Functional strength training versus movement performance therapy for upper limb motor recovery early after stroke: a RCT

Article history

The research reported in this issue of the journal was funded by the EME programme as project number 10/60/30. The contractual start date was in April 2012. The final report began editorial review in January 2017 and was accepted for publication in October 2017. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The EME editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

none

Copyright statement

© Queen’s Printer and Controller of HMSO 2018. This work was produced by Pomeroy et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2018 Queen’s Printer and Controller of HMSO

Objectives

The scientific driver for this trial was that detailed understanding of the interaction between the treatment and each patient’s residual neural function will more likely enable physical therapies to be targeted at recovery mechanisms in those stroke survivors most likely to respond. The specific objectives were to:

1. determine if upper limb motor recovery is enhanced more by FST + CPT than an equal dose of MPT + CPT commenced early after stroke

2. identify the similarities and differences in the neural correlates of clinical improvement in upper limb motor function in response to (1) FST + CPT and (2) MPT + CPT

3. determine whether or not any pretreatment neural characteristics or combination of (1) anatomical location of infarction, (2) volume of the stroke lesion, (3) residual structural corticocortical connectivity, (4) residual corticospinal connectivity and (5) brain–muscle functional connectivity [derived from Transcranial Magnetic Stimulation (TMS)], are sufficiently predictive of upper limb recovery after stroke to enable physical therapy to be targeted at those people most likely to respond.

Achieving these objectives before undertaking a Phase III randomised controlled trial (RCT) conforms with the Medical Research Council (MRC) Framework for Design and Evaluation of Complex Interventions to Improve Health. 40

Design

A randomised, controlled, observer-blind, two-group, multicentre trial. The randomisation order was generated before the trial began and allocation was via an independent telephone interactive voice response randomisation service. Clinical efficacy measures were made before randomisation (baseline), the working day (± 7 days) after the 6-week intervention ended (outcome) and 6 calendar months (± 14 days) after the index stroke (follow-up). Neural measures were made at baseline and outcome within 10 days following the clinical efficacy measures. All assessors were blinded at baseline as randomisation had not yet occurred. At outcome and follow-up all assessors were, where exceptional events did not dictate, blinded to the randomisation. When at all possible, randomised participants were included in outcome and follow-up measures. The trial procedure is illustrated in Figure 1. The trial is reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines and covers all the CONSORT checklist items.

FIGURE 1.

Flow chart to illustrate trial procedure.

Trial registration

The trial was registered on Current Controlled Trials with the unique identifier ISRCTN19090862 (www.controlled-trials.com). In addition, the protocol was published41 and the study was adopted by the UK Clinical Research Network and, therefore, appeared on the UK Clinical Research Network Study Portfolio.

Ethics and research governance approval

The Norfolk Research Ethics Service provided ethics approval on 22 February 2012 (reference number 11/EE/0524). All participants provided informed consent.

The trial was outside the scope of the Clinical Trials Directive defined by the Medicines and Healthcare Products Regulatory Agency as neither trial intervention included provision of a pharmacological compound.

The University of East Anglia (UEA) was the recipient of the funding. UEA together with the University of Glasgow Clinical Trials Unit [Robertson Centre for Biostatistics (RCB)] and the Norwich Clinical Trials Unit was responsible for the organisation and running of the trial. UEA had subcontracts with each partner university that laid out the delegated responsibilities of the partner and funding to be provided. Each partner university was responsible for procuring suitable management and governance provision with their local NHS trust centres. All members of the trial team had current good clinical practice (GCP) training. Those members of the trial who were actively involved in clinical contact with participants but were not employed by the NHS had either honorary clinical contracts or research passports.

All decisions regarding eligibility for entry, provision of written informed consent, inclusion, exclusion and attrition were documented as per the MRC Code of Good Practice in Clinical Trials and the CONSORT guidelines. Relevant trial documentation will be retained for a period of 10 years after the end of data collection to comply with the GCP regulations and to ensure availability of data for any subsequent systematic reviews and meta-analyses. The UEA will archive trial documents in a secure facility. The custodian will be Professor Pomeroy.

