Methods
Design
The FOCUS, AFFINITY and EFFECTS trials are multicentre, parallel-group, double-blind, placebo-controlled trials with broad entry criteria and follow-up to ascertain the primary and secondary outcomes at about 6 and 12 months (Fig.
1). This section describes the core protocol that the three trials share and the variations adopted to facilitate each trial in its national setting. The following description reflects the versions of the trial protocols in use on 1 March 2015.
Start-up phases
Each trial has completed a start-up phase to establish whether the protocol is feasible in each setting and to establish the trial management teams, IT systems to manage web-based randomisation, drug allocation, stock control, follow-up, data collection and verification, and important aspects of feasibility including recruitment, medication adherence, questionnaire completion and follow-up rates.
Main phase
The trials are powered to detect differences in a primary outcome based on an ordinal analysis of the seven-category modified Rankin scale (mRS 0, 1, 2, 3, 4, 5, 6) for the entire group [
26]. Because it is not feasible to enrol sufficient patients in each trial to reliably detect small effect sizes that would still be of clinical significance we plan to perform an individual patient data meta-analysis including the data from FOCUS, AFFINITY and EFFECTS. This will allow us to provide the most precise estimates of any risks and benefits to detect a smaller overall effect size than those detectable by the individual trials and also to determine the effects in subgroups.
Patient population
Patients will be identified by participating clinicians from in-patient stroke services and outpatient clinics in the UK (FOCUS), Australasia and Asia (AFFINITY) and Sweden (EFFECTS).
Our inclusion criteria are:
-
Males and females aged ≥18 years
-
A clinical stroke with brain imaging that is compatible with intracerebral haemorrhage or ischaemic stroke (including those with normal CT scans)
-
Randomisation can be performed between 2 and 15 days after stroke onset
-
Persisting focal neurological deficit is present at the time of randomisation severe enough to warrant treatment from the patient’s or carer’s perspective.
Our exclusion criteria are:
-
Subarachnoid haemorrhage (except where secondary to a primary intracerebral haemorrhage or enrolling in AFFINITY)
-
Unlikely to be available for follow-up for the next 12 months, e.g. no fixed home address
-
Unable to speak English (FOCUS) or Swedish (EFFECTS) AND no close family member available to help with follow-up forms
-
Other life-threatening illness (e.g. advanced cancer) that will make 12-month survival unlikely
-
History of epileptic seizures
-
History of allergy to fluoxetine
-
Contraindications to fluoxetine including:
-
Hepatic impairment (alanine aminotransferase >3 upper normal limit)
-
Renal impairment (creatinine levels >180 micromol/l and in AFFINITY also eGFR <30 ml/min/1.73 m2)
-
Hyponatraemia (sodium < 125 mmol/l) in AFFINITY
-
Pregnant or breast-feeding women of child-bearing age not taking contraception. Minimum contraception is an oral contraceptive
-
Previous drug overdose or attempted suicide
-
Already enrolled into a CTIMP
-
Current or recent (within the last month) depression requiring treatment with an SSRI antidepressant. AFFINITY also excludes patients requiring treatment or currently treated with any antidepressant
-
Current use of medications that have a serious interaction with fluoxetine
-
Use of a monoamine oxidase inhibitor (MAOI) during the last 5 weeks (e.g. phenelzine, isocarboxacid, tranylcypromine, moclobemide selegiline and rasagiline)
-
Pimozide
-
AFFINITY also excludes those taking tramadol unless the person is willing to stop
-
AFFINITY specifically excludes patients with a diagnosis of bipolar disease and patients receiving treatment with an antipsychotic medication or tamoxifen
-
EFFECTS excludes patients who are unable to consent for themselves, FOCUS and AFFINITY allow consent by a proxy, and AFFINITY allows waiver of consent in specific circumstances.
Co-enrolment
Inclusion in another research study does not automatically exclude a patient from participating in these trials. As long as inclusion in the other study would not confound the results of the trials or make attribution of adverse reactions difficult, co-enrolment is permissible. However, if a participant has already been enrolled into another CTIMP, they cannot be enrolled into the trials. If a patient is enrolled into one of the trials, they may not subsequently be enrolled into another CTIMP. Also, local researchers must avoid overburdening patients.
