Acceptability of the method of administration of a patient-reported outcome measure (PROM) with stroke survivors, a randomised controlled trial protocol

This study is a UK four- arm, pragmatic, multicentre, non-inferiority randomised controlled trial of the method of administering the PRHSQs at the 6-month post-stroke time point (clinically confirmed stroke diagnosis ≥ 4 months to ≤ 8 months). The ratio of allocation is 1:1:1:1, with face-to-face being the method of administration against which the non-inferiority of the three comparator methods (online, postal and telephone) will be discerned. Participants will be randomly allocated to one of four methods of administering these 15 Patient-reported Health Status Questions at eligibility and asked to consent via the post. By virtue of the fact that any of the four methods under investigation are established practices in UK stroke services, participants will not be informed about the other possible modes of administration. Addressing this research question requires implementation of a Zelen design [19] this acknowledges the issues of contacting patients prior to obtaining consent. The impact of this design is due to the nature of the intervention, approaching participants later than usual, as well as a larger-than-usual sample size. We acknowledge that informing participants of the alternative methods of administration may introduce bias, as the rate of return of the PRHSQs for the different assessment methods is the primary outcome of interest. Therefore, participants will knowingly consent to receive and complete the PRHSQs via a single method of administration. See Fig. 1 for the participant flow diagram The study protocol was written in full compliance with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) 2013 [20] and a completed SPIRIT Checklist and Figure 2013 (Fig. 2) [21] are available as a supplement (Additional file 1).

The study has gained acceptance by the Health and Care Research Wales (HCRW) Clinical Research Portfolio, thereby allowing for National Health Service (NHS) organisations in all four devolved nations, via the UK Clinical Research Network Portfolio of studies, to view the basic study design and to approach the central research team for permission to open as a research site and conduct the research. The researchers will conduct an open-door policy and will offer the study to any capable NHS organisation who expresses an interest, based on the following site inclusion criteria: (1) patients who meet the study inclusion criteria are routinely under the care of the principle investigator (PI) at the research site and (2) the research site is able to facilitate all four arms of the study.

The study will be undertaken in 14 centres across England and Wales. Centres acting as research sites will be those who were previously involved in potentially eligible participants’ post stroke care. In line with the time-frame of the study, the majority of potentially eligible participants will reside in the community and will be anticipating routine clinical follow-up from the centres acting as research sites.

Using a non-inferiority margin of 2.5%, with a power of 80% and 2.5% significance level and expecting the same response rate in each group, a minimum of 202 patients are required per randomised allocation group. Thus, a minimum total of 808 randomised eligible participants are necessary for statistical analysis. No adjustment will be made for carrying out three pair-wise tests. Secondary outcomes may not necessarily be powered.

The sequence was generated using a varying-sized permuted block design, stratified within centre. The sequence was concealed by the central research team on a secure system. The allocation sequences are concealed from analyst and central research team until planned unblinding following interpretation of the study findings. Participants and research staff at the research sites will not be blinded; however, allocation will be random as to not introduce bias.

Potentially eligible participants are randomised and allocated centrally by the research team via e-mail (sent within a secure e-mail network) for the random allocation of potentially eligible participants identified solely via a completely pseudonymous identifier (10-digit NHS Number). Research sites will then receive the 10-digit pseudonymous identifier, a 6-digit Participant Research Number (the first two digits being a site identifier and the subsequent four digits referring to the participant) and the method assigned to. The use of a pseudonymous participant research number is to aid compliance with data governance and to ensure blinding of the participants to the central research team.

Potentially eligible participants who have been found eligible by a member of their own clinical care team acting as principal investigator (PI) at the research site, will firstly be randomly allocated to each of the four arms of the study. Following random allocation, potentially eligible participants will be asked to consent by the PI to the research study based on the method of administration they have been randomly allocated. The potentially eligible participants will receive an Invitation Letter, a Participant Information Sheet and a Consent Form via the post. The participant facing documentation will invite the participant to consent to the study and receive their 6-month review and the PRHSQs via one of the four methods they have previously been randomly allocated. Additionally, a proxy consent option is noted in the Participant Information Sheet and Consent Form, thereby allowing for a designee to offer consent on behalf of the participant via the principles of informed consent if the participant requires physical assistance to complete the questionnaire.

Following the initial contact, participants who have not responded will be contacted a second time; two working weeks following the dispatch of the first letter. The second contact will contain the same information as the initial contact (Invitation Letter, Participant Information Sheet and Consent Form). Potentially eligible participants will be given a further four working weeks to reply to the second invitation to participate. If no consent is returned, the potentially eligible participant will be assumed not to wish to take part and will not be asked to complete the PRHSQs.

For all potentially eligible participants who are randomised and invited to participate in the study, we will collect a range of background demographic data and a small number of routinely collected health-related data. See Table 1 for details of data for collection.

All demographic data and routinely collected clinical data will be stored against a Participant Research Number in the research data base. Personally identifiable information will only be stored at the research sites and will not be stored for those who do not offer informed their consent before the cut-off period of 4 weeks following the posting of the second letter Therefore, no personally identifiable information will be held by the central research team.

All consented participants will receive the PRHSQs. The PI at site as a member of both the research team and the participant’s own care team will be responsible for arranging or delegating the completion of the questionnaires between 4 and 8 months post stroke. There are schema for each method by which the PI must abide; however, the methodologies for all four methods of administration differ slightly and are as follows:

Data will be collected at the research sites until February 2018. Potentially eligible participants who do not respond or consent will not be excluded from data analysis of their demographic and routinely collected non-identifiable clinical data as this data is essential for the determination of acceptability of the methods under investigation. However, individuals who do not consent will not receive the PRHSQs.