Participants

The combined inclusion and exclusion criteria for potential participants were people:

1. with a clinical diagnosis of stroke in the territory of the anterior cerebral circulation, cortical and/or subcortical as corroborated through routine clinical imaging

2. aged ≥ 18 years

3. who were between 2 and 60 days after stroke when providing informed consent

4. were able, before the index stroke, to use the paretic, contralesional, upper limb to lift and then drink from a cup

5. who were defined as medically stable as confirmed by the stroke service medical team responsible for the individual’s stroke care

6. with enough voluntary muscle contraction in the paretic upper limb to score at least 11 of the 33 points on the Motricity Index pinch section42

7. who were unable to complete the Nine-Hole Peg Test (9HPT) within 50 seconds41

8. with a score of 0 or 1 on the Extinction and Inattention subscale of the National Institutes of Health Stroke Scale (no obvious spatial neglect)

9. who could imitate action with the non-paretic (ipsilesional) upper limb. The research therapist sat alongside a potential participant to demonstrate five upper limb activities. The potential participant was asked to watch and then perform the activities. The accuracy of imitation was assessed on the 3-point scale used by Decety: 2 = correctly reproduced action, 1 = incorrectly reproduced action and 0 = not reproduced. 43 Those people scoring 8 out of 10 or above were considered as able to imitate and, therefore, eligible for inclusion in this trial.

Settings for recruitment, assessment and treatment

• Birmingham (centre 1).

  • Moseley Hall Hospital, Birmingham Community Healthcare NHS Trust.

  • Queen Elizabeth Hospital, University Hospitals Birmingham NHS Foundation Trust.

  • University of Birmingham Imaging Centre.

  • Participants’ own homes.

• Norfolk (centre 2).

  • Acute Stroke Unit, Norfolk and Norwich University Hospital NHS Trust.

  • Stroke Rehabilitation Ward, Mulberry Unit, and Early Supported Discharge Team, Norfolk Community Health and Care NHS Trust.

  • James Paget University Hospitals.

  • The Movement Analysis Laboratory at the UEA.

  • Participants’ own homes.

• North Staffordshire (centre 3).

  • Haywood Hospital, Staffordshire and Stoke-on-Trent Partnership NHS Trust.

  • University Hospital of North Midlands NHS Trust.

  • Participants’ own homes.

Screening and recruitment of participants

Potential participants were screened from the three centres and followed up on the condition that research governance and ethics approvals were in place for that setting. A two-stage screening process was used because some study criteria required movement assessment expertise that was not always present in the Comprehensive Local Research Network (CLRN) screening teams.

Stage 1 of the screening process was conducted by CLRN research nurses in liaison with stroke service clinical nurses and therapists. They screened for study criteria (a)–(f) as described in Participants (Figure 2). For some potential participants, the screening happened more than once. This was because medical and functional characteristics change over time after stroke so if somebody was unsuitable on one occasion it was possible that they could change and meet the trial criteria later. An individual’s progress was monitored until eligibility for this trial was clear (Figure 3). At the end of stage 1, informed consent was taken by either CLRN nurses or research therapists (enrolment). All people screened were categorised as unsuitable, refused informed consent, or provided informed consent.

FIGURE 2.

Stage 1: initial screening.

FIGURE 3.

Stage 1: participant eligibility pathway. Dark green indicates clinician or research network colleague, light green indicates research network colleague or researcher, dark blue indicates researcher and light blue indicates participant.

People providing informed consent undertook stage two of the screening process with a research therapist to ascertain concordance with study criteria (g), (h) and (i) (Figure 4). Those people who did not meet all three of the inclusion criteria were reassessed later if they agreed and the exclusion time limit had not been met. At the end of stage 2 individuals were categorised as unsuitable for the trial or as a recruited participant.

FIGURE 4.

Stage 2: post-consent suitability assessment. Dark green indicates researcher and light green indicates participant.

Randomisation

Before the trial began, the computer-generated randomisation sequence was created by an independent statistician at the Glasgow Clinical Trials Unit. Participants were allocated to FST + CPT or MPT + CPT in a 1 : 1 ratio, stratified by (1) time after stroke (up to 30 days or 31–60 days); (2) ability to use the paretic upper limb as assessed by the 9HPT, moving one peg or less in 50 seconds, or moving 2–8 pegs in 50 seconds; and (3) clinical centre. An independent telephone randomisation service was used to conceal group allocation order from investigators, research therapists and blinded assessors prior to the randomisation of a participant. The telephone randomisation service was contacted by the team member (research therapist or blinded assessor) immediately after they had completed the baseline measures with a participant.

Sample size and power

The sample size calculation that was used accounted for the expected clustered data structure (patients within therapist within treatment group). 44 This sample size calculation was based on actual ARAT score data from our previous early-phase trial28 and the expectation that the intraclass correlation coefficient (ICC) would be somewhat lower than 0.05 for patient outcomes. 45 On the assumption of an ICC of 0.01 in both treatment arms and three centres with a separate therapist for each randomised arm, a sample size of 99 participants per group would provide 80% power to detect a clinically important mean difference of 6.2 points in ARAT score change for data analysis using a two-sample t-test, with Satterthwaite correction, applying a 5% two-sided significance level and allowing for potentially different standard deviations (SDs) in the CPT + MPT (SD 7.9) and CPT + FST (SD 19.3) groups.