Randomisation
Having obtained consent, the randomising person collects the baseline data (see Table
1) on a randomisation form and enters these data into a trial-specific computerised central randomisation service by means of a secure 24/7 Web interface or a telephone call to the trial office during office hours. After the computer programme has checked these baseline data for completeness and consistency it allocates that patient a unique study identification number and a treatment pack number that corresponds to either fluoxetine or placebo. The trial-specific system applies a common minimisation programme to achieve a balance of four factors:
Table 1
Baseline data collected prior to randomisation in the three trials
Eligibility confirmed | + | + | + |
Consent confirmed | + | + | + Patient only |
Participant’s names | + | + | + |
Date of birth | + | + | + |
Gender | + | + | + |
Ethnicity | + | + | + |
Living arrangements | + | + | + |
Employment | + | + | + |
Co-morbidities (existing) | | | |
• Previous ischaemic stroke/TIA | + | + | + |
• Previous intracranial bleeding | + | + | + |
• Coronary heart disease | + | + | + |
• Current or past depression | + | + | + |
• Diabetes | + | + | + |
• Gastrointestinal bleeding | + | + | + |
• Hyponatraemia | + | + | + |
• Fractures | + | + | + |
Current medication | + | + | + |
NIHSS including subsections [ 28] | + | + | + |
Prior independence | 1 Question | mRS | 1 Question |
Ability to walk alone | + | + | + |
Ability to lift both arms | + | + | + |
Post-stroke disability (smRSq) [ 31‐ 33] | | + | |
Patient health questionnaire (PHQ) [ 43, 56] | 2-Question version | 9-Question version | 2-Question version |
Haemorrhage on brain imaging? | + | + | + |
OCSP classification for ischaemic stroke [ 57] | + | + | + |
Modified TOAST classification [ 58] | + | + | + |
Renal and liver function test results | | + | |
Contact details to facilitate central follow-up | + | | + |
Unique identifier to facilitate central follow-up | + | | |
Following randomisation, the trial co-ordinating centre (FOCUS, EFFECTS) or randomising centre (AFFINITY) generates and sends a letter to inform the General Practitioner of the patient’s enrolment in the trial, including a copy of the consent form, follow-up schedule and advice about treatment of depression in patients participating in the trials.
Treatment allocation
The minimisation algorithm randomly allocates the first patient to a treatment, but allocates each subsequent patient in a ratio of 1:1 to the treatment that leads to the least difference between the treatment groups with respect to the prognostic factors [
29]. To ensure that we retain a random element to treatment allocation, patients are allocated to the group that minimises differences between groups with a probability of 0.8. The systems contain a list of treatment codes for each centre and that match the stock of IMP held at that centre. At the end of the session each patient is allocated a treatment code that corresponds to either an active (fluoxetine 20 mg once daily) or placebo treatment pack with a 6-month supply of capsules.
Patients are prescribed the study medication (20 mg capsule of fluoxetine or placebo capsule) to be taken daily at a time that is likely to maximise their adherence, i.e. linked to an activity of daily living. If the patient is unable to swallow capsules and has an enteral feeding tube in place then the capsules may be broken open and the contents put down the tube according to accepted methods [
30].
Blinding
The patient, their families, the healthcare team including the pharmacist and anyone involved in outpatient assessments are blinded to the treatment allocation. Emergency unblinding systems are available for each trial. The relevant chief investigators will decide on a case-by-case basis whether unblinding is required to ensure patient safety.
Follow-up
Participants are followed up in all three trials at 6 and 12 months to collect the primary and secondary outcomes. However, the trials vary in the timing, frequency and method of monitoring the patients’ progress (Table
2).