Two levels of data management exist for this research study. The first level is that of the individual research site which will have access to identifiable patient data. This will allow the site to invite potentially eligible participants and to contact participants who consent to receive the PRHSQs via the method randomly assigned. Following completion of the PRHSQs the anonymous data received from the two self-administered questionnaires (postal or online) will be populated in the research data base against the corresponding Participant Research Number as detailed on the returned PRHSQs. The data from the two administered methods (face-to-face or telephone) will be captured via anonymised PRHSQs, identified solely by the Participant Research Number.. Data is to be initially coded and entered into a site level data base. The data base has been developed using Microsoft Excel 2013 to support ‘data validation’ and will not accept any data entered outside pre-defined ranges for each discrete data value to be collected. Additionally, ‘conditional formatting’ will be utilised to ensure that double entry is minimised..

The second level of data management is that of the central research team. No personally identifiable information will be transferred or held in the central research data base. Research sites will transfer data to the research team using the Participant Research Number as the unique discriminator. Data quality will be ensured by the trial statistician (Dr. Ben Carter, Senior Lecturer in Biostatistics, King’s College London) by conducting central data verification prior to data base lock. Data queries arising from data verification procedures will be submitted to sites in writing and sites will have two working weeks to respond in full.

All data will be handled by research sites and by the central research team in accordance with the Data Protection Act (1998) [23]. Subsequent to the closure of the study all trial materials will be archived for a period defined by Good Clinical Practice (ICH-GCP). The study will be conducted in accordance with the Declaration of Helsinki.

Site monitoring visits will occur throughout the conduct of the study. These monitoring visits will be conducted by members of the central research team and will ensure the correct conduct of the study and will identify any protocol deviations. Furthermore, site monitoring visits will support sites and will offer the opportunity for the central research team to improve the conduct of the study by identifying needs at the research site and offer any clarification or training necessary. The principal investigators (PIs) at research sites are responsible for monitoring protocol deviations and sites are required to report all deviations via the protocol deviation form. All deviations will be monitored and assessed by the central research team. Following assessment, corrective and preventative actions will be disseminated to the sites.

All harms will be reported utilising an adverse and serious adverse event (AE, SAE) form included in the research site file, which will be disseminated to all research sites during site initiation visits. We do not anticipate any trial-related AEs; however, sites will collect and report all AEs and SAEs to the central research team within 24 h of discovery as per the Health and Care Research Wales guidelines. In the event of a trial-related AE or SAE the research site will escalate this to their local NHS R&D department, National Research Governance Organisation and to the chief investigator (CI). Additionally, all PRHSQs can be unblinded via the research sites in the rare event that a concern for patient safety or the safety of others is identified.

Continuous data will be summarised descriptively, and also with a general linear model, and adjusted for patient age and sex. Dichotomous data will be summarised descriptively and will analysed using a logistic regression adjusted for: patient age and sex.

We note that many data will be gathered relating to the questionnaire itself. This useful clinical data warrants publication in its own right; however, it is not the focus of this study and will form a separate stand-alone analysis.

We will be reporting using a complete case analysis. Due to the nature of the research question, we are expecting missing data, and anticipate that this will be not missing at random. We will describe patterns of missing data. However, the primary objective of this study is to ascertain if the missing data is systematically more likely in the comparator groups, compared to the face-to-face allocation group.

Utilising unpublished routinely collected clinical data approximately 85% of follow-up appointments for all patients, who would meet the inclusion criteria, return for face-to-face administration of their 6-month post-stroke follow-up assessment at the ≥ 4-month to ≤ 8-months post-diagnosis time point. The reason for this high level of face-to-face follow-up is not grounded on evidence, however, disability and communication issues post stroke may have a role in the choice of this method. Nevertheless, there is a significant cost implication in the choice of method for follow-up as evidenced in a study by Lannin et al. (2013) investigating the cost-effectiveness of telephone vs mail methods of follow-up for the Australian National Stroke Registry [10].

Alongside the National Institute for Health and Care Excellence (NICE) accredited RCP stroke guidelines [25], a non-inferiority margin of 2.5% was decided upon as the maximum tolerated acceptable reduction in acceptability, alongside conflicting resource allocation and offering flexibility of patients to undergo follow-up.

The study is to be overseen by a Steering Committee composed of Dr. Manju Krishnan (Clinical Lead and Consultant Stroke Physician Abertawe Bro Morgannwg University Health Board and Deputy HCRW Stroke Research Lead) as chair, lay members, independent clinical and research specialists and key stakeholders. The Steering Committee will be responsible for examining research progress, monitoring progress against established timelines and monitoring recruitment, completion and follow-up rates. Meetings will take place quarterly and all recommendations arising from the Steering Committee will be acted upon immediately. Moreover, the Steering Committee will reserve the right to request and receive interim reports between quarterly meetings. The nature and design of the trial does not necessitate the creation of an additional Data Management Committee to supplement the work of the Trial Steering Committee. Therefore, the Trial Steering Committee will have oversight across the study, including data monitoring and data management.

We plan to disseminate the findings of the study via a peer-reviewed academic journal. The findings of which will be cascaded to all research sites and disseminated from the research sites to participants. The published findings will adhere to the Consolidated Standards of Reporting Trials (CONSORT) randomised control trial reporting guidelines [26, 27]. In addition, the published findings will adhere to both the CONSORT extension relating to non-inferiority and equivalence trials [28] and the extension relating to pragmatic trials [29].

Recruitment commenced in July 2017 and is planned to continue until February 2018. Protocol Version: 1.0 Date: 10 April 2017.