To account for clustering in the design (participants within a therapist within randomised treatment at each study site), a sample size inflation factor 1 + (m – 1) × ICC was applied, where m is the cluster size and ICC uses SSC software version 1 (Health Services Research Unit, University of Aberdeen, Aberdeen, UK). On the assumption that recruitment would be distributed evenly across therapists, the sample size was therefore inflated to 129 evaluable participants per group. To allow for an attrition rate of 10% (7% in our previous single-centre trial)28 we recruited 144 participants per group. The total sample size was 288.

Description of interventions

Measurement battery

Adverse reactions and adverse events

Participation in FST and MPT is considered low risk for adverse events (AEs). However, there is a small possibility that either therapy could be associated with an overuse syndrome, expressed as experience of pain or fatigue. These were therefore classified as potential adverse reactions as follows:

• Pain was considered an adverse reaction if (1) a participant reported onset or increase of paretic upper limb pain (verbally or behaviourally), (2) the pain was sustained over four consecutive therapy sessions and (3) the clinical team were unable to account for this in any other way than involvement in the trial. If pain occurred then the research therapist adjusted the trial therapy as appropriate or, if indicated, stopped it on either a permanent or temporary basis. The date of the adverse reaction was recorded as the date of the fourth therapy session.

• Fatigue was considered to have occurred if (1) a participant demonstrated a decrease of two levels in the Motricity Index upper limb score on two consecutive therapy sessions and (2) the clinical team could not account for this in any other way than involvement in the trial. This was addressed by the therapist adjusting the trial therapy as appropriate or, if indicated, stopping the trial therapy on either a permanent or temporary basis. The date of the adverse reaction was recorded as the date of the second therapy session.

To report serious adverse events (SAEs), the trial team used the Norfolk and Norwich University Hospital and UEA joint standard operating procedure (SOP). The latest version can be found www.nnuh.nhs.uk/Dept.asp?ID%20=%20681 (see SOP 205). All SAEs were assessed by the local principal investigator, countersigned by the chief investigator, reported to the sponsor and reported to the Data Monitoring and Ethics Committee (DMEC) and Trial Steering Committee (TSC). All SAEs were followed up until a documented end date and resolution could be provided, or the participant ended the trial. Participants were considered to have reached the end of the trial when the first of the following occurred:

• completion of assessment at 6 months after stroke

• withdrawal of consent

• SAE resulting in withdrawal of participant or death

• loss to follow-up.

If a SAE was still ongoing at the time a participant reached a trial end point, their trial end date was also used to record their SAE end data.

Statistical analysis

The statistical analysis plan (SAP) was agreed, signed and dated prior to the database lock and unblinding of the treatment allocations (see Appendix 4).

There was a change to the original analysis plan from taking account of clustering by therapist to only adjusting for the study site. The change was made because, for practical reasons, the participants were not clustered within the same therapist for all their sessions.

Trial management

Objective addressed

To determine if upper limb motor recovery is enhanced more by FST + CPT than an equal dose of MPT + CPT commenced early after stroke.

Dates of recruitment and follow-up periods

Screening for potential participants started on 1 April 2012 and ended on 25 January 2016. The first participant was randomised on 17 October 2012 and the last participant was randomised on 29 January 2016. The last follow-up assessment, at 6 months after stroke, was 6 July 2016.

Participant characteristics

The characteristics of participants are detailed in Table 1. In summary, the mean age (SD) of participants in this trial was 72.2 (12.5) years. The sex distribution was 64.6% male and 35.4% female. Informed consent was given within 30 days of stroke by 59% of participants and at ≥ 31 days by 41%. All participants had been diagnosed with stroke in the territory of the anterior cerebral circulation. The stroke was caused by a haemorrhage for 8.7% of participants and by ischaemia for 91.3%. The right-hand side of the body was more affected by the stroke than the left-hand side for 42% of participants.

Flow of participants through the trial

The CONSORT flow chart for this trial in respect to the primary outcome (ARAT score) is provided in Figure 5. In summary, a total of 5064 stroke survivors were screened for eligibility for this trial. Of these, 2929 (58%) were excluded because they did not meet the initial study inclusion criteria, with a further 872 (17%) excluded for additional reasons, such as being outside the trial catchment area or having severe cognitive and/or language impairment. This left 1263 identifiable eligible people; however, 536 (42%) of these people declined to talk to the research team. The remaining 481 (38%) provided informed consent and undertook the trial suitability screening assessment; of those, 138 (29%) did not meet the eligibility criteria and a further 55 (11%) either withdrew consent or otherwise were not randomised. The remaining 288 were recruited as participants and randomised into the trial. Consequently, 288 participants undertook the baseline assessments and were subsequently allocated randomly to the MPT group (n = 143) or the FST group (n = 145).