Table 2
Study assessment schedules
Consent and randomise | x | | | | | | | |
Contact details | x | f,e | | a | | | | |
Living circumstances | x | f | | x | e | x | | x |
Training (physio, etc.) | | | e | e | | e | | |
1 0 Outcome-smRSq (mRS) [ 26, 31‐ 33] | a | | | a | | x | | x |
Depression diagnosis | x | f,e | | a | x | x | | x |
| f,e | | | | | | | |
| a | | | a | | a | | a |
| | | | | | f,e | | f,e |
| | | | | e | e | | e |
Emotionalism | | | | | | e | | |
DSM IV for depression [ 42] | | | | | e | e | e | e |
| | | | | | x | | x |
Fatigue subscale SF36 [ 44, 45] | | | | x | e | x | | x |
| | | | | | a | | a |
| e | | | | | e | | |
| | | e | e | | x | | x |
EQD thermometer | | | | | | a | | a |
| | | | | | | | a |
Adverse events | | f,e | e | a,e | e | x | | x |
Adherence to IMP | | f | e | a,e | e | x | | |
All medications | x | f | | a | | x | | a |
Retrieve residual capsules (pill count) | | | | | e | x | | |
Physical therapy received | | | | e | | | | e |
Resource use over 12 months | | | | | | | | x |
At early follow-up, during the index admission and in the first month to identify adverse events, monitor adherence is carried out by the local centres in all three trials. However, each trial’s national coordinating centres follow up the patients at 6 and 12 months with postal and telephone questionnaires to measure the primary and secondary outcomes. Data are also collected from general practitioners and by data linkage to mortality and hospital admission data in all three trials. The reasons why patients stopped taking the trial medication will be recorded.
Primary and secondary outcomes
The primary outcome is functional status, measured with the modified Rankin scale (mRS) [
26] at the 6-month follow-up. We are using the simple modified Rankin scale questionnaire [
31‐
33] delivered by postal questionnaire, or via interview over the telephone or face to face to determine the mRS.
Secondary outcomes
-
Survival till the end of the trial. This will be determined by following patients up for 12 months through their GPs and telephone and postal questionnaire and thereafter through linkage to routine mortality data
-
Functional status (mRS) at 12-month follow-up
-
Health status with the Stroke Impact Scale (SIS) (for each of nine domains on which the patient scores 0–100) [
34‐
36]
-
Arm, hand, leg and foot strength
-
Hand function
-
Mobility
-
Communication and understanding
-
Memory and thinking
-
Mood and emotions
-
Daily activities
-
Participation in work, leisure and social activities
-
Overall rating of recovery
-
Adverse events/outcomes
-
Depression. Although the SIS includes a domain reflecting mood, in the trials we are collecting additional information on the diagnosis and treatment of depression during follow-up. Participants are asked if they have been diagnosed with depression since their last assessment, whether this has been treated and whether they have been started on an antidepressant medication. Mood is assessed during follow-up in FOCUS and EFFECTS with the Mental Health Inventory 5 [
37‐
39]. In addition EFFECTS uses the Montgomery-Åsberg Depression Rating Scale (MADRS) and patients scoring high have a diagnosis of depression confirmed based on the DSM-IV criteria [
40‐
42]. The PHQ-9 [
43] is administered at baseline (covering the 4 weeks before stroke), 1, 3, 6 and 12 months in AFFINITY
-
Recurrent stroke including ischaemic and haemorrhagic strokes
-
Acute coronary syndromes
-
Epileptic seizures
-
Episodes of hyponatraemia (<125 mmol/l)
-
Upper gastrointestinal bleeding
-
Other major bleeds (lower GI, extracranial, intracranial but extracerebral)
-
Poorly controlled diabetes including hyperglycaemia (>22 mmol/l) and hypoglycaemia
-
Falls resulting in injury
-
New fractures
-
Attempted suicide/self harm
-
Death
-
Fatigue (Vitality subscale of SF36) [
44,
45]
-
Cognition—the SIS, which incorporates an assessment of memory and thinking, is used for all three trials. In AFFINITY cognition during follow-up is assessed with the Modified Telephone Interview for Cognitive Status (TICSm) [
46]. EFFECTS assess cognition with the Montreal Cognitive Assessment (MoCA) [
47]
-
Health-related quality of life measured with the five-level Euroqol 5D (EQ5D-5 L) to generate utilities [
48]
Each trial is collecting data about resource use over the first 12 months to enable us to carry out health economic analyses.
Provisional analysis plan
A detailed analysis plan for each trial, and for an individual patient data meta-analysis, will be developed and reported by the chief investigators and an independent statistician prior to the database being locked at the end of follow-up for final analysis.
The primary analyses will retain patients in their original assigned treatment groups. We will use ordinal regression to compare functional status (mRS scores) at the 6-month follow-up, adjusted for those factors included in our minimisation algorithm [
49].
