between the three areas, we need to approach 66, randomise 33-34 and have 16-17 available per area for analysis.
See Figure 1: Trial flowchart.
As per the 2013 Office of National Statistics data described earlier, we know that 8.6% of all deaths are home deaths due to neoplasms in those aged over 15. [1] Deaths due to neoplasms are seen as a useful proxy for expected deaths. Therefore, the 3 recruitment areas have the following numbers available per annum – North Wales 653, Gloucestershire 517, Vale of Glamorgan 349.
Statistical analysis: Primary analysis will be concentrated on the feasibility metrics and adherence outcomes based on the thresholds defined in Table 2. There will be limited preliminary analysis of intervention outcomes. Point and 95% confidence interval estimates will be calculated and used to estimate variability and direction of effect to further inform the sample size calculation for a definitive study.
Summary statistics of all outcomes will be used to inform the approximate models of analysis that would be used in a full trial. Models will be specified once the data is better understood through the feasibility trial, (e.g. numbers of episodes where as-needed medication used, proportion of participants that never required as-needed medication). A preliminary analysis of the outcomes will be completed using an intention to treat approach. All analysis undertaken will be pre-specified in a statistical analysis plan that will be written and agreed before data collection is completed.
As this is a feasibility trial there will be no imputation of missing data. Missing data will be considered as a criterion for assessing the suitability of measures. Descriptive statistics will be produced for each of the outcome measures, to evaluate the appropriateness of the measures for inclusion in a definitive RCT.
Progression to full trial: Clear progression rules are defined to determine whether an application for a future substantive trial powered to study effectiveness and cost effectiveness should proceed. Our progression rules will relate to the following measures; which we considered important to feasibility:
- reaching our target (16-17) for the number of patients recruited per site.
We have also established clear assessment criteria for establishing the acceptability of the potential primary outcome measures.
The table below summarises the objectives, action plan and criteria for progression to a full trial.
9. Governance
Trial governance procedures adhere to the NIHR guidelines and include a Trial Management Group (TMG), an independent Trial Steering Committee (TSC) and an independent DMEC. SAEs will be reported to the TSC and DMEC in line with NIHR guidance.[56]
To conform to the Data Protection Act 2018 and the General Data Protection Regulation (GDPR) (Regulation [EU] 2016/679), all data will be anonymised and stored securely. No published material will contain patient identifying information.
Peer review: This protocol has had high-quality (independent, expert and proportionate) peer review through the NIHR HTA funding application process. The independent members of the TSC and DMEC will provide an element of continuous peer review.
Bangor University is sponsoring the study, and the Head of School, School of Healthcare Sciences is acting for and on behalf of the Study Sponsor.
10. Quality Assurance and Quality Control
Monitoring, Audit & Inspection
A Trial Monitoring Plan will be developed and agreed by the TMG and TSC based on the trial risk assessment. Site monitoring will be done by performing site visits (at least once per site, with a specific focus on consent recording and handling of data and site files) as well as remotely by exploring the trial dataset.
The sites will be expected to assist the sponsor in monitoring the study. These may include hosting site visits, providing information for remote monitoring, or putting procedures in place to monitor the study internally. Monitoring will be conducted across all sites, and will include a focus on enrolment rates, numbers of withdrawals, and numbers of reported Adverse Events (AEs).
Responsibilities for monitoring will be defined and documented in the Trial Monitoring Plan.
11. Data Handling
Procedures are in place to protect participant confidentiality before, during and after the trial.
Data collection tools and source document identification: Source data will be captured on paper at the relevant time points. A study specific MACRO database will be developed to allow researchers to enter data online. MACRO allows controlled access to the data by all centres and stores a full audit trail. The electronic data captured in the MACRO database will be stored on servers maintained by Bangor University and will be subject to the university IT disaster recovery procedures.
Access to Data and data management: Paper data at sites will be stored in locked filing cabinets separately from identifiable participant data. Access to the MACRO site will be secure and password controlled.
Access to MACRO will be defined on two different levels, access to input (researchers at sites) and access to full data set which will be limited to those core team members involved in data and trial management.
A detailed data management plan will include the definition of the data quality checks that will be performed on the data throughout the life course of the trial. These will include source data validation, random data checks and timelines for data entry.
Access to the final trial dataset: The trial statisticians will have full access to the dataset. The Chief Investigators (CIs) and trial manager will have access to the full dataset after the analysis has been completed. The DMEC will have access to the full dataset as required. The TSC will have access to the full dataset prior to the individual sites having access.
Data sharing: During the course of the trial data sets may be requested from the trial team. A data request form will form part of the data management plan and will document the approval and retrieval process for data sets during the conduct of the trial. All requests will have to be approved by the CIs. All data requests and data sets issued will be retained for completeness.
Data archiving: Archiving of trial documents will be authorised by the Sponsor following submission of the end of study report. As per the sponsor’s research data management policy, research data and records will be retained “for as long as they are of continuing value to the researcher and the wider research community, and as long as specified by research funder, patent law, legislative and other regulatory requirements. The minimum institutional retention period for research data and records is five (5) years after publication or public release of the work of the research, unless required by the funder to retain for longer.”[57]
In line with legal requirements, trial documents will be archived centrally at a secure facility with appropriate environmental controls and adequate protection from fire, flood and unauthorized access. Archived material will be stored in tamper-proof archive boxes that are clearly labelled. Electronic archiving will be provided by the sponsor for post-project deposit and retention of data. Destruction of essential documents will require authorisation from the Sponsor.
12. Publication Policy
Dissemination plan: The results of the study will be first reported to trial collaborators. The main report will be drafted and agreed by the trial co-ordinating team and the final version will be agreed by the HTA before submission for publication, on behalf of the collaboration.
The study findings will be disseminated through publication in highly cited and open access peer reviewed journals and submissions to national and international conferences. In addition, dissemination of our work to clinical and academic colleagues will be via professional societies, newsletters, existing networks and professional web-sites. Relevant NHS organisations and healthcare providers e.g. Clinical Commissioning Groups and National Institute for Health and Care Excellence (NICE) will be informed of the study outcomes.
All carer participants, if they so wish, will be sent an accessible summary of the findings from the study within six months of study completion. The same summary will be made available to public/patient forums to inform patient groups across the area.
It is expected that the TMG will ensure a high level of awareness of our work in the relevant media whilst exploring the use of social media to disseminate outcomes, encourage public/patient involvement and promote future research to improve patient care at the end of life.
Authorship eligibility: Authorship (individually named or group) on the final trial report and manuscripts submitted for publication will be in accordance with the authorship criteria defined by the International Committee of Medical Journal Editors.[58]
13. Indemnity
Bangor University has appropriate Clinical Trials Indemnity and Professional Indemnity insurance in place that will cover members of the research team to conduct the research as per protocol. Health and Care Research Wales staff has NHS contracts and will be responsible to ensure that their work is appropriately insured. NHS staff working with patients involved in the intervention will not be expected to do anything that is not covered by their contracts and will remain covered by the NHS insurance arrangements.