FIGURE 5.

All centres: CONSORT 2010 flow diagram. a, Data are excluded from analysis if missing items prevent the ARAT total score from being calculated. UL, upper limb.

At the primary end point of the outcome (at the end of the 6-week intervention phase), 245 (85.1%) participants undertook the primary measure (ARAT score), of whom 119 had been allocated to MPT + CPT and 126 to FST + CPT. Attrition reasons were:

• withdrew as a result of a SAE or AE (MPT + CPT, n = 5; FST + CPT, n = 5)

• unwilling to continue (MPT + CPT, n = 6; FST + CPT, n = 6)

• withdrew consent (MPT + CPT, n = 4; FST + CPT, n = 2)

• lost to outcome/other (MPT + CPT, n = 5 FST + CPT, n = 3)

• did not attend for assessment, but remained in trial (MPT + CPT, n = 4; FST + CPT, n = 3).

At 6 months after stroke, the follow-up time point, 208 (72.2%) participants undertook the primary measure (ARAT score), of whom 104 were allocated to MPT + CPT and 104 to FST + CPT. Attrition at the follow-up point had occurred for the following reasons:

• withdrew as a result of a SAE or AE (MPT + CPT, n = 15; FST + CPT, n = 8)

• unwilling to continue (MPT + CPT, n = 13; FST + CPT, n = 11)

• withdrew consent (MPT + CPT, n = 4; FST + CPT, n = 3)

• investigator withdrawal (MPT + CPT, n = 0; FST + CPT, n = 1)

• lost to outcome/other (MPT + CPT, n = 7; FST + CPT, n = 18).

Adverse reactions, adverse events and serious adverse events

No SAEs met the criteria for reporting to the National Research Ethics Service. All SAEs were reviewed by the DMEC. Only four participants experienced an adverse reaction of pain or fatigue during the intervention phase (Table 2).

Adverse events, related AEs, SAEs and unexpected AEs during the intervention phase are detailed in Tables 3–6. No clinically important differences are detectable between the FST + CPT and MPT + CPT groups.

Those AEs, related AEs, SAEs and unexpected AEs that occurred during the follow-up phase are detailed in Tables 7–10. No clinically important differences are detectable between the FST + CPT and MPT + CPT groups.

Primary clinical efficacy analysis

The primary outcome measure was ARAT score. There were 126 participants for the FST + CPT group and 114 for the MPT + CPT group with data at both baseline and outcome (Table 11). At follow-up, the number of participants with data at baseline had reduced to 104 for the FST + CPT group and to 100 for the MPT + CPT group (Table 12). Both groups showed improvement (mean and SD) from baseline at outcome [FST + CPT = 9.70 (SD 11.72) and MPT + CPT = 7.90 (SD 9.18)], but there was little difference between the groups and this did not reach statistical significance (p = 0.298). Further improvements were evident at follow-up with mean (SD) changes from baseline of 11.10 (SD 14.68) for the FST + CPT group and 10.3 (SD 10.74) for the MPT + CPT group. Again, this difference was small and did not reach statistical significance (p = 0.743).

Secondary clinical efficacy analysis

Secondary clinical efficacy outcome measures (Tables 13–18) showed a similar pattern to the ARAT scores (primary outcome) in terms of their mean improvements. SDs indicated more variability in response for grip and pinch force. Differences between groups were small and did not reach statistical significance.

Objective addressed

To identify the similarities and differences in the neural correlates of clinical improvement in upper limb motor function in response to (1) FST + CPT and (2) MPT + CPT. To reiterate, baseline measures were made immediately before randomisation and outcome measures made at the end of the 6-week intervention phase (see Chapter 2, Design).

Flow of participants through the correlates analysis

Neural variables at baseline and outcome

The numbers of participants for whom neuroimaging variables were available at outcome and follow-up are shown in Tables 21–23. The number of participants varies from 45 for corticocortical and corticospinal anatomical connectivity at outcome (see Table 21) to 84 for volume of the stroke lesion at baseline (see Table 23).

Change in neural variables derived from neuroimaging between baseline and outcome for those participants with data at both visits and ARAT score data at both visits is shown in Table 24. There were no statistically significant differences between the two groups for change in any of these variables.