In secondary analyses we will compare the two treatment groups with respect to the following outcomes at 6 and 12 months:
There is evidence that functional outcome at 6 months post stroke is strongly associated with long-term survival [
50]. Therefore, if fluoxetine treatment is associated with improvements in functional status at 6 months it would be important to establish whether this translates into longer survival.
Subgroup analyses
The functional status (mRS) at 6 months will be compared with ordinal regression in the following subgroups:
-
Age (≤70, > 70 years)
-
Baseline probability of a good outcome on mRS calculated with the six simple variable model [
27]—to see whether effects remain constant across the range of stroke severities (<0.15 vs. 0.15-1 probability of being alive and independent at 6 months)
-
Ischaemic vs. haemorrhagic stroke
-
Patients who were unable to consent for themselves since this subgroup will allow us to answer the question whether routine use of fluoxetine is likely to benefit patients in whom a formal assessment of mood is impossible because of communication and cognitive problems.
In addition we are particularly interested to know whether the effect of treatment on neurological function is modified by specific neurological deficits present at baseline. Because patients may have a combination of neurological deficits, individual patients may appear in more than one subgroup
-
Subgroup 1: Patients with a motor deficit affecting the face/arm or leg
-
Subgroup 2: Patients with aphasia
The functional status (mRS) at 12 months will be compared with that at 6 months to establish whether any benefits observed at 6 months are maintained. If no difference is observed in the functional status between the treatment groups at 6 months, then secondary analyses will aim to establish whether there are differences with respect to the secondary outcomes, and if so whether these are maintained at 12 months.
We will also perform analyses of potential mediating factors, e.g. the role of depression. We will seek to answer the question whether any benefits are mediated by improvement in mood (based on MHI5, MADRS or PHQ9) and also whether any apparent loss of benefits in mRS or SIS between 6 months to 12 months is because of a deterioration in mood.
We envisage that levels of missing data in the primary outcome will be exceedingly low from previous experience of acquiring the mRS by postal and telephone questionnaire [
51‐
55] and the primary analysis will be a complete case analysis. If we see higher levels of missing data than expected, we will use a suitable analysis, based on the likely missing data mechanism. We will consider whether to extend missing data methods to secondary outcomes at a blinded review of the Statistical Analysis Plan immediately before the database lock.
Economic analyses
Each trial will develop a detailed health economic analysis plan before completion of data collection to determine the cost-effectiveness of fluoxetine in the trial setting.
Sample size/power calculations
FOCUS, AFFINITY and EFFECTS are planning to enrol at least 3000, 1600 and 1500 patients respectively. Table
3 shows the effect sizes [expressed as a common odds ratio (COR)]. These numbers would provide 90 % power with an alpha of 0.05. These estimates are based on the distribution of outcomes in the seven categories of the mRS (0–6) (observed in both treatment groups combined amongst the first 451 patients enrolled and followed up at 6 months in the FOCUS trial).
Table 3
Sample size calculations derived from an ordinal regression and based on 90 % power and alpha 0.05
EFFECTS | 1500 | 1.35 | 49.5 | 42.1 | 7.4 |
AFFINITY | 1600 | 1.34 | 49.4 | 42.2 | 7.2 |
FOCUS | 3000 | 1.23 | 48.4 | 43.2 | 5.2 |
Pooled | 4500 | 1.19 | 48.0 | 43.6 | 4.4 |
Pooled | 6000 | 1.16 | 47.7 | 44.0 | 3.7 |
If FOCUS, AFFINITY and EFFECTS combined enrol 6000, we would have 90 % power (alpha 5 %) to detect a COR of 1.16, equivalent to a 3.7 % absolute difference in percentage with mRS 0–2 (44.0 % to 47.7 %).
The trial steering committees (TSC) will regularly review the target sample size and adjust this based on accruing blinded data on
-
the enrolment into specific pre-specified subgroups
-
completeness of follow-up
-
distribution of mRS categories in the population of enrolled subjects (i.e. both treatment groups combined), overall and in specific patient categories (e.g. those with motor deficits, aphasia, etc.)
For example, if the distribution of mRS is different from that anticipated, then the sample size might need to be increased to maintain the power of the trial to detect the specified effect size. This approach has the advantage that such sample size adjustments can be made without reference to the accumulating blinded data and avoid the need for conditional power calculations, which can be unreliable.
Study funding
The three trials are funded by grants from charitable organisations and government funding bodies (see acknowledgements for details). They do not receive any funding from the pharmaceutical industry.
