Appendix 6Project 2: evidence submitted to the Academy of Medical Sciences for the review on research governance
The following case study was submitted as evidence to the Academy of Medical sciences for the review on research governance.162
Case study: cluster randomised trial evaluation of a patient- and carer-centred system of longer-term stroke care (Longer-Term Stroke care trial)
Introduction
The LoTS care trial is part of a NIHR-funded programme of research that aims to improve outcomes after stroke by addressing the longer-term needs of patients and carers living at home in the community. It is a cluster RCT evaluating the clinical effectiveness and cost-effectiveness of a new system of care (LoTS care intervention) in comparison to usual practice, as delivered by SCCs, for 800 stroke patients (with and without carers) living at home in the community. The system of care involves the use of a structured assessment linked to a manual containing reference guides and evidence-based treatment algorithms to identify individual patient and carer needs and create action plans for those patients and carers.
Trial organisation
The trial is managed by the Bradford Teaching Hospitals NHS Foundation Trust and the University of Leeds and is sponsored by the Bradford Teaching Hospitals NHS Foundation Trust. It is adopted by the SRN (prior to the NIHR co-ordinated system for gaining NHS permission). The trial involves 32 services across the UK, each of which is composed of a SCC-led community service and the hospitals that refer patients to them, such that 62 separate trusts are involved. After all approvals are in place, SCC services are randomised to either the LoTS care intervention or the control (usual practice) (cluster randomisation); after a training period for intervention services to embed this practice into their service, services then use the practice to which they have been randomised with all patients referred to them throughout the duration of the trial. Patients are recruited to the trial in hospital (for the majority of patients) or in the community by a stroke researcher who is separate from the clinical team, before their first appointment with the SCC service. Patients (and carers) who consent to the trial have a baseline assessment undertaken by the researcher and are asked to complete questionnaires assessing psychological and functional outcomes at baseline and 6 and 12 months after recruitment.
Trial timelines
The NIHR-funded research programme runs for 5 years from October 2007 until September 2012. The initial application for ethics approval for the trial was made in January 2008 and the first patient was recruited in July 2009 (phase 1 services) and February 2010 (phase 2 services). The steps leading up to the first patient being recruited are summarised in and are discussed in more detail below. Recruitment of 800 patients is required to be completed by the end of 2010, with 12 months’ follow-up completed in January 2012, allowing time for analysis and writing up of results by the end of the programme. The original programme timescale was 55 months; however, this was extended to 60 months to accommodate the protracted set-up and approval times at the start of the trial.
Longer-Term Stroke care trial: summary of timelines to first patient recruited
Ethics approval
The review process was efficient (1–2 months). However, approval was delayed because of initial rejections, which appeared to be based on a lack of understanding of the cluster randomised design of the trial. In addition, separate ethics approvals were required for Scotland and the rest of the UK because of the inclusion of adults lacking capacity.
Local set-up
This involved establishing interest and the suitability of SCC services and the hospitals from which they received referrals, determining which acute trusts and primary care trusts are involved, identifying principal and co-investigators and preparing R&D applications. In the majority of services the principal investigator was required to take responsibility for the trial over the whole service, with co-investigators at each of the sites involved. In some services, the principal investigator was able to take responsibility for the trial only at an individual trust, resulting in more than one principal investigator for the service. The intervention is delivered in the primary care trust whereas the majority of patients are recruited from acute trusts, so agreement was required as to who would take responsibility. Delays were caused by differing trust policies regarding who could be principal investigator; many acute trusts would accept principal investigators employed by their own trust only and others required that the principal investigator must be a consultant, despite this being a non-CTIMP (Clinical Trial of an Investigational Medicinal Product) study delivered by stroke research nurses and SCCs. There were also delays in the issuing of honorary contracts for those principal investigators not employed by the acute trusts. Set-up times ranged from 0.5 to 12.2 months, with a median time of 4.1 months. Set-up times for individual services are given in and .
Longer-Term Stroke care trial: timelines for set-up and R&D approval (phase 1)
Longer-Term Stroke care trial: timelines for set-up and R&D approval (phase 2)
Research and development approval (NHS permission)
Sixty-nine R&D approvals from 62 trusts were required to cover the 32 SCC services involved in the trial. The number of trusts involved in a service ranged from one to four. Because of the unpredictability of R&D approval timelines, it was required that all approvals at a service had to be in place before that service could be randomised. The randomisation had to be staggered into two phases according to the time taken to obtain R&D approvals, with 14 services in phase 1 and 18 services in phase 2.
All R&D applications were submitted after ethics approval (Scotland or the rest of the UK as appropriate) was in place. The time taken to receive all R&D approvals for a service ranged from 1.3 to 10.4 months (median 2.8 months), with R&D approvals from individual trusts taking between 0.5 and 10.4 months (median 2.5 months). In total, 42% of trusts gave approval in < 2 months, 35% in 2–3 months, 10% in 3–4 months and the remaining 13% in 7–10 months. R&D approval times for individual services and trusts are given in and .
Common problems that were encountered included:
lack of understanding of the cluster randomised trial design
changes requested to the model non-commercial agreement (Schedule 2: division of responsibilities) and delays in sign-off
local requirements for documentation in addition to the standard forms and documents
reluctance to provide approval without a named stroke researcher; in some cases this was necessary as the SRN/trial funds could not provide a researcher until approval was in place
insistence that consent could be taken only by a physician rather than by a stroke researcher
requirement to provide documentary evidence of indemnity despite the trial having a NHS sponsor
pressure from R&D departments (and the NHS sponsor) to prevent researchers from implementing ethics-approved amendments until R&D departments have also approved the amendments, and delays in approving amendments by R&D departments.
Conclusions
The LoTS care trial is a publicly funded trial with the potential to improve the longer-term care of stroke patients and carers. Substantial delays were caused in the time to first patient recruited by the time taken to obtain R&D approvals from the large number of trusts involved and local policies disproportionate to the risk of the study. It is suggested that a target of R&D approval within 2 months be implemented. Of the 62 trusts involved in the LoTS care trial, 77% gave R&D approval within 3 months and 42% within 2 months, suggesting that 2 months is an achievable target, in addition to bringing R&D approval in line with ethics and Medicines and Healthcare products Regulatory Agency approval timelines.