The number of participants for whom neural variables derived from TMS were available at outcome and follow-up together are shown in Tables 25 and 26. The number of participants ranges between 72 for RMT for paretic extensor carpi radialis (pECR) at outcome (see Table 26) and 110 for presence of a MEP of pBB and pECR at baseline (see Table 25).

Change between baseline and outcome for RMT for pBB and pECR muscles for those participants with data at both visits and ARAT score data at both visits is shown in Table 27. There are no statistically significant differences between the two groups for change in either of these variables.

The correlations between change, from baseline to outcome, between neural variables and ARAT score ranged from r = –0.147 (p = 0.385) for all participants’ corticospinal connectivity to r = 0.199 (p = 0.320) for the MPT + CPT group for RMT pBB (Table 28). Consequently, the neural correlates of improvement in ARAT score were similar for the two groups.

Objective addressed

To determine if any pretreatment parameters or any combination of pretreatment parameters [(a) anatomical location of infarction, (b) volume of the stroke lesion, (c) residual structural corticocortical connectivity, (d) residual corticospinal connectivity and (e) brain–muscle functional connectivity (derived from TMS)] are sufficiently predictive of upper limb recovery after stroke to enable physical therapy to be targeted at those people most likely to respond.

Participants with neural variables data at baseline

Data on the number of participants at baseline and the values of the neural variables are presented in Tables 21–28.

Change at outcome from baseline for Action Research Arm Test score

For those people with both ARAT score and neural variable data at baseline and outcome there were no statistically significant interaction effects (Table 29). The data for change in ARAT score underlying these interaction effects are provided in Tables 30–34. There are no statistically significant differences between the subgroups for any of the baseline characteristics.

Summary of main results

Strengths and limitations

Bias protection for this trial was provided by the blinding of assessors to group allocation, concealment of randomisation order, group allocation via independent telephone randomisation service, adhering to the intention-to-treat principle and reporting all planned outcomes. This trial evaluated a behavioural intervention and, therefore, it was not possible to blind research therapists, participants or clinical staff to the allocated intervention. Although this is perceived as a distinct limitation in drug trials, it is recognised that double blinding is not possible in trials of many stroke rehabilitation interventions. Because of staffing challenges in centres at some points in the trial, it was not always possible to have blinded assessment of the behavioural outcomes.

The sample size for this trial was estimated by a power calculation informed by data from our early-phase trial. 28 The estimated sample size allowed for an attrition rate of 10% at the primary end point of outcome. In the event our attrition rate at outcome was 12.5%, which is substantially lower than the 25% reported by the well-regarded EXCITE (Extremity Constraint-Induced Therapy Evaluation) trial of CIMT. 56

This trial was congruent with the requirements for stroke rehabilitation trials. 40 A strength for all aspects of this trial is that the interventions, both MPT and FST, were described in sufficient detail to enable both research replication of the trial and transferability of results to clinical practice. 48,57,58 Thus, this trial has followed the recommendations for improving rehabilitation research. 57,59

The experimental FST + CPT in this trial was consistent with key principles of neural plasticity9,60 and was therefore well placed to (1) exhibit clinical efficacy and (2) form a useful probe of correlates of, and indicators for, upper limb recovery.

Relationship to previous studies

Progress in stroke rehabilitation has been hampered by a paucity of large-scale projects in which neuroscientists and clinicians have been able to work together and so inform each other’s approach to a single clinical problem, such as the treatment of upper limb weakness. 55 For this trial, we combined neuroscience and clinical science expertise in the largest study of its kind to investigate underlying mechanisms of motor recovery in a representative sample of patients early after stroke with substantial to moderate upper limb motor impairment using well-characterised physical therapies.

Generalisability

The generalisability of this trial is considered in the wide inclusion criteria with direct relevance to clinical practice and the conduct of the trial in three different clinical centres.

Objective 1

The trial reported here found small differences in the clinical efficacy of upper limb recovery between FST + CPT and MPT + CPT, but these did not reach statistical significance. Both groups showed increases in ARAT score (primary outcome measure) above the clinically important change. However, variation around the mean change from baseline scores was substantial in both groups.

Objective 2

The neural correlates of change were similar for the two forms of physical therapy. There were no statistically significant correlations between change in the neural correlates and change in clinical efficacy measures.

Objective 3

The trial reported here found that none of the pretreatment neural characteristics of interest predicted response to either FST + CPT or MPT + CPT.

Implications for health care

The findings of the trial reported here confirm clinical impressions and emerging research evidence of variation in response to specific therapies among people early after stroke.