Ethics approvals
Each trial has received approval for its protocol and trial materials from the relevant local ethics committees and regulatory authorities in their respective countries [FOCUS: Scotland A Research Ethics Committee (for UK) (21/12/2011); AFFINITY: Western Australia, Royal Perth Hospital Human Research Ethics Committee (HREC) (24/02/2012), New South Wales, Victoria & Queensland, Western Sydney Local Health District HREC (30/04/2013), South Adelaide Clinical HREC (01/09/2014), New Zealand, Central Health and Disability Ethics Committee (17/04/2014)]. EFFECTS: Stockholm Ethics Committee (30/09/2013). No centres can start recruitment until they have received relevant ethics and regulatory approvals. Informed consent is obtained before the patient is enrolled except where it has been waived. Consent procedures had to comply with national requirements, so that in FOCUS and AFFINITY approval was obtained for consent by either the patient or proxy, AFFINITY also has approval for waiver of consent, whilst in EFFECTS patients have to be capable of consenting for themselves. The intensity and methods of monitoring were agreed between trial investigators and the relevant organisation in each country.
Organisation
Each trial has established its own Trial Steering Committee (TSC) to oversee the conduct and progress of the trial. Each trial has its own independent Data Monitoring Committee (DMC) to oversee the safety of participants in the trial. During recruitment, interim analyses of the baseline and follow-up data will be supplied, in strict confidence, to the chairmen of the data-monitoring committees, along with any other analyses that the committees may request. In the light of these analyses, the data-monitoring committee will advise the chairmen of the Trial steering committees whether, in their view, the randomised comparisons have provided (1) 'proof beyond reasonable doubt' that for all, or some, the treatment is clearly indicated or clearly contra-indicated and (2) evidence that might reasonably be expected to materially influence future patient management. Following a report from the DMC, the steering committees will decide whether to modify entry to the study (or seek extra data). Unless this happens, however, the TSC, the collaborators and central administrative staff will remain ignorant of the interim results.
The terms of reference of the DMCs specify that any of the chairmen should confer with the chairmen of the DMCs of the other trials if they have concerns about the accruing data. The chairmen may elect to share blinded or unblinded reports and to request combined analyses of accruing data in the three trials.
The FOCUS trial is coordinated from the Edinburgh Clinical Trials Unit (ECTU) (Neurosciences Division) and sponsored by ACCORD (a joint organisation between University of Edinburgh and NHS Lothian). The AFFINITY trial is coordinated from the Stroke Research Unit in Perth (School of Medicine and Pharmacology, University of Western Australia, and Sir Charles Gairdner Hospital, Health Department of Western Australia) and the Neurological & Mental Health Division, George Institute for Global Health, in Sydney, Australia. The EFFECTS trial is coordinated from the Karolinska Institutet in Stockholm but its randomisation and data storage are hosted by the ECTU in Edinburgh.
Discussion
The investigators of these three trials agreed to a collaborative approach including adhering to a common core protocol and combining their results to allow a prospective individual patient data meta-analysis. We judged that a larger number of participants would be required to answer our secondary research questions than could be recruited in any trial within a reasonable time period if based in one country. Also, we were keen to have sufficient power to ensure we detected even a modest overall effect, since this would still have important implications for patients, their families, and health and social services. The degree of collaboration is illustrated by the fact that the co-chief investigators of each trial participate in each of the three trial steering committees and have, where allowed, been named as co-applicants on each of the funding applications. This means that each of us has an interest in each of the three trials succeeding.
An alternative strategy that we considered was to establish a single international trial. However, we decided that our chosen strategy has several advantages.
We are able to vary trial procedures to fit local conditions. For instance, in the UK, the FOCUS trial is supported by a well-organised network of principal investigators and research staff in local centres funded by the National Institute of Health Research. In Australia and Sweden the individual trialists have to identify the local support networks. Regulations vary between countries with respect to whether proxy consent is acceptable and how patient identifiable data can be shared between centres and coordinating centres, which influences the feasibility of centralised follow-up and the intensity of monitoring required by regulators.
Ideally, the three trials would have collected identical data items at the same time points. Since the vast majority of data items are common to all three trials our approach did allow for some variation. For instance the PHQ9 was favoured in Australasia whilst the MHI5 was favoured in the UK and the MADRS in Sweden. These small variations should not preclude a valid combined analysis. Most importantly the eligibility criteria, method of randomisation, interventions and principal outcome measures were almost identical.
Our model allows us to confirm the results of one trial in another two providing valuable evidence of external validity. It will also allow us to explore any heterogeneity in trial results that may reflect the settings in which each trial was carried out. For instance, the intensity of background physical therapies, adherence to the trial drug, duration of hospitalisation and costs of delivering hospital and social services vary between countries and may alter the effectiveness or cost effectiveness of fluoxetine in treating stroke patients.
Our model negates the need to gain approvals for transporting supplies of active drug and placebo across international jurisdictions. It allowed the investigators in each country to apply for, and secure, funding for their own trial, which has provided more resources than were likely to be available from any one funding agency. Governmental and charitable funding agencies often put limits on how much of their grant funding can be spent abroad. Our approach also encouraged local ownership of the trial, which will hopefully facilitate faster recruitment and more complete follow-up.
Our DMCs are also collaborating to maximise patient safety. If safety concerns are raised as data accrue within one trial, the DMC charters encourage the chairs to share data for confirmation of a problem or for reassurance.
We urgently need new treatments to reduce the burden of disability after stroke. These three multicentre, investigator-led, charity- and government-funded randomised trials will determine whether the routine administration of fluoxetine 20 mg daily (for 6 months) in stroke survivors between 2 and 15 days after stroke improves recovery at 6 months, whether any benefits persist after fluoxetine has been discontinued and whether there are benefits in specific subgroups. If fluoxetine is effective, the results of the trial could be rapidly implemented throughout the world at very little cost to health services.
Acknowledgements
The start-up phase of FOCUS was funded by the UK Stroke Association (TSA 2011101) and the main phase is funded by the NIHR Health Technology Assessment Programme (project no. 13/04/30). The views and opinions expressed herein are those of the authors and do not necessarily reflect those of the NIHR Health Technology Assessment Programme. The start-up phase of AFFINITY was funded by a National Health and Medical Research Council Programme Grant (application ID: APP1013612) and a University of Sydney bridging support grant. The main phase is funded by a National Health and Medical Research Council Project Grant (APP1059094), Australia. The views and opinions expressed herein are those of the authors and do not necessarily reflect those of the NHMRC. During the completion of this work, Maree L. Hackett was in receipt of a National Heart Foundation Future Leader Fellowship 100034. The start-up phase of EFFECTS was funded by the Konung Gustaf V:s och Drottning Victorias Frimurare, Hjärt-Lungfonden and Stroke Riksförbundet and the main phase by Vetenskapsrådet, the Swedish research council (application ID: 921-2014-7072).
A large number of people have already contributed to the development of the protocol and establishment of these trials (see Additional file
1: Appendix). In particular we would like to thank Catriona Graham for producing the combined sample size calculations (Table
3). The NIHR research networks in the UK are central to our efforts to meet recruitment targets in FOCUS. Our collaborations are growing as more centres join the trials and we would like to acknowledge their contributions.
On behalf of the FOCUS (Fluoxetine Or Control Under Supervision), AFFINITY (Assessment oF FluoxetINe In sTroke recoverY) and EFFECTS (Efficacy oF Fluoxetine – a randomisEd Controlled Trial in Stroke) collaborative groups (see Additional file
1: Appendix).
Competing interests
The chief investigators of these three trials declare no competing interests other than having received grant funding to support the trials. None of the authors have any non-financial competing interests.
Authors’ contributions
GM contributed to the design of the trials, carried out the systematic review of the previous trials in humans, which underpinned the rationale for these trials, provided details of FOCUS and critically appraised the drafts. MH contributed to the design of the trials, carried out the systematic review of the previous trials in humans, which underpinned the rationale for these trials, provided details of AFFINITY and critically appraised the drafts. EL contributed to the design of the trials, provided details of EFFECTS and critically appraised the drafts. VM contributed to the design of the trials, provided details of EFFECTS and critically appraised the drafts. GH contributed to the design of the trials, provided details of AFFINITY and critically appraised the drafts. MD contributed to the design of the trials, provided details of FOCUS and wrote the first draft of the combined manuscript. All of the authors, apart from VM (deceased), have read and approved the final manuscript.