Research recommendations

There is still an urgent need for evidence to guide decisions about (1) appropriate prescription of physical therapy for individuals and (2) the recovery mechanisms at which physical therapy should be targeted.

Grant co-applicants

• Jane Burridge contributed to production of protocol and reviewed the final version of the report.

• Roger Lemon contributed to production of protocol and reviewed the final version of the report.

• Paulette vanVliet contributed to production of protocol and reviewed the final version of the report.

Research colleagues

• Andrew Walker, formerly School of Health Sciences, UEA. Trial manager who set up the trial and oversaw the initial data collection stage.

• Karla Miller, University of Oxford. MR physicist who supervised the setting up of the neuroimaging protocol across all sites.

• James Kolasinski, University of Oxford. Neuroscientist who carried out the analysis of the neuroimaging data.

Clinical research colleagues

• All of the research support staff in the NHS trusts in which this research was located, with particular thanks to Sue Thompson, Barker Unit, Haywood Hospital (Staffordshire and Stoke-on-Trent Partnership Trust), Stoke-on-Trent, and Pel Fordham at the UEA, who managed and supported the administration and quality of the trial documentation and supported participant liaison for imaging appointments.

• The CLRN staff for support given to trial recruitment, especially Lesley Maloney, Research Manager for the Norfolk Community Health and Care NHS Trust.

• The Norwich Clinical Trials Unit for clinical trial support and guidance.

• The RCB for data handling and statistical analysis.

Trial research therapists and assessors

• Carlo Areddu, Research Therapist.

• Alison Aries, Blinded Assessor.

• Deepa Barnabas, Blinded Assessor.

• Scott Brown, Research Therapist.

• Zoe Carmichael, Research Therapist.

• Elisabeth Cato, Research Therapist.

• Elizabeth Chandler, Research Therapist.

• Kathryn Cogman, Research Therapist.

• Kathryn Collins, Blinded Assessor.

• Ellen Colton, Research Therapist.

• Francine Cox, Research Therapist.

• Hayley Crane, Research Therapist.

• Claire Grocott, Blinded Assessor.

• Jane Hargreaves, Blinded Assessor.

• Claire Havis, Research Therapist.

• Rachel Hodder/Austin, Research Therapist.

• Claire Laxton, Research Therapist.

• Laura Mason, Research Therapist.

• Emma Mccoll, Research Therapist.

• Stephanie Mills, Research Therapist.

• Hannah Morgan, Research Therapist.

• Jen Pearson, Research Therapist.

• Francesca Quinn-Thomas, Blinded Assessor.

• Rosie Salmon, Research Therapist.

• Alyssa Snelson, Research Therapist.

• Jessica Swart, Blinded Assessor.

• Rachel Warren, Research Therapist.

• Laurence Wood, Research Therapist.

• Catherine Worth, Research Therapist.

• Katie Young, Research Therapist.

Clinical colleagues

• Principal investigators not listed elsewhere:

  • Bruce Jarvest, Principal Investigator, Imaging Department, University Hospital of North Midlands (Royal Stoke Hospital).

  • Dr Paul Malcolm, Principal Investigator, Radiology Department, Norfolk and Norwich University Hospital NHS Foundation Trust.

  • Dr Claire Sutton, Principal Investigator, Queen Elizabeth Hospital Birmingham and Moseley Hall Hospital Birmingham.

  • Ingrid Watmough, Principal Investigator, Community Rehabilitation, Norfolk Community Health and Care NHS Trust.

  • Dr Kneale Metcalf, Principal Investigator, Stroke Medicine, Norfolk and Norwich University Hospital NHS Foundation Trust.

• The rehabilitation teams in Staffordshire and Stoke-on-Trent Partnership Trust; Haywood Hospital, Stoke-on-Trent (in particular Sneyd Ward, Broadfield Ward and the Stroke Early Supported Discharge Team; and Janet Butler for providing admission and discharge date information); Moseley Hall Hospital, Birmingham Community Healthcare NHS Foundation Trust; Norfolk Community Health and Care NHS Trust (in particular Beech Ward and Early Supported Discharge); James Paget University Hospital NHS Trust and the Norfolk and Norwich University Hospital NHS Foundation Trust.

• The MRI teams in the Norfolk and Norwich University Hospital NHS Foundation Trust (in particular Richard Greenwood); University Hospital of North Midlands NHS Trust (in particular Dr Peter Wright, MRI and Non-ionising Medical Physicist; Sarah Prescott, Lead MRI Clinical Scientist, Medical Physics; and Harry Poole, Clinical Scientist, MRI Physics); and University of Birmingham Imaging Centre (in particular Nina Salman, Chief Radiographer and Dr Andrew Bagshaw, Scientific Director).

Other acknowledgements

We are grateful to the many other people who made this research possible.

• Pel Fordham, for organising participant appointments, assisting with trial meetings and assisting with writing this report.

• Lynn Aylett, for supporting data processing and returns to the RCB.

• The Research and Enterprise and Finance Departments, UEA and Keele University.

• All the research support staff in the NHS trusts in which this research took place.

Contributions of authors

Valerie M Pomeroy (Professor of Neurorehabilitation, UEA) was the chief investigator and was the principal investigator at the UEA, Norfolk Community Health and Care, and Norfolk and Norwich University Hospitals NHS Foundation Trust. She undertook the overall design of the study, supervised trial conduct, supervised conduct of trial in Norfolk, contributed to recruitment, wrote drafts and undertook the final editing of the report.

Susan M Hunter (Senior Lecturer, Institute for Applied Clinical Sciences, School of Health and Rehabilitation, Keele University) was the principal investigator at Keele University and NHS Staffordshire and Stoke-on-Trent Partnership Trust. She supported design of the study, provided training in intervention delivery to research therapists across all sites, supervised the conduct of the trial in North Staffordshire, contributed to recruitment of participants, provided intervention in the (long-term) absence of research therapists, conducted baseline and blinded outcome assessments in the (long-term) absence of research assessors, collected TMS data locally, contributed to the report and agreed the final version.

Heidi Johansen-Berg (Professor of Cognitive Neuroscience, University of Oxford) contributed to the overall study design, jointly supervised design conduct and analysis of neuroimaging aspects of the study, contributed to the report and agreed the final version.

Nick S Ward (Professor of Clinical Neurology and Neurorehabilitation, University College London Institute of Neurology) led neuroimaging aspects of the production of the protocol and conduct of the trial, contributed to the report and agreed the final version of the report.

Niamh Kennedy (Lecturer in Rehabilitation Neuroscience, UEA) led neurophysiological aspects of the production of the protocol and conduct of the trial, contributed to the report and agreed the final version of the report.

Elizabeth Chandler (Research Therapist, UEA) contributed to all trial operating procedures and development of CRFs, managed day-to-day conduct of the trial in Norfolk, produced recruitment data during the interim period between full-time trial managers, contributed to recruitment and provided experimental intervention, contributed to the report and agreed the final version of the report.

Christopher J Weir (Professor of Medical Statistics and Clinical Trials, University of Edinburgh) contributed to the design of the study, advised on the SAP, commented on drafts of the report and agreed the final copy version of the report.

John Rothwell (Professor of Human Neurophysiology, University College London Institute of Neurology) contributed to neurophysiological aspects of the protocol and oversaw the quality of data collection contributed to the report and agreed the final version of the report.

Alan Wing (Professor of Human Movement, School of Psychology, University of Birmingham) contributed to the production of the protocol, oversaw one of the clinical centres, contributed to the report and agreed the final version of the report.

Michael Grey (formerly Reader in Motor Neuroscience, University of Birmingham; currently Reader in Rehabilitation Neuroscience, UEA) led the conduct of neurophysiological measures in a key centre for this trial and is leading aspects of the analysis, contributed to the report and agreed final version of the report.

Garry Barton (Professor of Health Economics, UEA) led all aspects of the health economics investigation embedded in this trial, contributed to the report and agreed the final version of the report.

Nick Leavey (Clinical Trial Manager, UEA) managed the conduct of the trial across the three recruiting centres, contributed to the report and agreed the final copy version of the report.

Publications

Hunter SM, Johansen-Berg H, Ward N, Kennedy NC, Chandler E, Weir CJ, et al. Functional strength training and movement performance therapy for upper limb recovery early poststroke–efficacy, neural correlates, predictive markers, and cost-effectiveness: FAST INdiCATE Trial. Front Neurol 2018;8:733.

Pomeroy VM, Ward NS, Johansen-Berg H, van Vliet P, Burridge J, Hunter SM, et al. FAST INdiCATE Trial protocol. Clinical efficacy of functional strength training for upper limb motor recovery early after stroke: neural correlates and prognostic indicators. Int J Stroke 2014;9:240–5.

Van Vliet P, Hunter SM, Donaldson C, Pomeroy V. Using the TIDieR checklist to standardize the description of a functional strength training intervention for the upper limb after stroke. J Neurol Phys Ther 2016;40:203–8.

Data sharing statement

All available data can be obtained by contacting the corresponding author.

Patient Data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research. The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, the MRC, NETSCC, the EME programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the EME programme or the Department of Health and Social Care.

FAST-INdICATE training: delivering trial intervention – November 2013, UEA, London

Topics identified for discussion and actions agreed during the training

1. Therapist-directed and participant-administered therapy.

   • Specific equipment that was required for FST would be provided by the research therapist. However, written instructions would not be provided by the therapist, but the participant or carer could make notes to remind them of what exercises they should do, and when/how often they should do them.

   • It was noted that hand dominance and side of hemiplegia potentially impacted on the ability to complete this participant-administered therapy.

   • The presence/involvement of a carer or spouse was likely to be helpful in participants administering their own therapy/exercises.

   • Variability and non-standardisation of this directed therapy was important because of personalised care.

2. Equipment left at participant’s house for directed therapy.

   • There is a potential threat to blinding if equipment that is obviously specific to FST is left at the participant’s home and seen by the clinical team. Its presence may confound the content of CPT, as the clinical team might be tempted to use this equipment in their CPT programme with that participant, thus increasing the dose of FST and changing the nature of actual CPT.

   • There was a suggestion that the content of CPT should be analysed for both groups to see if it is affected by group allocation (i.e. whether or not there is more strength training activity for the FST group), which might be a result of the presence of strengthening equipment in the home.

3. Documenting therapy.

   • It was agreed that the number of repetitions of an exercise/activity in the FST group would not be reported.

4. Progression of dexterity activities in high-functioning participants.

   • This was discussed in the context of generating ideas for progression exercises to improve dexterity in high-functioning participants and the difficulty this presents was acknowledged.

5. Difference between strength and endurance training/activities in FST.

   • Discussed and number of repetitions for strengthening to be based on six repetition maximum, whereas number of repetitions for endurance to be based on three sets of 10–12 repetitions.

6. Use of feedback and the difference between knowledge of performance and knowledge of results.

   • Knowledge of performance is characterised by a focus on movement (body part, direction/plane, e.g. bend, straighten elbow), whereas knowledge of results is characterised by a focus on quantitative measure of goal achievement (e.g. moving an object a set distance).

7. Use of active-assisted exercise/movement and facilitation in FST.

   • Neither active-assisted nor facilitated movement is acceptable as part of FST; however, providing support to a body part in order to minimise the effect of gravity as part of a progressive strengthening regime is acceptable as long as the support is simply supporting the weight of the limb and not assisting with movement.

8. Adverse event reporting.

   • Pain and fatigue are to be monitored at each therapy visit to identify adverse reactions to the therapy:

     • A change in two levels on the Motricity Index that occurs on four consecutive occasions is to be considered an adverse reaction of fatigue (AE on the CRFs).

     • The presence of new pain over four consecutive occasions is to be considered an adverse reaction of pain (AE on CRFs) only if the pain cannot be explained by the clinical team as resulting from a different cause. If a participant reports new pain, the research therapist or PI will contact the clinical team to determine whether the new pain is likely to be cause by the trial intervention or if there is another explanation for that pain. If that new pain is present on four consecutive visits by the research therapist is considered to be attributable to the trial therapy, then an AE form should be completed. If the pain is not considered to be attributable to the trial intervention, then it is not necessary to complete an AE form but an e-mail trail of the discussion with the clinician should be kept and stored.

     • If there is pre-existing pain at the start of the trial, this should not be recorded on the pain and fatigue form; a record should be made of this pain only if it changes in intensity or nature.

9. Use of ‘special’ equipment and ‘middle therapies’ that could constitute a different intervention.

   • It was agreed that neither pulleys, nor FES, nor water bath/hydrotherapy should not be used for MPT or FST exercises.

   • TheraBand^® (TheraBand, Akron, OH, USA) is acceptable for strength training as part of FST.

   • Throwing and catching a ball is an acceptable activity for FST or MPT; however, the use of the ball and progression of exercise would be subtly different for each intervention.

Training evaluation: research therapists, UEA, 7 November 2013

After the training had been completed in November 2013, the research therapists and trial manager were asked to summarise what they had gained most from the training. The following were the verbal responses received:

• Greater understanding of the trial interventions – able to now differentiate between MPT and FST. This was reassuring and insightful.

• Astounded by the quality of the team.

• Reaffirming similarities of therapy between centres.

• Team building – knowing everyone better.

• Getting to know people; links for further discussions.

• Discussion at length to clarify treatment – clinically reason choices.

• Clarification re MPT and function and scope of what can be done.

• New ideas for new members of the team.

• Insight into MPT.

• Increased scope of what can be done in MPT and FST.

• Clarification re paperwork, especially AEs.

• Focus on feedback given to patients.

• Increased treatment options.

FAST-INdICATE training: London, 6–7 July 2015